According to McKinsey, 89 % of large companies globally have a digital and AI transformation underway. They have captured 31 percent of expected revenue lift and 25 percent of expected cost savings. The gap is not explained by insufficient investment, inadequate technology, or lack of ambition. It is explained by a structural problem that most transformation programs are not designed to address: the absence of deliberate coordination across systems, data, and business processes before scale is attempted.

The organizations that have made measurable progress share a recognizable pattern. They treat digital transformation not as a technology program but as an architecture challenge — one that requires specific foundations to be in place before meaningful capability can be built on top of them.

How Architectural Discipline Works in Practice

Shell offers one of the clearest documented examples of what this looks like at enterprise scale.

Their approach is explicit about sequencing. Documented accounts of Shell’s data strategy describe starting advanced analytics work as early as 2012 by building data platforms, data governance, and data quality infrastructure first — before focusing on what the data would produce. Their CIO has articulated the broader shift: five to ten years ago, they developed IT strategies; now they develop business strategies that are digitally enabled. Their Enterprise Cloud Platform team explicitly credits a pivot from technology-centric thinking to standardized enterprise capabilities as the key success factor in their cloud transformation.

The specifics are documented. Shell is a founding member of OSDU — the Open Subsurface Data Universe — an industry standard for subsurface data management. They built and open-sourced their core data platform via the Linux Foundation Energy. They operate over 100 AI applications on a common enterprise platform rather than as isolated deployments. A documented case study of their cloud platform transformation recorded a 77 percent productivity increase after full deployment of their standardized enterprise approach. Equinix has documented a similar philosophy — prioritizing digital infrastructure and platform coherence before attempting capability scale.

What distinguishes these organizations is not the technology they chose. It is the order in which they built things, and the discipline they applied to ensuring each layer was in place before the next was attempted. Data foundation before analytics. Integration architecture before application deployment. Platform standards before scale.

That sequencing is the discipline most organizations skip — and skipping it is the most reliable predictor of transformation stall.

The Coordination Gap

The failure mode is structurally consistent. An organization announces transformation objectives. Business units interpret this as permission to move independently. Cloud migration, IoT investment, new applications, and AI experiments launch in parallel. Each initiative progresses against its own mandate. The coordination failure surfaces around month twelve — when the AI pilot cannot access production data, the new application cannot surface information owned by a team that was never consulted, and integration costs have compounded beyond what was budgeted.

This is not a technology failure. It is a coordination failure — predictable when transformation is designed as a portfolio of independent projects rather than a systematically sequenced program.

Bain’s 2024 analysis found that 88 percent of business transformations fail to achieve their original ambitions. Gartner attributes a significant share of digital transformation failure specifically to poor data governance. The consistency of these numbers across industries and geographies points to a structural diagnosis, not an execution one.

The Five Transformation Pillars

The organizations that build and sustain transformation momentum share a common architectural foundation. The five pillars below represent what needs to be in place before transformation programs can deliver at scale. They function as a readiness assessment: organizations can evaluate which pillars are established, which are partial, and which are absent. The absence of two or more is the most reliable indicator that scale will not be achieved.

Strategic alignment. Technology investment is explicitly connected to business capabilities and strategic priorities. Business units are not interpreting transformation independently — they are working from a shared enterprise view of where the portfolio is heading and what each initiative is meant to enable.

Data foundation. Data governance, data quality, and integrated data architecture are in place before advanced capabilities are attempted. This is the foundation Shell built first. It is also the foundation most organizations attempt to build retroactively — after AI pilots have stalled and the root cause has been diagnosed.

Integration architecture. An enterprise integration framework exists before applications are deployed. Standard patterns for how systems communicate — APIs, data contracts, messaging standards — are defined and enforced. Without this, each new system creates new integration requirements solved independently, and the cost compounds with every addition to the portfolio.

Sequenced investment. Initiatives are launched in dependency order, not political visibility order. Infrastructure before applications. Data foundation before AI. Integration layer before new platforms. This sequencing is the mechanism by which organizations like Shell avoided the compounding remediation costs that characterize most large-scale transformation programs.

Enabling governance. Architecture principles, standard patterns, and a lightweight review process exist to guide technology decisions across the enterprise. The goal is not control — it is consistency. Well-designed governance reduces decision paralysis, accelerates delivery, and prevents the accumulation of technical debt that eventually consumes the capacity for innovation.

These five pillars are interdependent. A strong data foundation without integration architecture means data remains siloed. Integration architecture without enabling governance degrades as teams build around standards rather than within them. The value is in the combination, not any single component.

What EA Actually Requires

Enterprise architecture is not an IT function that sits upstream of project delivery. Done well, it is a business capability — the organizational capacity to think about how technology investments relate to each other and to strategy, at enterprise scale, over time.

Three things are consistently underestimated in practice.

Executive sponsorship with authority. EA crosses organizational boundaries and challenges existing autonomy. A team with strong frameworks but no authority to act on them becomes a documentation exercise. McKinsey’s research consistently identifies leadership alignment as one of the highest-leverage factors in transformation success. Without C-level sponsorship, the five pillars cannot be built — each requires investment that does not produce visible output before it enables everything else.

A bias toward enabling over gatekeeping. The most common way EA programs fail is by becoming bottlenecks. Architecture reviews that slow projects without adding value lose legitimacy quickly. Shell’s platform approach demonstrates the right model: standardized foundations that give teams freedom to build on top, not a review board that controls every decision. The goal is to help teams make faster, more consistent decisions — not to centralize judgment.

Communication as a core discipline. Gartner research finds that projects with poor technical-to-business communication are 67 percent more likely to exceed budget and 89 percent more likely to miss strategic objectives. The most common EA failure is not technical — it is the inability to make architectural reasoning legible to executive decision-makers. Translating the five pillars into business-language evidence — cost of integration remediation, AI pilots that failed to scale, redundant investments discovered late — is the work that creates organizational permission to act.

Implications by Role

The five pillars surface differently depending on where you sit. The assessment is only useful if it leads to action — and the action differs by role.

For transformation leaders and CIOs. Run the readiness assessment honestly before the next program is funded. Which pillars are genuinely in place? Partial foundations are more dangerous than absent ones — they create the illusion of readiness without the substance. Data foundation and integration architecture are almost always the right starting point. They are prerequisites for everything else and the most common source of scale failure when absent.

For enterprise architects. The framework is a communication tool as much as a diagnostic one. The inability to make the case for foundational investment in business terms — not architectural terms — is the most common reason EA programs lose organizational support. Each pillar has a business-language evidence set: cost of retroactive integration, AI programs that could not reach production, redundant capabilities discovered after investment. That evidence is the lever for securing the sponsorship and budget that EA work requires.

For program and portfolio leaders. Sequencing inversion is largely a portfolio governance problem rather than an architectural one. Individual program teams sequence based on their own constraints. No one owns the dependency map across the portfolio. The fix requires someone with authority over the enterprise sequencing view and the mandate to delay a program being launched out of order. Without that, the five pillars cannot be maintained even when they are built.

The Question Worth Sitting With

The five transformation pillars address the coordination problem that most digital transformation programs refuse to name. But they introduce a tension that is genuinely unresolved: how do you maintain architectural coherence when the technology landscape — particularly AI — is moving faster than any governance model was designed to handle?

The organizations that built strong architectural foundations through the first generation of digital transformation are now facing a second-order challenge. The data platforms, integration layers, and cloud architectures built with discipline are already being stress-tested by agentic AI, real-time edge computing, and distributed data models that were not in scope when those foundations were designed.

The question is not whether the foundational approach was right. The evidence on that is settled. The question is whether enterprise architecture as a discipline can evolve fast enough to govern what comes next — or whether the governance models themselves will become the next inherited structure that organizations have to work around.

The transformations that succeed over the next decade will be the ones where that question stays live.

Sources & Further Reading

On transformation performance

• McKinsey & Company — How top-performing companies approach digital transformation (Smaje & Zemmel, March 2024)

• Bain & Company — 2024 analysis on business transformation success rates

• Gartner — Digital transformation failure rates and the role of data governance

On Shell’s approach

• Equinix Digital Leaders Summit — Digital transformation insights from Shell and Equinix

• Highberg — Productization in Shell’s Cloud

• CGI — Shell’s strategy for a business-led IT organization

On EA and communication

• Gartner — Research on technical-to-business communication and project outcomes

Foundational reading

• Ross, Weill & Robertson — Enterprise Architecture as Strategy (Harvard Business Press)

• Gregor Hohpe — The Architect Elevator

• The Open Group — TOGAF Standard

Haider Ali writes on enterprise systems, digital transformation, and architecture at scale. He focuses on how large organizations align technology investment with business strategy.

Author of Unfinished

First Insight · LinkedIn · haiderali.co

On February 16, 2026, India hosted the first-ever global AI summit in the Global South. Over 100 countries sent delegations. More than 20 heads of state arrived. Tech CEOs from Google, OpenAI, Anthropic, and DeepMind gathered at Bharat Mandapam in New Delhi. The stated mission: translate AI discussions into development outcomes. The reality: a masterclass in the gap between aspiration and execution.

I’ve spent two decades designing systems that serve hundreds of thousands of users across distributed geographies. I’ve seen frameworks persist long after they’ve stopped serving their purpose—a pattern I explore throughout my book Unfinished. The India AI Impact Summit wasn’t just another conference. It was a live demonstration of what happens when inherited thinking about how summits “should” work collides with the complexity of actually making them work.

The Three Sutras: People, Planet, Progress

The summit anchored itself around three foundational pillars—Sutras, meaning guiding principles in Sanskrit. Each was meant to define how AI could be harnessed for collective benefit:

People: AI must serve humanity in all its diversity, preserving dignity and ensuring inclusion.

Planet: AI innovation must align with environmental stewardship and sustainability.

Progress: AI’s benefits must be equitably shared, advancing global development.

These weren’t empty platitudes. They translated into seven thematic working groups covering AI for economic growth, democratizing resources, social inclusion, safety and trust, human capital development, scientific advancement, and resilience. India announced initiatives like training 500 PhD scholars and 5,000 postgraduates in AI research. The country’s AI-powered technology sector projected revenues of $280 billion for 2025. Nearly 89% of new startups launched in 2024 integrated AI into their products.

The architecture was sound. The intentions were genuine. But architecture means nothing if the execution infrastructure can’t support it.

When VIP Culture Meets Scale

Here’s what actually happened: New Delhi’s already notorious traffic became completely gridlocked. Why? Because when dozens of heads of state and global CEOs need to move through a city simultaneously, police close roads entirely—a practice locals call “VIP movements.” Speakers missed their own sessions. Delegates spent hours stuck in traffic. Yoshua Bengio, one of AI’s “godfathers,” delivered his address via blurry video link from the Canadian embassy because he couldn’t physically reach the venue.

On day one, exhibitors were thrown out of the venue with no warning at midday to accommodate Prime Minister Modi’s visit. Gates closed until 6 PM. One founder had his display tech stolen during the chaos. People reported two-hour entry queues after three-hour drives. The overcrowded rooms, ever-changing entry policies, and poor communication infrastructure created what attendees described as a “third-class citizen” experience for anyone not classified as a VIP.

This is what I call the implementation paradox: the gap between what we design on paper and what actually works when humans encounter it at scale.

The Robot That Wasn’t

Perhaps the most revealing moment came on February 18. Galgotias University showcased a robot dog at their exhibition pavilion, presenting it as indigenous innovation. Social media users immediately identified it as the Unitree Go2—a commercially available product from Chinese company Unitree Robotics. The university apologized, claiming their representative was “ill-informed.” They were directed to vacate their stall.

The incident exposed something deeper than misrepresentation. It revealed the pressure to demonstrate innovation credentials on a global stage—and what happens when that pressure meets inadequate verification systems. In my work designing AI-augmented systems for enterprise environments, I’ve learned that the most critical failures aren’t technical. They’re process failures that allow unchecked claims to reach production.

What Actually Shipped

Despite the operational chaos, the summit produced tangible outcomes. Eighty-eight countries and international organizations signed onto a diplomatic declaration committing to inclusive AI development. India set a Guinness World Record with 250,946 pledges for an AI responsibility campaign in 24 hours—far exceeding the initial 5,000-pledge target.

Sarvam AI, an Indian laboratory, launched new language models including 30-billion and 105-billion parameter models using mixture-of-experts architecture. The Research Symposium on AI and its Impact brought together leading researchers to discuss sovereign AI infrastructure, global adoption challenges, and policy priorities. These weren’t performative announcements. They represented real progress in a country positioning itself as a key platform for shaping the global AI agenda.

The India AI Impact Expo featured over 300 exhibitors from 30 countries across more than 10 thematic pavilions. Applications spanned healthcare, agriculture, education, and sustainable industry. The event ran six days instead of five due to overwhelming public response.

The Unspoken Reality

Here’s what the official narratives won’t tell you: Amnesty International pointed out that while India was lauded for technological progress, human rights concerns around AI deployment in the country—including facial recognition and public sector automation that excludes marginalized communities—were “papered over.” The summit’s push toward sovereignty, innovation, and democratization, they argued, feeds a global trend of turning AI into a power accumulation race rather than collective action for rights protections.

This tension isn’t unique to India. It’s the fundamental challenge of AI governance: How do you balance rapid development with genuine safeguards? How do you ensure technology serves people when the very definition of “serving people” is contested terrain?

What This Tells Us About AI’s Future

The India AI Impact Summit revealed something more important than its documented achievements or failures. It showed us that the future of AI won’t be determined primarily by technical capabilities. It will be determined by our ability to translate high-level principles into working systems that serve real people in real contexts.

The summit’s title shifted from “AI Safety” (Bletchley Park, 2023) to “AI Action” (Paris, 2025) to “AI Impact” (New Delhi, 2026). According to legal analysts at Crowell & Moring, these changing titles reflect a broader shift away from governance toward practical implementation and measurable outcomes.

But implementation requires more than good intentions and impressive pavilions. It requires:

Infrastructure that matches ambition. You can’t host 100+ country delegations and dozens of VIPs without transportation systems that actually move people to where they need to be.

Verification systems that work. You can’t celebrate indigenous innovation without processes that catch misrepresentation before it reaches the exhibition floor.

Cultural awareness at scale. You can’t design for “People, Planet, Progress” while maintaining VIP cultures that leave most participants feeling like obstacles to be managed.

In my work architecting digital experiences that serve hundreds of thousands of users, I’ve learned that the hardest problems aren’t about technology. They’re about the inherited assumptions we carry about how things “should” work—even when those assumptions actively prevent things from working.

The Real Test

India’s next challenge isn’t launching more AI models or hosting more summits. It’s demonstrating that the principles articulated in those three Sutras can actually shape how AI gets developed and deployed—not just in policy documents, but in systems that real people use every day.

The summit succeeded in positioning India as a serious player in global AI conversations. It created space for Global South perspectives that are often sidelined in technology discussions dominated by American and Chinese companies. It generated commitments, showcased innovations, and set records.

But it also reminded us that designing for impact requires more than ambitious frameworks. It requires the unglamorous work of making systems actually function when they encounter scale, complexity, and the messy reality of human needs.

The Switzerland AI Summit scheduled for 2027 will be watching. The question isn’t whether they can avoid traffic jams and verification failures. The question is whether these gatherings can evolve beyond diplomatic theater toward something that genuinely advances AI in service of humanity.

Because right now, we’re still figuring out how to make the summit itself work—let alone the technology it’s meant to govern.

About the Author

I work at the intersection of design, technology, and human experience—crafting intelligent systems that amplify human capability rather than replace it. As a Digital Experience Design Architect, my practice is grounded in a belief that the most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and create.

My approach combines rigorous methodology with creative vision. I question assumptions, challenge conventional wisdom, and seek patterns that others might miss. Whether exploring user research methodologies, designing enterprise systems, architecting digital experiences, or examining broader societal challenges, I maintain a critical lens that asks not just “what works” but “why it works” and “for whom does it work best.”

Each article I write reflects this philosophy: technology should expand our creative horizons, design should serve genuine human needs, and innovation should be tempered with wisdom about its implications.

For more on AI implementation, enterprise design, and questioning established thinking:

• Subscribe to User First Insight

• Connect on LinkedIn

• Visit stayunfinished.com

There’s a particular irony in watching an energy company become an AI infrastructure powerhouse. While Silicon Valley races to build the next chatbot, Aramco—the company that has powered global industry for over 90 years—is quietly building something far more foundational: the actual infrastructure that makes AI possible at scale.

I’ve been watching this transformation from the inside as a Lead Digital Experience Design Architect at Aramco’s Saudi Accelerated Innovation Lab (SAIL), and what’s happening here challenges almost everything the tech industry assumes about digital transformation.

The Reality of AI Infrastructure

Aramco recently published their vision for building the AI future, and while the headlines focus on investments and scale, what struck me most was what wasn’t said: the complex, foundational reality of making AI actually work in industrial contexts.

The article mentions $1.8 billion in AI-driven Technology Realized Value in 2024 alone. But here’s what that number obscures: 442 identified use cases, with 200+ solutions deployed and 100+ still in development. That’s not a moonshot story. That’s the story of systematic, persistent implementation across one of the world’s most complex operations.

In Unfinished, I wrote about how real transformation happens through accumulated small decisions rather than singular breakthrough moments. Aramco’s AI journey is a perfect case study: AI models analyzing drilling plans, autonomous systems preventing failures, reservoir simulations informing engineering decisions. Each one unremarkable on its own. Collectively, they’ve helped maintain Aramco’s position as one of the lowest upstream carbon intensity producers globally.

The Infrastructure Problem No One Talks About

Here’s where it gets interesting: the International Energy Agency projects that AI and data centers will account for around 20% of total electricity growth in advanced economies between 2024 and 2030. Twenty percent.

This is where energy companies stop being legacy players and become essential enablers. Building intelligent systems requires more than algorithms—it requires reliable, affordable power at scales most tech companies have never had to think about. It requires physical infrastructure that doesn’t fail when training large language models for weeks at a time.

Aramco’s planned investment in HUMAIN—the PIF-owned AI infrastructure company—isn’t just about building data centers. It’s about constructing the complete, integrated digital foundation that will form vital national infrastructure: compute-as-a-service, cybersecurity, cloud services, and the power to run it all reliably.

This is the kind of foundational work that determines which AI ambitions succeed and which ones remain unrealized.

What Design Architects See That Others Miss

Working at the intersection of design, technology, and industrial-scale operations has taught me something crucial: the success of AI implementation has less to do with algorithmic sophistication and more to do with understanding context, constraints, and human capability.

Aramco’s commitment to training more than 6,000 AI developers through partnerships with Imperial College, Caltech, and KAUST reveals an understanding that many tech companies lack: people are the foundation of every successful digital transformation. Not just any people—specialists who understand both the global state-of-the-art and local context, who can build systems that work in specific conditions with specific constraints.

This mirrors what I explored in my work on enterprise-scale design challenges: you can’t design for generic “users.” You have to design for real people with specific contexts, needs, and capabilities. The same principle applies to AI deployment at industrial scale.

The Promise and the Reality

Aramco’s vision positions Saudi Arabia as a regional and global hub for AI excellence. Through Aramco Digital, they’re providing solutions in connectivity, cybersecurity, and cloud computing to empower the industrial sector and accelerate the Kingdom’s industrial transformation.

But here’s what I appreciate about their approach: it’s grounded in actual operational experience. When you’ve been managing some of the world’s most complex industrial operations for 90 years, you develop a certain pragmatism about technology. You understand that reliability matters more than novelty. That integration beats isolated brilliance. That the systems that persist are the ones designed for maintainability, not just launch-day demos.

What This Means for AI’s Future

The tech narrative around AI tends toward either utopian promises or dystopian warnings. What gets lost is the mundane reality: AI’s impact will be determined by infrastructure, implementation capability, and the unglamorous work of making systems actually function in real-world conditions.

Aramco’s journey from energy provider to AI infrastructure builder suggests a different future than the one Silicon Valley imagines—one where the companies that understand physical infrastructure, reliable operations, and long-term capability building become essential players in determining how AI shapes industries.

For those of us working at the intersection of design and technology, this is a reminder: the most important transformations aren’t the ones that make headlines. They’re the ones that build foundations that last.

Haider Ali is a Lead Digital Experience Design Architect at Aramco’s Saudi Accelerated Innovation Lab (SAIL) and author of Unfinished. His work explores the intersection of design, technology, and human experience, with a focus on building AI-powered systems that augment human capability rather than replace human judgment.

For more writing on design systems and intelligent interfaces, subscribe to User First Insight or explore Black & White. Connect on LinkedIn, follow on Medium, or visit haiderali.co and stayunfinished.com.

Read the original article: “Building the AI future”

The pitch deck was slick. Another AI prototyping platform claiming to transform prompts into production-ready interfaces faster than I could refactor a component library. As a Digital Experience Design Architect who’s spent years bridging the gap between human intent and digital execution, I’ve learned to approach such promises with measured skepticism. Yet the demos kept coming—venture-backed tools promising to compress weeks of wireframing into minutes of prompt engineering.

Rather than dismiss them outright or embrace them blindly, I decided to run these tools through their paces on a real project: redesigning the learner profile hub for our enterprise training platform. Not a hypothetical exercise or a marketing landing page, but a complex interface serving thousands of professionals navigating certification paths, accessing course materials, and tracking their learning journeys. The kind of nuanced challenge that exposes whether a tool delivers substance or just screenshots.

What emerged from weeks of systematic testing was neither the revolution promised nor the disaster skeptics predicted. Instead, I discovered a technology that mirrors our instructions with impressive fidelity while consistently missing the judgment calls that separate competent interfaces from exceptional experiences. This is that story—complete with the patterns I uncovered, the failures that taught me most, and a framework for integrating AI responsibly without abandoning the craft we’ve spent careers refining.

The Test: Real Work, Real Stakes

My evaluation wasn’t academic. The profile hub redesign carried real consequences—it would shape how thousands of learners interact with our platform daily. I needed to understand not just whether AI could generate interfaces, but whether those interfaces could handle the complexity of enterprise learning: multi-role permissions, progress tracking across certification paths, integration with live session scheduling, and the countless micro-interactions that keep learners engaged.

I approached this systematically, testing three distinct categories of tools against progressively detailed specifications:

The tool landscape ranged from pure design generators producing Figma-ready mockups to code-first platforms spinning up React components, plus conversational AI that could discuss, iterate, and refine designs through dialogue. Each promised a different flavor of acceleration—visual, functional, or collaborative.

The prompt progression started deliberately vague—just the purpose and audience—then evolved through detailed specifications with enumerated components and interaction states, finally culminating in prompts augmented with sketches, mockups, and existing design artifacts. This progression mirrored how real projects evolve from ambiguous briefs to concrete specifications.

The Revelation in Specificity

The relationship between prompt precision and output quality proved more dramatic than expected. With broad instructions, every tool defaulted to its most familiar patterns—social media profiles, marketing heroes, generic dashboards. The AI wasn’t creating; it was pattern-matching against its training data and serving up the statistical average of “profile page.”

But something shifted when I provided detailed specifications. Suddenly, the tools began inferring adjacent details I hadn’t explicitly requested. When I specified progress tracking for certification paths, several tools automatically included expiration warnings for time-sensitive credentials. When I described the need for upcoming session access, they added contextual preparation materials. These weren’t random additions—they reflected genuine understanding of the problem space.

The most striking results came when I included visual artifacts. A rough sketch transformed vague layouts into precise component arrangements. A mid-fidelity mockup ensured proper visual hierarchy. Yet this accuracy exposed an uncomfortable truth: by the time you’ve created detailed visual references, you’ve already done the hardest design work. The AI becomes a translation service, not a design partner.

The Persistent Gaps

Even the best outputs—those generated from detailed specs with visual references—felt technically correct but experientially hollow. They captured structure without soul, layout without logic.

Grouping and relationships consistently failed. Elements that belonged together drifted apart. The password for supplemental materials sat three columns away from the access link. Exam preparation resources scattered across disparate modules instead of forming a coherent study center. The AI followed instructions literally but couldn’t infer the narrative arc of user tasks.

Visual rhythm and emphasis proved equally challenging. Some outputs flooded interfaces with our brand colors, creating visual chaos where restraint was needed. Others delivered such low contrast that accessibility validators would have failed instantly. The tools understood color values but not color purpose—when to amplify, when to recede, when to guide attention versus when to maintain calm.

Interaction semantics revealed the deepest gaps. Disabled states looked nearly identical to active ones. Loading indicators appeared without context. Focus states violated keyboard navigation patterns. The syntax was correct—buttons had hover states, forms showed validation—but the semantics were broken. A learner encountering these interfaces would constantly second-guess whether the system was responding to their actions.

The Weight of Training Data

These shortcomings aren’t random; they’re artifacts of how AI learns. Most prototyping tools train on publicly available interfaces—marketing sites, SaaS dashboards, component libraries. They learn what appears most frequently, and that statistical bias shapes every output.

The result? A gravitational pull toward visual mediocrity. Sans-serif type, neutral grays, rounded corners, bright CTAs—the aesthetic of playing it safe. For enterprise brands investing millions in differentiated experiences, this homogenization is unacceptable. We’re not building another project management tool; we’re crafting learning experiences that need to feel distinctly ours.

Language ambiguity compounds the problem. When I wrote “profile hub,” some tools interpreted it as a public social profile, elevating bio sections and contact details while burying course logistics. The tools weren’t wrong—they were just mapping to the most statistically common interpretation of ambiguous terms. This forced me to write with defensive precision, anticipating every possible misinterpretation.

Finding the Sweet Spots

Despite these limitations, I discovered specific scenarios where AI prototyping delivers genuine value:

Divergent exploration benefits most. When I need to see twenty different approaches to information hierarchy, AI can generate that variety in an afternoon. I treat these outputs like a design sprint’s crazy eights—fuel for discussion, not final direction.

Stakeholder alignment accelerates dramatically. Executives struggle with static mockups but immediately grasp interactive prototypes. AI can transform a Figma board into a clickable demonstration that makes abstract concepts tangible, even if the details need refinement.

Testing scaffolds emerge quickly. Rather than hand-coding prototypes for usability testing, I can generate functional approximations that are good enough to validate core flows and gather user feedback. The fidelity isn’t production-ready, but it doesn’t need to be.

A Framework for Responsible Integration

Based on this evaluation, I’ve developed a framework for integrating AI prototyping without compromising design quality:

Immediate Actions (This Quarter)

• Develop prompt templates that encode your design system’s principles, accessibility requirements, and interaction patterns

• Create evaluation rubrics for assessing AI outputs against your quality standards

• Document failure patterns to build organizational memory about what these tools consistently miss

Building Capability (Next 6 Months)

• Run parallel tracks comparing AI-assisted and traditional design approaches on the same briefs

• Quantify the differences through usability testing, measuring task completion, error rates, and user satisfaction

• Identify the breakpoint where AI assistance shifts from accelerating to compromising quality

Strategic Evolution (12-18 Months)

• Co-develop your design system with AI consumption in mind—rich metadata, clear naming conventions, usage examples

• Build proprietary training sets from your successful projects to fine-tune models on your specific patterns

• Establish governance models that maintain human accountability while leveraging AI speed

The Leadership Imperative

For design leaders, AI prototyping presents both opportunity and obligation. The opportunity: to accelerate exploration and democratize certain aspects of interface creation. The obligation: to ensure that speed doesn’t compromise the judgment, empathy, and craft that define exceptional experiences.

This means evolving how we develop our teams. Prompt engineering becomes a new literacy—not replacing visual design or interaction principles, but extending them. We need designers who can articulate intent with precision, evaluate outputs critically, and know when to override the machine’s suggestions.

More importantly, we must celebrate and protect the human advantages. The ability to group related elements based on user mental models. The judgment to know when breaking consistency serves clarity. The empathy to anticipate confusion before users encounter it. These aren’t just nice-to-haves; they’re the differentiators between products that merely function and those that genuinely serve.

Beyond the Hype Cycle

After weeks of testing, my position is clear: AI prototyping belongs in our toolkit, but it doesn’t replace our toolkit. It can draft layouts faster than any human, generate variations at unprecedented scale, and translate specifications into code with impressive accuracy. But speed without judgment produces competent mediocrity—interfaces that look professional from a distance but frustrate users up close.

The future isn’t human versus machine but human plus machine, with judgment at the center. Our role as designers evolves from purely creating to orchestrating—knowing when to leverage AI’s speed, when to override its suggestions, and how to infuse outputs with the meaning and nuance that only human understanding provides.

The tools will improve. The models will learn better patterns. The interfaces will become more sophisticated. But the need for human judgment—for someone who understands not just how interfaces look but how they feel, not just what they display but what they mean—that need only grows stronger.

We’re not just designing interfaces; we’re designing experiences that shape how people learn, work, and grow. That responsibility demands more than statistical pattern matching. It demands the kind of thoughtful, empathetic, critical engagement that defines our craft. AI can assist that mission, but it cannot own it.

The design guardian must remain in the room.

This analysis emerges from ongoing work at the intersection of AI and enterprise design. For deeper exploration of these themes, subscribe to User First Insight, follow Black & White Perspective for broader systemic analysis, or read Unfinished: Notes on Designing Experience in a World That Never Stops Changing. The conversation continues at haiderali.co and stayunfinished.com.

I needed to cancel my RunwayML subscription. I’d signed up for their $15/month Standard plan to test some video work, and after finishing my project, it was time to move on. What followed wasn’t just frustrating—it was a masterclass in how broken UX patterns erode trust.

The Workspace plan details page showed a clear Cancel plan button positioned prominently alongside “Manage billing” and “Update plan.” I clicked it.

The system took me to a subscription management page that displayed my plan details, end date, and two action buttons: “Update subscription” and “Don’t cancel subscription.”

Notice what’s missing? An actual way to cancel.

The UI promised me a cancellation flow. The reality delivered a page designed to prevent exactly that action. After cycling through several screens—back to workspace settings, into billing information, hunting through payment methods—I still couldn’t find the off-switch. The “link” option next to the payment method suggested I’d need to navigate into a Stripe portal somewhere, but that path wasn’t marked or obvious.

This wasn’t just poor UX. It was a broken promise, executed in real-time.

The Anatomy of a Broken Promise

Over two decades of working in design and product development, I’ve witnessed the gap between what systems promise and what they deliver. This RunwayML cancellation flow exemplifies five critical failures that compound into user distrust:

Broken information scent. The button label says Cancel plan. The destination page offers Don’t cancel subscription. That’s a promise broken in one click. Information scent—the visual and textual cues that guide users toward their goals—works only when the trail leads somewhere. When it dead-ends with the opposite action, users don’t just feel lost; they feel manipulated.

Affordance mismatch. Primary CTAs signal irreversible actions. We’ve trained users for decades to expect that a prominent “Cancel” button means “end this now.” Using such buttons to navigate to a page that discourages cancellation violates that learned expectation. The cognitive load isn’t just about finding the right path—it’s about relearning what buttons mean in this particular system, and questioning whether the system is trustworthy at all.

Invisible system boundaries. The product appears to offload cancellation to Stripe’s customer portal without making that handoff explicit. If completing a core task requires leaving your application, that transition must be clearly marked and seamlessly executed. The small “link” label next to the payment method hints at this, but it’s buried in a section labeled “Payment Method”—not “Subscription Management.” Users looking to cancel aren’t scanning for payment links; they’re looking for cancellation controls.

No end-state preview at the decision point. When I clicked “Cancel plan,” I expected to see what would happen next: confirmation of my end date, clarification that I’d keep access until November 21st, and assurance I wouldn’t be charged again. Instead, I saw subscription details I’d already read and a button telling me not to do what I came to do. The absence of clear next-step information doesn’t just create uncertainty—it converts a straightforward business transaction into an anxiety-inducing maze.

Compounding frustration. Confusion at cancellation time doesn’t merely generate support tickets. It transforms satisfied users into vocal critics. I completed my video project successfully. I had no complaints about RunwayML’s core functionality. But this final interaction—this moment where the system actively worked against my clear intent—colors the entire relationship. What should have been a neutral exit became a story about deceptive UX I’m now sharing publicly.

Why This Matters Beyond UX

In my book Unfinished, I explore how established frameworks often persist long after they’ve stopped serving us. The cancellation UX pattern I encountered isn’t accidental—it’s the product of misaligned incentives, technical debt, and a fundamental misunderstanding of what retention actually means.

Making cancellation difficult isn’t a retention strategy. It’s borrowed time, paid back later in damaged reviews, eroded brand trust, and regulatory scrutiny. Real retention starts with an honest off-switch. When users trust that leaving is as simple as staying, they’re more likely to return when their needs change.

After decades of studying how people interact with digital systems, I’ve come to understand that friction at critical moments reveals organizational priorities. When a company makes it hard to leave, they’re saying: “We don’t trust our product to bring you back.” That admission echoes through every subsequent interaction—and in my case, through this very article.

The Off-Switch Test

I’ve developed a simple framework for evaluating cancellation UX—a lens any team can apply to their billing flow:

One screen, two clicks, three answers. Users should move from the initial cancellation action to confirmed cancellation in two clicks or fewer. At the decision point, show three critical pieces of information: effective end date, refund or renewal status, and how to reactivate if they change their mind.

RunwayML’s flow failed this test immediately. Clicking “Cancel plan” should have taken me to a confirmation screen, not a discouragement screen.

Label consistency matters. If the first button says “Cancel plan,” every subsequent step must use identical phrasing—not “Don’t cancel subscription.” Inconsistent labels suggest inconsistent outcomes. They erode trust word by word, click by click.

Make handoffs explicit. If cancellation lives in Stripe’s customer portal, the “Cancel plan” button should open that portal directly, deep-link to the subscription management page, and auto-focus the cancellation action. Add contextual copy: “Opening secure billing portal (powered by Stripe).” Don’t make users hunt through payment method sections for a vague “link” that might or might not lead to cancellation controls.

Write transparent copy. Microcopy should state the obvious: “You’ll keep access until November 21, 2025. You won’t be charged again.” Offer a “Remind me before renewal” option as a humane alternative to immediate cancellation. Sometimes users just need reassurance about timing, not a complete exit.

Enable undo and reactivation. A 7-day undo period and one-click reactivation reduce fear and support load. They signal confidence in your product’s value. Users who know they can easily return are less panicked about leaving—and more likely to actually come back.

Send receipts that reassure. Instant email confirmation with the end date and next steps converts anxiety into closure. It’s the digital equivalent of a handshake at the end of a transaction.

What RunwayML Should Show Instead

Imagine clicking “Cancel plan” and seeing this:

Modal header: Cancel your Standard subscription?

Body copy: Your plan will end on November 21, 2025. You’ll keep full access until then and won’t be charged again. You can reactivate anytime with one click.

Need more time to decide? We can remind you 3 days before your renewal.

Buttons:

• Confirm cancellation (primary)

• Remind me later (secondary)

• Keep my plan (tertiary link)

What RunwayML Should Show Instead

Imagine clicking “Cancel plan” and seeing this:

Modal header: Cancel your Standard subscription?

Body copy: Your plan will end on November 21, 2025. You’ll keep full access until then and won’t be charged again. You can reactivate anytime with one click.

Need more time to decide? We can remind you 3 days before your renewal.

Buttons:

• Confirm cancellation (primary)

• Remind me later (secondary)

• Keep my plan (tertiary link)

Success state after confirmation: All set. Your Standard plan ends November 21, 2025. We’ve sent confirmation to your email. Come back anytime—your workspace stays ready.

This isn’t complicated. It’s honest.

Business Value Beyond User Experience

Transparent cancellation flows deliver measurable business outcomes:

• Fewer support tickets and chargebacks (no more “how do I actually cancel” emails)

• Lower involuntary churn (clear dates reduce panicked cancellations)

• Higher reactivation rates (users leave with trust intact, not resentment)

• Improved NPS and CSAT at the most sensitive moment in the customer journey

These aren’t soft benefits. They’re the difference between users who quietly disappear—or worse, loudly complain—and users who return when their needs change.

The Real Test

If your team hasn’t run the Off-Switch Test recently, try canceling your own product as a new user would. Use an incognito window. Don’t lean on your internal knowledge of where things are hidden. Click only what’s labeled as cancellation. Follow only the visible paths.

The result isn’t just a UX audit. It’s a window into your actual retention strategy—the one you’re demonstrating through design decisions, not the one you’re describing in strategy documents.

Design is a promise. The cancellation flow is where you prove whether you meant it.

Quick UX Checklist for Product Teams

• Cancel” action completes in ≤2 clicks from billing

• Same action label across every step (no “Cancel plan” → “Don’t cancel subscription”)

• Deep-link to external portals if cancellation lives there

• Show end date, charge status, and access retention up front

• Offer undo period and one-click reactivation

• Send instant confirmation email with clear next steps

• Provide “remind me before renewal” as a middle option

About the Author

Haider Ali is a Lead Digital Experience Design Architect with over 22 years of experience shaping digital experiences across enterprise, government, and consumer products. Working at a leading innovation lab in the energy sector, he explores the intersection of design, technology, and human experience, building AI-powered systems that augment human capability rather than replace human judgment.

He is the author of Unfinished: Notes on Designing Experience in a World That Never Stops Changing, a book exploring how established frameworks persist long after they’ve stopped serving us—and how we can build better systems by questioning what we’ve inherited.

Through his publication User First Insight, Haider examines the intersection of design, technology, and trust, translating complex system failures into actionable frameworks that teams can use to build more honest, human-centered products.

Connect with Haider on LinkedIn, follow his writing on Medium, visit haiderali.co, or explore his book at stayunfinished.com.

The conversation about UX team structure often presents false choices: centralized versus decentralized, embedded versus independent, autonomy versus alignment. We debate reporting lines, draw org charts, and search for the perfect model as if structure alone determines success.

But here’s what five years of building UX practice at enterprise scale has taught me: the model matters far less than your willingness to keep evolving it.

I work in an organization serving over 200,000 users across a geographically distributed operation. We’re responsible for digital and AI initiatives that span the entire enterprise—from core products like our intranet to design systems, governance frameworks, and consulting on complex projects throughout the organization. The scale alone makes traditional centralized models impossible. Yet complete decentralization would sacrifice the consistency and cohesion that enterprise experiences demand.

So we’ve built something different. Not because we followed a framework, but because we kept asking: What actually works here? For these people? In this context?

The Reality of Distributed Governance

Our structure is fundamentally decentralized—UX professionals are distributed across departments and geographies, embedded in product teams throughout the organization. But we maintain a central governance function that does three critical things:

We train practitioners across departments to serve as UX representatives within their teams. These aren’t junior designers learning basics; they’re professionals we develop to maintain standards while adapting to local needs.

We consult on complexity. When product teams face genuinely difficult challenges—architectural decisions, cross-platform experiences, novel interaction patterns—we engage directly. Not as gatekeepers, but as specialized expertise activated when it matters most.

We own the connective tissue. The intranet experience. How services integrate into larger ecosystems. The design system that ensures consistency without constraining creativity. Governance documents that provide direction without dictation.

This isn’t a model I found in an article or extracted from a case study. It’s what emerged from constantly asking: Is this still working? Where are the gaps? What needs to change?

What Traditional Models Miss: The Gradient Between Structure and Practice

Most UX organizational frameworks assume a level of control that enterprises can’t actually exercise. They optimize for reporting clarity, role definition, and clean boundaries. But they miss something crucial: at scale, structure becomes less important than systems of influence.

Consider the typical managed-integration approach, where UX reports to UX leadership while embedding in product teams. It’s elegant in theory. In practice, with hundreds of practitioners across vast geographies supporting dozens of initiatives simultaneously? The reporting line becomes an abstraction. What matters is:

• Knowledge transfer systems that help distributed teams learn from each other

• Governance that guides without constraining local adaptation

• Core products that demonstrate standards rather than mandate them

• Strategic consulting capacity that can parachute in when needed

We don’t achieve consistency through command and control. We achieve it through deliberate cultivation of practice, supported by infrastructure that makes good work easier than inconsistent work.

The Tooling Philosophy: Permanence is Fiction

Here’s where enterprise UX often calculates incorrectly: we invest enormous energy selecting “the right tools,” as if choosing correctly means we’re done choosing.

Our approach is different. We treat tools as temporary scaffolding for current thinking.

Today—literally as of October 21, 2025—we’re heavily using Make for design workflows, experimenting with vibe coding through Windsurf, Cursor, and Lovable, and employing AI extensively for research synthesis, heuristic analysis, and workflow automation. Six months from now? The list will look different.

This isn’t chaos. It’s acknowledging that in my book Unfinished, I argue that design practice must embrace constant transformation rather than resist it. The same applies to how we structure teams and select tools. We’re not building permanent architecture; we’re building adaptive capacity.

The moment you stop experimenting—the moment you declare “this is how we do things here” and close the door on evolution—you’ve begun calcifying.

When Decentralization Creates New Problems

I won’t pretend our approach is perfect. Decentralization at this scale creates real challenges:

Visibility becomes difficult. With practitioners embedded across the organization, understanding what’s actually happening—where the struggles are, where the innovations emerge—requires deliberate systems, not just good intentions.

Quality becomes variable. When you can’t directly manage every practitioner’s daily work, you rely on training, governance, and culture to maintain standards. Sometimes that’s insufficient.

Resource allocation becomes complex. You can’t simply assign designers to teams when those designers already belong to other organizational units. Influence must substitute for authority.

Consistency requires active maintenance. Our design system isn’t just documentation—it’s an ongoing conversation about what consistency means when contexts vary wildly.

We address these through regular training cycles, active consulting engagement on complex projects, and owning the experiences that set organizational standards. But I’m not claiming we’ve “solved” decentralization. We’re constantly adjusting our approach as we discover what works and what doesn’t.

The Cultural Dimension: What Works Depends on Context

One aspect rarely discussed in Western UX literature: cultural context fundamentally shapes what “good” structure looks like.

Working in the Middle East, serving a workforce that spans cultures, languages, and contexts, we learned early that certain assumptions don’t travel. Visual language that seems universal isn’t. Interaction patterns that feel intuitive in one context confuse in another. Even the concept of “user-centered design” requires translation—not just linguistically, but philosophically.

Our governance approach reflects this. We maintain professional, clear visual language that works across contexts. We avoid assuming shared cultural references. We test across user segments that Western organizations might treat as homogeneous.

This isn’t about being “sensitive” or “inclusive” in some abstract sense. It’s recognition that enterprise UX at global scale demands humility about your assumptions. The best model is the one that works for your users, not the one that worked for someone else’s users.

The Framework We’re Building: Incremental Evolution Over Grand Redesign

Rather than perfect our current structure, we’re developing a framework for continuous structural evolution. The core principles:

1. Structure serves strategy, not vice versa. When organizational priorities shift, our structure should shift with them. Rigidity is failure.

2. Governance guides through demonstration, not mandate. Our core products show what good looks like. Our design system makes consistency easy. Our consulting engagements spread practices through doing, not documenting.

3. Decentralization requires strong connective tissue. You can’t just distribute practitioners and hope for coherence. You need shared systems, regular forums, and deliberate knowledge sharing.

4. Metrics matter, but not the ones you think. We track impact through adoption of standards, reduction in redundant work, and practitioner capability growth—not hours logged or tickets closed. (Note: specific numbers remain confidential, but the trajectory shows consistent improvement in all three areas.)

5. Tools are disposable; capability is permanent. Invest in people’s ability to learn and adapt, not in perfect tool selection.

This framework isn’t finished. It won’t ever be finished. That’s the point.

What Most UX Leaders Get Wrong

Here’s my contrarian take after years of building UX practice at enterprise scale: Most UX leaders optimize for the wrong thing.

They optimize for structural clarity when they should optimize for adaptive capacity. They perfect reporting lines when they should perfect learning systems. They seek the right model when they should build the capability to keep evolving models.

The conventional wisdom provides valuable direction—understanding centralized versus decentralized tradeoffs, recognizing the challenges of matrix reporting, appreciating the value of embedded practitioners. But direction isn’t dictation. The path you take depends entirely on your context, your culture, your constraints, and your users.

An organization of 500 needs different structure than one of 50,000. A product company has different needs than an enterprise with internal users. A culture that values hierarchy requires different approaches than one that prizes autonomy.

The wisdom isn’t in choosing the right model. The wisdom is in building your capacity to keep questioning whether your current model still serves its purpose.

The Invitation: Share What Actually Works

I’m developing this framework further because I suspect many large organizations face similar challenges—too big for pure centralization, too complex for simple decentralization, too dynamic for static structures.

I want to hear from practitioners and leaders building UX at scale:

• What’s your organizational size, and what structure have you adopted?

• Where does your model succeed? Where does it struggle?

• What have you learned that contradicts conventional wisdom?

• How do you maintain consistency while enabling local adaptation?

This isn’t academic curiosity. It’s recognition that the best insights emerge from collective intelligence, not individual genius. The frameworks that matter most are built collaboratively, tested in real contexts, and evolved through shared learning.

Closing: The Only Constant is Evolution

In Unfinished, I argue that design must embrace constant transformation—that the systems we build exist in a state of permanent becoming rather than completed being. Our UX practice reflects this philosophy directly. We don’t have a perfect structure. We have a structure that works reasonably well right now, with systems in place to help us recognize when it stops working and evolve accordingly.

This is the actual work of enterprise UX leadership: not finding the right answer, but building the organizational capacity to keep asking better questions. Not implementing the perfect model, but creating conditions where people can experiment, learn, fail safely, and improve continuously.

The tools will keep changing. The challenges will keep evolving. But the mission remains constant: creating experiences that genuinely serve human needs, at whatever scale necessary, through whatever structure makes that possible.

Technology should expand our creative horizons. Structure should enable rather than constrain. And wisdom comes not from following the book, but from understanding when to write your own.

Ready to explore further? Subscribe to User First Insight for perspectives on design, technology, and human experience in enterprise contexts. For broader explorations of how we navigate complex systems and societal challenges, follow Black & White Perspective. And if these ideas resonate, Unfinished: Notes on Designing Experience in a World That Never Stops Changing offers deeper exploration of designing for constant transformation.

Connect with me on LinkedIn for professional conversations, follow my writing on Medium for additional insights, and visit haiderali.co and stayunfinished.com to continue the conversation.

This isn’t just content—it’s an invitation to question, evolve, and reimagine what becomes possible when we approach both structure and practice with critical thinking and adaptive action.

There’s a moment in design when you stop analyzing and start feeling. Your hand finds the mouse. The cursor flows. You’re not consulting your style guide or remembering some accessibility guideline—you’re just making choices that feel right. That’s vibe design.

If you’ve never heard the term, you’ve definitely experienced it. It’s what happens when you’re deep in a project, the coffee is warm, the music is good, and somehow, without conscious effort, your design just works. Not just functionally—though it does that—but aesthetically, intuitively, harmoniously. Your product has a groove.

What Is Vibe Design?

Vibe design is the practice of making design decisions based on intuition, aesthetic judgment, and the feeling of rightness rather than strict adherence to external rules or frameworks. It’s the opposite of template-based design. It’s not about ignoring user research or accessibility standards; it’s about having internalized them so completely that they become instinct.

When you vibe design, you’re making micro-decisions constantly: Should this button be primary or secondary? Does this user flow need three steps or two? Should this color be pushed more toward blue or kept neutral? Should the spacing breathe here or feel compact? A good vibe designer doesn’t just follow their design system—they know when to bend it, when to extend it, when to trust their gut.

It’s the difference between following a design kit and designing.

The Vibe Design Manifesto

Clarity over cleverness. Your design should make users smile when they interact with it. Not because it’s surprising, but because it’s intuitive and feels effortless.

Consistency over rigidity. A design system that feels alive and coherent, even if it bends a few traditional rules, beats a system that’s technically perfect but lifeless and mechanical.

Elegance over exhaustiveness. A well-considered micro-interaction beats seventeen edge-case states no one will ever see. Know when not to design.

Trust over enforcement. Good vibes come from working with teams that trust your judgment. Design tokens are great, but they shouldn’t strangle your creative instincts.

The Skill Behind the Feeling

Here’s the thing about vibe design: it looks like magic until you realize it’s actually just experience crystallized into intuition.

A junior designer looking at a vibe designer’s work might think, “How did they know that button placement would work so well?” The answer is: they’ve made the wrong placement a hundred times. They’ve watched users get frustrated with cluttered interfaces. They’ve experienced the subtle friction of bad hierarchy and the quiet joy of a perfectly balanced layout. Their vibe isn’t random; it’s pattern recognition compressed into feel.

This is why vibe design can’t be learned from a course. You can’t memorize it. You have to live it. You have to design interfaces badly, recognize why they felt wrong, iterate toward better solutions, and let that experience settle into your aesthetic sensibility. You need to watch users interact with your designs. You need to feel their confusion. You need to feel their delight.

The Vibe Design System

Some of the most beautiful products I’ve encountered came from teams that shared a design vibe. They didn’t have 200-page brand guidelines. They had a Figma where someone might say, “I’m thinking we approach this pattern with more breathing room,” and everyone felt it. The product that emerged was unmistakable—you could interact with it and think, “Oh, that’s a [company name] product.” It had personality.

This is rare. Most design systems operate like bureaucracies where a design director descends from on high with The Rules. Everything is pixel-perfect and on-brand, but it feels corporate. It’s functional in the way a hotel is functional—everything matches, but nobody gets excited about it.

Vibe design teams are different. There’s a rhythm. A shared understanding of what feels right. When you review a design, you can tell immediately if it’s been made by someone in the vibe or someone just checking boxes against a design system.

The Dangers of Bad Vibes

Of course, vibe design can go wrong. Bad vibes are a real thing.

A product designed by someone who hasn’t earned their vibe yet is chaos. It’s trendy for the sake of trendiness. It’s inconsistency masquerading as personality. It’s a designer who thinks the accessibility guidelines don’t apply to them because they haven’t internalized the principles yet—just the aesthetics.

The difference between good vibes and bad vibes is usually: Can I hand this off to another designer and have them maintain it? Can users from different backgrounds navigate it effortlessly? If the answer is no—if the design is so idiosyncratic that only the original designer understands it, or if it’s beautiful but inaccessible—then it’s not vibe design. It’s just ego.

Good vibes are generous. They make the experience easier for everyone, not harder. They reduce confusion and cognitive load.

Living in the Vibe

The best part about vibe design is the state of flow it creates. When you’re doing it well, you’re not fighting your design system or your tools or your user research. You’re working with them. You’re in the groove. You’re making hundreds of small decisions that, collectively, feel inevitable.

That feeling—that’s worth protecting. It’s worth pursuing. It’s worth building teams and selecting tools and organizing projects around.

In a world where design is increasingly templated and industrialized (design systems, AI-generated interfaces, component libraries), there’s something deeply human about vibe design. It’s the space where research becomes intuition, where guidelines become voice, where design stops being decoration and becomes craft.

The best vibe designers are the ones who’ve studied their users obsessively, internalized their needs, understood their contexts—and then let all of that fade into the background so they can make decisions from a place of deep knowing rather than conscious thinking.

The Vibe Approach in Action

This philosophy extends beyond individual products and into entire ways of working. When launching my book Unfinished: Notes on Designing Experience in a World That Never Stops Changing, I faced a choice: follow the traditional publishing playbook or trust my vibe about what felt right.

Rather than spending months building a website through conventional agencies or bloated platforms, I leaned into the vibe approach. I worked with tools and partners that felt right—moving quickly, iterating based on feel, making hundreds of small decisions guided by intuition rather than committee. What would typically take months happened in weeks. The website at stayunfinished.com emerged not from a rigid plan, but from a shared vibe with collaborators who understood the vision and trusted the process. The result? A digital home for the book that actually reflects its core message about designing in a world of constant change—nimble, intentional, alive.

That experience crystallized something I’d always believed: vibe design isn’t just about interfaces and interactions. It’s about entire systems, entire processes, entire ways of working.

The Convergence

Interestingly, the best products are made by teams where developers and designers share a vibe. The code feels like the design. The design respects the code. They’re not fighting each other or the user. They’re orchestrated. When you use a product like that, you don’t think about the technology or the visual design separately—you just think about how right it all feels.

That’s the goal. That’s the vibe.

So the next time you find yourself making a design decision that just feels right, don’t second-guess it. Trust it. Feel that vibe. Protect it. Because that’s where the best design lives.

What’s your vibe? Does your team have a shared design rhythm, or is everyone doing their own thing? I’m curious how you think about design beyond the pixel-pushing and the design system rules.

About the Author

I work at the intersection of design, technology, and human experience—crafting intelligent systems that amplify human capability rather than replace it. As a Digital Experience Design Architect, my practice is grounded in a belief that the most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and create.

My approach combines rigorous methodology with creative vision. I question assumptions, challenge conventional wisdom, and seek patterns that others might miss. Whether exploring user research methodologies, designing enterprise systems, architecting digital experiences, or examining broader societal challenges, I maintain a critical lens that asks not just “what works” but “why it works” and “for whom does it work best.”

Each article I write reflects this philosophy: technology should expand our creative horizons, design should serve genuine human needs, and innovation should be tempered with wisdom about its implications.

Connect & Explore Further

Subscribe to User First Insight for perspectives on design, technology, and human experience in enterprise contexts.

For broader explorations of sustainability, global politics, and societal challenges, follow Black & White Perspective where I examine clear perspectives on the issues that matter and practical ways to solve them.

Read the book: Unfinished: Notes on Designing Experience in a World That Never Stops Changes offers deeper exploration of design philosophy in an age of constant transformation.

Connect with me: Find professional conversations on LinkedIn, additional insights and case studies on Medium, and see these ideas in practice at haiderali.co and stayunfinished.com.

This is more than content—it’s an invitation to question, to evolve, and to reimagine what becomes possible when we approach both technology and society with critical thinking and thoughtful action. The tools keep changing, the challenges keep evolving, but the mission remains constant: creating experiences and solutions that genuinely improve how people live, work, and create.

The integration of Artificial Intelligence (AI) into the user experience (UX) research workflow represents one of the most significant paradigm shifts since the establishment of HCI principles decades ago. This transformation promises unprecedented speed and scale in data analysis, but it simultaneously introduces complex, long-term ethical and legal ramifications that demand immediate strategic attention. For senior product leaders and researchers, the challenge is no longer merely understanding the rights and well-being of participants, but ensuring that the advanced computational systems used also respect the dignity of the people involved in research activities.

The foundation of modern ethical research must be built upon internationally recognized regulatory consensus. The UNESCO Recommendation on the Ethics of Artificial Intelligence provides a comprehensive, human-rights centered approach that spans the entire AI lifecycle—from initial research, design, and development through to eventual deployment, evaluation, and end-of-life termination. This broad mandate compels organizations to view AI not as a feature set, but as a structural component of society requiring explicit governance.

Fundamentally, the strategic value of AI in user research lies in its capacity for augmentation, not replacement. AI’s inherent speed and analytical scale are capable of leveling the playing field for bringing new products to market, enhancing the need for human intelligence by streamlining repetitive workflows. This augmentation frees researchers to focus on strategic interpretation and the “human elements” inherent to design and product management. For AI to be leveraged successfully for profound, actionable insights, it must always be paired with continuous, conscious human oversight.

Aligning Macro Ethics with Micro Interactions

A critical challenge for implementing ethical AI is bridging the gap between high-level philosophical principles and daily operational mechanics. Regulatory bodies provide the macro ethical values—such as Fairness, Accountability, and Transparency (FAT)—that must guide enterprise policy. However, the immediate impact on a user or researcher interacting with an AI tool is defined by micro design principles.

The Nielsen Norman Group (NN/g), a leading authority on Human-Computer Interaction (HCI), has identified specific interaction ethics that must be embedded directly into the AI tool stack to ensure a safe and respectful experience for research participants. These principles—User Control, Error Recovery, and Feedback Loops—dictate how the ethical mandate translates into interface design. Allowing a user to override an AI decision (User Control), for example, is the tactical manifestation of their macro right to human determination and autonomy. If AI makes a mistake, the interface must provide clear paths to correct those errors (Error Recovery), reinforcing trust and accountability. Finally, incorporating continuous user feedback (Feedback Loops) ensures the ongoing improvement and ethical realignment of the underlying models.

Effective ethical practice, therefore, requires designing interfaces and workflows that allow users and researchers to actively exercise their macro rights through these specific, measurable micro controls.

Foundational Ethical Frameworks: Design Principles and Human Oversight

Ethical implementation of AI requires adherence to foundational principles that prioritize human well-being over algorithmic efficiency. These frameworks provide the operational blueprint for responsible AI development and procurement, addressing issues that arise both at the user interface level and the organizational policy level.

Human Oversight and Determination

A core principle established by international bodies is the absolute necessity of Human Oversight and Determination. Member States must ensure that AI systems do not displace ultimate human responsibility and accountability. Regardless of how sophisticated an algorithm becomes, the ultimate authority and legal accountability must remain placed upon natural or legal persons—the developers, the researchers, or the organization—and not the AI system itself.

This mandate translates into the “Human-in-the-Loop” requirement for UX research. AI must be viewed fundamentally as a decision-support system, designed to augment human judgment, not replace it. This is especially vital in sensitive domains, such as healthcare, where algorithmic errors or ‘hallucinations’ can directly impact patient safety and quality of care. By mandating human oversight and continuous auditing, organizations proactively mitigate the risk of unintended harm caused by autonomous AI decisions.

Accountability and Proactive Governance

Responsibility and accountability must be operationalized through formalized internal processes. Organizations have an ethical obligation to ensure AI is used responsibly, which requires mitigating the risks of bias, discrimination, and privacy violations. To promote accountability, organizations must develop specific ethical guidelines for AI use and establish formal oversight committees tasked with monitoring compliance.

Proactive due diligence is mandatory. Before any AI system is deployed in a research context, the organization must conduct a thorough ethical assessment to identify and mitigate potential risks. Corporations, such as Microsoft, have formalized this commitment through comprehensive frameworks like the Responsible AI Standard, which is built upon six key principles: fairness, reliability, inclusiveness, privacy, transparency, and accountability. Applying these principles consistently requires internal ethics reviews for all AI projects and the dedicated training of employees in ethical AI development.

The organizational structure supporting ethical AI must be collaborative. The development and deployment of AI systems require the involvement of diverse stakeholders, including users, researchers, and ethicists. This involvement helps ensure that the AI is designed and used ethically, reflecting a broad range of perspectives and minimizing blind spots related to social and cultural context.

Integrated Ethical Principles for AI in UX Research

Synthesizing principles from governance and interaction design yields a clear set of requirements for ethical practice:

Integrated Ethical Principles for AI in UX Research

PrincipleDefinition/ChallengeMitigation StrategyTransparency & Explainability (T&E)Algorithmic opacity (trade secrets, complexity) hinders trust and oversight. Appropriate level of T&E needed, balancing privacy.Label AI-derived vs. human insights; maintain clear documentation of AI’s role and data usage. Ensure awareness when a decision is informed by AI.Fairness & Non-DiscriminationAI models can perpetuate and amplify biases from historical data, leading to skewed outcomes.Conduct thorough ethical assessments; regularly audit AI models for bias; involve diverse stakeholders in system design.Privacy & Purpose LimitationRisk of hidden scope creep/data reuse for model training without consent. Need to protect highly sensitive data.Obtain tiered consent explicitly covering AI use/training; ensure strong data security and minimize collected PII. Ensure user data is explicitly not used for LLM training unless consented.Human Oversight & DeterminationAI system autonomy must not displace ultimate human responsibility.Implement human-in-the-loop mechanisms for AI decision validation ; establish internal accountability protocols and oversight committees. Researchers must retain the ultimate capacity to override AI decisions.

Mitigating Core Ethical Risks: Privacy, Bias, and Opacity

The convergence of AI with user research data significantly amplifies traditional ethical challenges, primarily centered around privacy breaches, algorithmic bias, and decision-making opacity. Managing these risks requires concrete, operational solutions integrated into the research workflow.

A. Data Privacy and the Peril of Purpose Creep

AI fundamentally complicates data privacy due to its capability for massive, pervasive data processing, the potential for using personal information for secondary purposes (purpose creep), and the technical difficulty in ensuring comprehensive data deletion. This is acutely true in sectors handling highly sensitive information; for example, healthcare data mining carries a high risk of exposing sensitive genetic or medical information without sufficient patient knowledge or explicit consent.

A crucial tension exists between legal mandate and technical necessity. The General Data Protection Regulation (GDPR) mandates purpose limitation, prohibiting the reuse of data gathered for specific purposes. Conversely, training robust deep learning models often requires vast amounts of data, a process frequently strengthened by reusing data collected for other, often unrelated, purposes. This fundamental conflict necessitates researchers be highly cautious and transparent regarding data provenance and usage.

Mandating Explicit Consent and Transparency

To protect research participants, researchers must move beyond standard legal disclaimers toward obtaining tiered, explicit consent that clearly outlines all AI involvement. This consent must specifically address AI analysis, transcription, data processing, and any use of third-party tools. Furthermore, given the significant risk of unconsented data reuse, researchers must explicitly guarantee that user data is not being used to train Large Language Models (LLMs) or internal proprietary AI tools without the participant’s express, separate permission. Researchers should treat vague or blanket consent statements as high-risk indicators, as they often miss potential downstream uses or data repurposing.

Vetting commercial AI tools is a non-negotiable step. Researchers must prioritize vendors that are transparent with their data collection practices and formally compliant with major regulatory laws like GDPR and the California Consumer Privacy Act (CCPA). Required operational actions include adopting data minimization practices (only collecting the necessary Personally Identifiable Information, PII), establishing clear data retention policies, and making opt-out or limitation of sharing easily available to participants. Leading commercial tools recognize this necessity and actively market their adherence to regulations like GDPR, CCPA, and standards such as SOC 2 Type II audits, which verify security and privacy controls.

B. Algorithmic Bias and Reinforcement

The risk of algorithmic bias arises when AI models are trained on historical data sets that reflect and thereby entrench societal prejudices based on sensitive attributes like race or gender. If unaddressed, this bias amplification leads to unfair or discriminatory outcomes in critical, real-world decisions, such as loan applications, hiring screenings, or medical diagnostics.

Mitigation Strategy: Auditing, Assessment, and Inclusion

Counteracting inherent model bias requires a continuous commitment to auditing and assessment. Organizations must conduct thorough ethical assessments prior to the deployment of any AI system. This is not a one-time process; the ethical implications of AI use must be regularly monitored and evaluated to identify and address any emerging issues.

Crucially, the design process itself must be inclusive. Developing AI systems requires involving diverse stakeholders—including users, ethicists, and researchers—to ensure the final design is both ethical and inclusive, reflecting a range of human experiences rather than the homogenous perspective of a narrow development team.

C. Algorithmic Opacity (The Black Box Problem)

Algorithmic opacity describes the difficulty stakeholders have in understanding how an AI system arrives at a decision. Proprietary algorithmic systems are often technically complex, protected as trade secrets, and managerially invisible to external oversight. This inscrutability—which can sometimes be intentional or simply a function of the deep complexity and “high dimensionality” of deep learning models—erodes consumer trust. Studies show that approximately 78% of consumers actively prefer companies that practice transparency in their AI systems.

The Need for Transparency and Explainability (T&E)

The ethical deployment of AI is contingent upon its Transparency and Explainability (T&E). Stakeholders must be made aware when a decision or insight is generated or influenced by AI. It is understood that achieving T&E involves a careful balancing act, as the level of disclosure must be appropriate to the context and may sometimes conflict with other principles such as privacy and security.

To operationalize T&E, UX teams must mandate that AI research tools provide trackable insights and maintain clear, verifiable documentation of the AI’s role at every stage of the user testing process. In published research, attribution guidelines must be established, clearly labeling AI-derived insights versus those derived solely from human analysis. Furthermore, relying solely on autonomous black boxes is unacceptable; internal controls must implement human-in-the-loop mechanisms to validate AI decisions, ensuring that the ultimate human judgment remains the final arbiter.

Regulatory Compliance Checklist for AI User Research Tools

To provide actionable steps for vetting tools and designing research protocols, the following regulatory compliance checklist integrates global requirements with practical research duties:

Regulatory Compliance Checklist for AI User Research Tools

Regulation/StandardCore RequirementActionable Research StepGDPR (EU)Strict data management, purpose limitation, right to erasure/deletion.Update consent forms to explicitly mention AI usage and retention policies. Ensure data isn’t used for AI training without permission.CCPA/CPRA (California)Right to opt-out/limit data sharing, data minimization, transparency in privacy notices.Process the minimal necessary PII; make opt-out or limitation of sharing easily accessible. Ensure strong data security measures are in place.Auditing Frameworks (e.g., SOC 2 Type II)Regular, independent audit of security, availability, and privacy controls.Prioritize vetted commercial tools that maintain and publish regular compliance audits (e.g., Lookback confirms SOC 2 Type II adherence).Data Reuse EthicsProhibits reusing data for secondary purposes (like model training) without proper, explicit consent.Avoid blanket consent statements; secure new, separate consent if training is required, or explicitly guarantee data will not be used for model training.

Technical Defenses: Quantifying Risk and Preserving Data Utility

When research necessitates sharing or analyzing data sets containing quasi-identifiers, ethical practice requires employing advanced anonymization techniques. This technical diligence addresses the crucial trade-off between maximizing individual privacy (reducing re-identification risk) and preserving the data’s utility (maintaining its usefulness for analysis).

The Limitations of Simple Anonymization

The most common initial step—the simple removal of direct patient identifiers—is almost universally insufficient to protect individual privacy. Researchers have demonstrated that, using external data sources, individuals can often be re-identified based on unique combinations of seemingly innocuous data elements (quasi-identifiers, such as age, gender, and general location). Therefore, anonymization techniques must be paired with controlled data access and rigorous technical evaluation before datasets are released or used for secondary analysis. The overall objective of advanced data anonymization is to enable analysis and publishing while guaranteeing that individual privacy is not compromised.

Syntactic Privacy Models and Their Evolution

Early technical models focused on defining privacy based on the structure (syntax) of the data:

K-Anonymity

This is the foundational syntactic privacy model. A dataset satisfies k-anonymity if, for the set of quasi-identifiers chosen by the researcher, the information released for each user is indistinguishable from at least $k-1$ other users who also appear in the release. This technique primarily limits the re-identification risk to a probability of $1/k$.

However, k-anonymity has limitations. It is vulnerable to homogeneity attacks (where all sensitive attributes within an indistinguishable group are too similar) and suffers from compositionality issues (combining two k-anonymous datasets does not guarantee the combined data remains k-anonymous).

L-Diversity and T-Closeness

To address k-anonymity’s shortcomings, specifically its failure to protect against attribute disclosure when sensitive values are homogenous, refinements were developed.

1. L-Diversity: This extension requires that within every equivalence class (the group of $k$ records), there must be at least $l$ unique values for each sensitive attribute. This measures the diversity of sensitive values within the group.

2. T-Closeness: This refinement goes further, ensuring that the distribution of sensitive attributes within an equivalence class is closely aligned with the overall distribution of the entire dataset. This prevents sophisticated inference attacks that exploit subtle distributional differences.

Tools like ARX are available to evaluate various combinations of these techniques, recommending optimal generalization and micro-aggregation levels to minimize re-identification risk while preserving data utility.

The Mathematical Guarantee: Differential Privacy (DP)

Differential Privacy (DP) represents the modern gold standard in data protection, offering a rigorous mathematical framework that provides provable privacy guarantees. DP works by introducing carefully calibrated statistical noise into query results or the dataset itself, ensuring that the inclusion or exclusion of any single individual’s data record does not substantially change the output.

Experimental evidence suggests that DP often outperforms earlier syntactic models like k-anonymity in achieving a more favorable balance between data utility and disclosure risk. However, DP is inherently complex to implement correctly and requires a deep understanding of statistical modeling.

Quantifying the Privacy-Utility Trade-Off

The decision of which privacy model to use must be data-driven, relying on metrics that quantify the risk of re-identification. This ability to place a quantifiable metric on ethical data handling elevates the process from soft compliance to a quantifiable engineering task. UX research teams handling sensitive quantitative data must secure access to data scientists capable of performing these rigorous risk assessments.

Risk Quantification Metrics

Specialized metrics must be used to assess the risk inherent in de-identified data before it is released for analysis:

• K-map: This metric assesses re-identifiability risk by computing the overlap between a given de-identified dataset of subjects and a larger re-identification—or “attack”—dataset.

• Delta-presence ($\delta$-presence): This metric estimates the probability that a specific individual from a larger population is present in the released dataset, helping evaluate population-level risk.

• ITPR (Information Theoretic-based Privacy Metric): A proposed metric designed to effectively quantify both the re-identification risk and the sensitive information inference risk associated with a dataset.

The Role of Synthetic Data

Synthetic Data Generation—the process of creating entirely artificial datasets that mimic the statistical properties and structure of the original sensitive data, but contain no real individual records—is an emerging technical bypass to the inherent privacy/utility trade-off. Synthetic data attempts to preserve data utility while eliminating the direct link to real individuals.

However, the creation and use of synthetic data introduce distinct ethical issues that require careful consideration. Researchers must evaluate the fidelity of the synthetic data to the original population and recognize the potential for synthetic data to inadvertently reflect or even amplify biases that were present in the source training data.

Technical Privacy Models: Utility vs. Risk Trade-Off

ModelPrimary Protection GoalMechanismKey Limitation/Trade-OffK-AnonymityRe-identification riskEnsures each record is indistinguishable from at least $k-1$ others using generalization.Vulnerable to homogeneity attacks (if sensitive attributes are uniform); compositionality issues.L-Diversity / T-ClosenessAttribute disclosure riskEnsures sensitive attributes are sufficiently diverse (l-diversity) or match population distribution (t-closeness) within groups.Necessary extensions to fix k-anonymity’s weaknesses.Differential Privacy (DP)Provable privacy guaranteeIntroduces carefully calibrated noise to query results or the dataset itself.Mathematically rigorous; can potentially outperform syntactic models in data utility/risk trade-off.Data SynthesisEliminates real recordsCreates entirely artificial data mimicking the statistical properties of the original.Utility can be highly variable; raises novel ethical issues regarding the fidelity and provenance of the synthetic data.

Strategic Governance: Building Accountability Frameworks

Technical measures are insufficient without robust, top-down enterprise governance. Ethical AI requires formalized policies, training, and audit mechanisms designed to mitigate systemic risks across the organization.

A. The Mandate for Enterprise AI Governance and Risk Assurance

Robust governance frameworks are essential for ensuring fair, equitable, and effective AI innovation while managing potential adverse incidents. Companies that rush to implement AI solutions without clear governance risk legal and ethical minefields, which can lead to significant reputational damage and real-world harm.

Successful AI governance must be integrated into existing business processes and policies to be effective, avoiding unnecessary duplication of effort. Governance is fundamentally a process of change management, requiring continuous education and communication to empower employees to continuously reflect on the ethical implications of their actions.

A more robust approach for large, multinational organizations involves shifting governance towards Risk Assurance. Risk assurance requires a harmonized internal audit process that asks open-ended questions about how different business units actively identify, manage, and mitigate AI-related risks. This audit model is adaptable locally, allowing business areas to reflect their unique regional risks, yet still subjecting all parts of the organization to a consistent standard of inquiry. By harmonizing risk audits globally, an organization prevents managers from simply outsourcing ethically complex or risky projects to jurisdictions with weaker standards, thereby ensuring consistent quality management and upholding enterprise reputation across all regions.

To manage resources efficiently, organizations should adopt a risk-based approach to defining the scope of governance. This involves classifying AI systems as low-, medium-, or high-risk and attaching proportionate governance requirements to each level. By using the familiar organizational concept of “risk assessment,” the governance requirements can be smoothly integrated into existing quality management processes.

B. The IIA AI Auditing Framework: Formalizing Oversight

The Institute of Internal Auditors (IIA) has developed an AI Auditing Framework that provides a rigorous structure for ensuring accountability and control across the organization. This framework is not just for auditors; it serves as a critical checklist for product leaders and senior researchers who are procuring or developing AI tools.

The IIA framework is built upon three overarching components: AI Strategy, Governance, and the Human Factor. Internal audit objectives must provide assurance over seven core elements:

1. Ethics: Ensuring consistency with the organization’s stated values, ethical responsibilities, and legal mandates.

2. Data Quality: Assessing the reliability, provenance, and integrity of the training and operational data used by the models.

3. The Black Box (Transparency): Reviewing policies and procedures to ensure the underlying algorithms, internal functions, and mechanisms that enable the AI are identified, understood, and documented.

4. Measuring Performance: Providing assurance on how performance metrics are established, monitored, and what level of performance deviation (model drift) is considered acceptable after deployment.

5. Cyber Resilience, AI Competencies, and Data Architecture & Infrastructure.

The requirements of this framework place an auditing imperative on UX research teams. Since internal audit is tasked with assessing the Black Box and Data Quality , research teams must proactively demand that commercial AI vendors provide the necessary technical documentation (model versions, training data sources, and performance benchmarks) to satisfy these internal audit requirements. Effectively, the IIA framework becomes an indispensable vendor vetting checklist.

From an organizational standpoint, formal governance requires a documented process that users must follow when requesting the use of AI, supplementing the core policy. This formal approval process helps the organization maintain a critical inventory of all AI users and departments, formalizing expectations for development, deployment, and monitoring. Finally, organizations must develop AI-specific incident response plans to address and mitigate potential compliance breaches related to AI systems.

The Future Role of the Ethical Researcher: Augmentation and Literacy

The current trajectory of Generative AI adoption confirms its inevitable integration into enterprise operations. High exposure rates across industries (79% of all respondents report some exposure) and widespread use in tasks like analyzing market data (74% of sales professionals) and generating basic content underscore that GenAI will continue to automate and accelerate analysis in user research. The challenge for the research community is not resisting this change, but governing it ethically.

The Paradox of Speed and Ethical Debt

The core utility of GenAI in user research is speed—the ability to accelerate insights and enhance customer interactions. However, this accelerated pace, often driven by the competitive imperative to maximize profits, carries an inherent risk of incurring ethical debt. When organizations prioritize rapid deployment, they may be tempted to outsource or ignore risky projects, undermining the governance structures needed for safe implementation.

Therefore, the primary strategic challenge for UX leaders over the next three to five years is to impose necessary friction on AI adoption. This friction takes the form of mandatory ethical assessments, rigorous human validation steps, and formalized audit pathways (like the IIA framework), explicitly designed to prevent speed from overwhelming safety and ethical consideration.

The Human-Centric Mandate

As AI systems accelerate data processing, the human researcher’s role becomes indispensable as the ultimate ethical gatekeeper and interpreter. The focus shifts from executing low-level tasks to strategic interpretation and ensuring that products are ultimately “built for humans, not ‘synthetic’ users”. The risk of “losing the human touch,” reducing interpersonal relationships, and removing the human factor in critical interactions (such as customer service or healthcare) remains a key public concern that researchers must actively counteract through careful design and deployment.

Ethical governance acknowledges the inherent difficulty in quantifying ethics directly, leading to a strategic pivot toward process-based KPIs. Instead of attempting to measure abstract concepts like “fairness” in a vacuum, successful governance measures compliance with the necessary ethical processes (e.g., “Was a formal ethical assessment conducted?”, “Was the training data audited for bias?”). This focuses the effort on proactive risk identification and management rather than subjective, box-ticking exercises.

The Necessity of Ethical Literacy

The complexity of foundational AI models and the nuanced debate surrounding technical privacy compliance are beyond the scope of traditional UX training. The senior researcher can no longer outsource the responsibility of vetting AI “add-ons”.

To ensure responsible governance, ethical literacy must become a core competency of the senior researcher. This allows them to accurately evaluate the trustworthiness of commercial tools, scrutinize the vendor’s internal data usage policies, and understand the crucial technical trade-offs inherent in privacy defenses such as $k$-anonymity versus Differential Privacy. Promoting public and professional understanding of AI and data through open education, digital skills training, and AI ethics curricula is critical to fostering a culture of continuous reflection and ethical action.

About the Author

I work at the intersection of design, technology, and human experience—crafting intelligent systems that amplify human capability rather than replace it. As a Digital Experience Design Architect, my practice is grounded in a belief that the most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and create.

My approach combines rigorous methodology with creative vision. I question assumptions, challenge conventional wisdom, and seek patterns that others might miss. Whether exploring user research methodologies, designing enterprise systems, architecting digital experiences, or examining broader societal challenges, I maintain a critical lens that asks not just “what works” but “why it works” and “for whom does it work best.”

Each article I write reflects this philosophy: technology should expand our creative horizons, design should serve genuine human needs, and innovation should be tempered with wisdom about its implications. I write to share insights, provoke thought, and invite others into conversations about how we can build a future where human creativity and technological capability work in genuine partnership.

For those eager to explore further:

Subscribe to User First Insight for perspectives on design, technology, and human experience in enterprise contexts. For broader explorations of sustainability, global politics, and societal challenges, follow Black & White Perspective where I examine clear perspectives on the issues that matter and practical ways to solve them. My book Unfinished: Notes on Designing Experience in a World That Never Stops Changes offers deeper exploration of design philosophy in an age of constant transformation. Connect with me on LinkedIn for professional conversations, follow my writing on Medium for additional insights and case studies, and visit haiderali.co and stayunfinished.com to see how these ideas manifest in practice.

This is more than content—it’s an invitation to question, to evolve, and to reimagine what becomes possible when we approach both technology and society with critical thinking and thoughtful action. The tools keep changing, the challenges keep evolving, but the mission remains constant: creating experiences and solutions that genuinely improve how people live, work, and create.

References

1. https://www.uxdesigninstitute.com/blog/what-are-user-research-ethics/

2. https://www.userinterviews.com/blog/ai-panel-recap

3. https://www.lookback.com/

4. https://www.unesco.org/en/articles/recommendation-ethics-artificial-intelligence

5. https://medium.com/commencis/the-future-of-ai-in-ux-design-a-comprehensive-guide-to-integrating-artificial-intelligence-into-2339f168dd56

6. https://www.frontiersin.org/journals/computer-science/articles/10.3389/fcomp.2022.1068361/full

7. https://tsaaro.com/blogs/ai-ethics-on-the-global-stage-a-deep-dive-into-unescos-recommendation-on-the-ethics-of-artificial-intelligence/

8. https://trustarc.com/resource/ai-applications-used-in-privacy-compliance/

9. https://pmc.ncbi.nlm.nih.gov/articles/PMC11457458/

10. https://adamfard.com/blog/ethical-ai-in-ux-research

11. https://onlinedegrees.sandiego.edu/ethics-in-ai/

12. https://maze.co/collections/ai/ethics-user-research/

13. https://www.unesco.org/en/artificial-intelligence/recommendation-ethics

14. https://pmc.ncbi.nlm.nih.gov/articles/PMC12255135/

15. https://jhmhp.amegroups.com/article/view/6448/html

16. https://www.mdpi.com/2075-4698/14/12/271

17. https://www.userinterviews.com/blog/the-user-researchers-guide-to-gdpr

18. https://academic.oup.com/jrsssa/article/185/4/1644/7069388

19. https://www.theiia.org/globalassets/site/content/research/foundation/2024/harnessinggenerativeai_1.pdf

20. https://www.ulapx.ai/resources/why-algorithmic-transparency-is-critical-for-ai-trust-and-compliance-part-2

21. https://www.theiia.org/globalassets/site/content/tools/professional/aiframework-sept-2024-update.pdf

22. https://www.tandfonline.com/doi/full/10.1080/2573234X.2025.2461507?src=exp-la

23. https://pmc.ncbi.nlm.nih.gov/articles/PMC6450246/

24. https://www.oaepublish.com/articles/jsss.2020.20

25. https://arxiv.org/pdf/2304.07210

26. https://cloud.google.com/sensitive-data-protection/docs/concepts-risk-analysis

27. https://arxiv.org/pdf/2510.11299

28. https://fpf.org/blog/the-curse-of-dimensionality-de-identification-challenges-in-the-sharing-of-highly-dimensional-datasets/

29. https://maze.co/collections/ai/ethics-user-research/

30. https://pmc.ncbi.nlm.nih.gov/articles/PMC11094102/

31. https://www.nist.gov/document/comments-draft-nistir-8312-ashley-p-moore-usagm-office-chief-information-officer-cio-1

32. https://www.intelligence.gov/ai/ai-ethics-framework

33. https://www.theiia.org/globalassets/site/content/tools/professional/aiframework-sept-2024-update.pdf

34. https://www.mdpi.com/2227-9709/12/2/36

35. https://research.aimultiple.com/generative-ai-ethics/

36. https://www.nu.edu/blog/ai-statistics-trends/

37. https://www.salesforce.com/news/stories/generative-ai-statistics/

Survey-based research faces a fundamental credibility crisis in enterprise UX. Studies show self-reported behavior overestimates actual behavior by 2x, while the correlation between what users say they prefer and how they actually perform is just r=0.44—meaning stated preferences predict only 25% of actual usability. For large organizations like Aramco and similar global energy companies conducting intranet redesigns, this gap between self-report and reality demands a critical reassessment of traditional survey methodologies and a strategic pivot toward AI-powered behavioral analytics and self-learning systems.

This research examines five critical dimensions: the fundamental limitations of survey-based research for understanding workflows, current AI-powered tools that can augment or replace surveys, the emerging landscape of self-learning systems that improve without explicit feedback, a critical analysis of when surveys fail versus succeed, and practical frameworks for reflecting on methodology in ongoing projects. The evidence reveals that while surveys excel at measuring attitudes and satisfaction at scale, they systematically fail to capture actual user behaviors—the very insights essential for effective enterprise intranet redesign.

Survey-based research cannot capture actual workflows

Traditional surveys suffer from an insurmountable limitation: they collect attitudinal data about what users think and feel, not behavioral data about what users actually do. Nielsen Norman Group explicitly warns that surveys are “no substitute for observational methods” and that “self-reported data is not enough for a good redesign, and can be misleading.” For large multinational organizations like Aramco, Shell, ExxonMobil, and similar global enterprises with 50,000+ employees across diverse roles from field workers to engineers to executives, surveys systematically miss the contextual factors that determine intranet success or failure.

The evidence is stark. Research by Brenner and DeLamater found that self-reported rates of behaviors like exercise and attendance were double the actual frequency when compared to administrative records—a 100% overreporting rate even in anonymous surveys. Nielsen and Levy’s foundational research established only a 0.44 correlation between users’ measured performance and stated design preference, meaning you can predict just 25% of how well a design works from knowing how much users like it. In 298 Nielsen Norman Group studies measuring both objective and subjective metrics, 30% showed paradoxes where users performed worse than average but liked designs more than average.

Surveys fundamentally cannot capture what happens in real workflows. They miss physical workspace adaptations like sticky notes with passwords on desks or paper “cheat sheets” employees create to overcome system limitations. They cannot observe real interruptions, multitasking, or the environmental factors affecting task completion. Crucially, surveys cannot reveal workarounds employees develop, the cross-tool workflows where the intranet fits into a larger ecosystem, or the emotional responses of frustration and confusion as they occur in the moment. As one researcher noted, when users were observed using banking apps, they said security requirements were too rigorous—but when surveyed separately, those same users wanted more security. The observation revealed actual behavior; the survey captured aspirational identity.

The response rate crisis compounds these problems. Organizations report that 60% of employees don’t view internal content, while survey response rates typically fall between 10-40%. This creates non-response bias that can overestimate employee satisfaction by 10-15 percentage points when disengaged employees opt out. The “squeaky wheels” over-respond while the satisfied silent majority remains unrepresented, leading to design changes that fix problems for vocal minorities rather than actual user needs.

What surveys provide versus what they miss

Surveys excel at three specific functions: measuring satisfaction levels and perceived ease of use at scale, establishing quantitative benchmarks for tracking metrics over time, and reaching diverse populations cost-effectively. For an organization like Aramco, a well-designed survey can gather attitudinal data from 500+ employees representing all departments, locations, and job levels—achieving statistical validity with a 95% confidence level and 5% margin of error. Standardized instruments like the System Usability Scale (SUS) provide reliable comparative data across iterations.

But surveys systematically fail to answer the critical questions for intranet redesign. They cannot reveal why behind behaviors—a user abandoning a form might indicate poor UX, technical issues, or simply a phone interruption, and the survey cannot distinguish. They miss workflow and task completion details essential for identifying friction points. They fail to evaluate information architecture effectiveness because users cannot accurately report on findability without actually attempting to find things. Most critically, they cannot capture unarticulated needs—the latent problems users haven’t consciously identified but observational research would reveal.

The gap exists because of psychological and cognitive factors. Identity theory suggests self-reports measure not just what people do, but who they think they are. When surveys ask about normative behaviors like productivity or exercise, respondents answer based on their ideal self rather than actual behavior. Nielsen’s framework shows users’ self-reported data is “three steps removed from truth”: people bend truth toward social acceptability, they report what they remember (and human memory is fallible), and in reporting what they remember, they rationalize their behavior. As Nielsen concludes: “Users do not know what they want. To design the best UX, pay attention to what users do, not what they say.”

Ten major types of response bias systematically distort survey data: recall bias where memory for small details decays, recency bias giving more weight to recent events, social-desirability bias where users conform to norms, prestige bias making themselves seem impressive, acquiescence bias tending toward agreement, order effects favoring options at beginning and end of lists, current-mood bias affecting all responses, central-tendency bias avoiding extreme ratings, demand characteristics where awareness of researcher aims changes responses, and random-response bias where users guess when uncertain. For enterprise contexts, add reference bias where peers influence self-assessment standards, and cultural response differences where Asian respondents show 30% higher midpoint selection while Mediterranean respondents exhibit 25% more extreme responding.

Alternative methods capture what surveys cannot

For organizations conducting enterprise intranet redesigns, a multi-method approach combining behavioral observation with attitudinal measurement is essential. Field studies and ethnographic research involve observing employees in their natural work environment as they perform real tasks. For global energy companies like Aramco, BP, or Chevron, this means shadowing field workers at operations sites, engineers accessing technical specifications, and administrators processing permits. Nielsen Norman Group reports that field studies with just 10 users typically reveal major pain points and big-picture issues, discovering the realistic context of interruptions, the workarounds employees have developed, and the human relationship patterns of who asks whom for help.

Contextual inquiry applies a “master-apprentice” model where users teach researchers their work in 60-90 minute sessions. The technique involves asking users to “imagine I am your student—show me one step after the other, everything I need to know” while the researcher asks questions throughout. For large organizations like Aramco and similar enterprises, this means having employees teach researchers how to submit expense reports, prepare for safety briefings, or find technical specifications for equipment. This method excels at revealing complex processes and the “why” behind behaviors that surveys cannot capture.

Usage analytics provides quantitative data about actual intranet behavior from server logs and specialized analytics tools. Key metrics include active users versus registered users, most and least visited pages, failed searches revealing findability problems, and click paths showing user journeys. Nielsen Norman Group found that lost productivity from poor intranet usability costs up to $15 million annually for companies with 10,000 users compared to top-rated intranets. Analytics reveals actual behavior patterns without self-report bias, establishing baselines before redesign and benchmarks for measuring improvements post-launch.

Card sorting and tree testing address information architecture specifically. Card sorting with 30 participants reveals how different user groups—engineers, field workers, and administrative staff at multinational corporations—naturally organize content, testing whether categories like “Operations,” “Technical,” or “Field Services” resonate across multilingual interfaces. Tree testing validates the proposed navigation structure by having 50-100+ participants navigate text-only hierarchies to find specific items, measuring task success rates and directness. The Scottish Government achieved top 3 ranking among 15 intranets for task completion and speed by using these methods.

Iterative usability testing with 5-8 participants per round remains the gold standard for identifying specific interface issues. Testing should occur in three rounds: paper/wireframe prototypes allowing rapid changes, interactive prototypes testing key interactions, and high-fidelity prototypes validating the complete experience. For global enterprises operating across multiple regions, testing must account for mobile usage among field workers, multilingual interfaces, and culturally appropriate design patterns. The Nielsen Norman Group has tested 57 intranets over 18 years with 285+ employees across multiple countries, consistently finding that observational testing reveals issues surveys never capture.

The recommended research plan for large multinational organizations involves parallel tracks during discovery: quantitative baseline surveys combined with qualitative field studies and contextual inquiry. Information architecture design should use card sorting to generate options and tree testing to validate structures. Design and prototyping requires three rounds of iterative usability testing while maintaining continuous analytics monitoring. This comprehensive approach balances survey efficiency for broad attitudinal measurement with observational methods that reveal actual workflows and usability issues.

AI-powered tools dramatically accelerate research analysis

The UX research landscape has transformed with AI integration. According to 2024 reports, 56% of UX researchers now use AI tools, up 36% from 2023, with 91% open to using them. The primary driver is efficiency: 58% cite improved team efficiency, 57% report faster turnaround times, and real users document analysis time reductions from two weeks to just two days.

Looppanel leads the qualitative analysis category with 95%+ accurate transcription across 17 languages in 3-5 minutes, AI-powered automatic note-taking organized by research questions reducing review time by 80%, smart thematic tagging that auto-categorizes research, and one-click executive summaries with evidence-backed insights. At $27/month with unlimited collaborators, it represents exceptional value for teams conducting frequent user interviews. One researcher reported: “Analysis time reduced from two weeks to just two days.”

BuildBetter.ai offers an all-in-one approach with universal AI search across all research data including calls, tickets, and documents, integration with 100+ tools including Zoom, Slack, Jira, and Salesforce, and an AI chat assistant for querying research data. Teams report 43% more time on revenue-driving activities, 18 hours saved per two-week sprint, and 26 fewer meetings per month. At $7.99-$200/month with unlimited seats, it’s particularly cost-effective for organizations needing to consolidate diverse data sources.

Dovetail functions as an enterprise research repository with centralized storage, AI-powered tagging and pattern detection, and thematic analysis across large datasets. Product Manager Eric Liu reported: “Dovetail reduced my workload from 100 hours to 10 hours to share customer insights.” However, users note limitations: high per-seat costs, complex taxonomy requiring significant upfront planning, AI features “not well integrated into researcher workflow,” and 90% transcription accuracy lower than competitors like Looppanel at 95%+.

For behavioral analytics and session replay, the market offers several tiers. FullStory provides pixel-perfect session replay with high fidelity, OmniSearch with AI-powered filtering, automatic detection of rage clicks and error clicks, and integration linking session replay to support tickets. It excels for deep technical debugging and customer support use cases but costs approximately 3x LogRocket pricing. Quantum Metric serves large enterprises needing real-time analytics and anomaly detection, AI-driven prioritization based on business impact, and quantification of revenue impact from issues, starting around $50,000/year. Microsoft Clarity offers completely free unlimited session recordings and heatmaps with no data volume charges, making it ideal for budget-conscious teams or proof-of-concept projects, though with fewer advanced features than paid alternatives.

For continuous feedback and sentiment analysis, tools like Sprig enable in-product feedback with AI-powered analysis of open-ended responses, real-time sentiment and emotion detection, and in-app surveys with heatmaps and session replays. Qualtrics XM and Medallia serve enterprise needs with advanced text analytics, sentiment analysis across 100+ languages, pattern detection across customer and employee feedback, and integration with CRM, support, and HR systems.

The critical consideration for enterprise adoption involves security, compliance, and integration requirements. Essential certifications include SOC 2 Type II, GDPR compliance, and for healthcare contexts, HIPAA compliance. Organizations must ensure AI tools don’t use research data for model training—opt-out options are essential. Integration with collaboration tools (Slack, Teams, Zoom), design tools (Figma, Sketch), product management platforms (Jira, Asana), and intranet systems (SharePoint, Confluence, Google Workspace) determines practical adoption success.

Self-learning systems represent the future of continuous insight

The most significant shift in user research involves moving from periodic surveys to continuous, always-on listening systems powered by AI and self-learning algorithms. Organizations are recognizing that traditional surveys face a critical response crisis: declining response rates often under 10% create nonresponse bias, the average time lag from survey deployment to actionable insights spans 21 months, and users increasingly ignore survey requests due to oversaturation. This crisis accelerates adoption of passive behavioral analytics as supplement or replacement for explicit feedback.

Microsoft Viva exemplifies mature AI-driven workplace personalization. Viva Skills combines Microsoft Graph capturing employee activity signals across Microsoft 365 with LinkedIn Skills Graph mapping 39,000 unique skills globally. The system uses AI to infer employee skills from work activities including emails, documents, meetings, chats, and collaboration patterns—providing organizational leaders with dashboards showing skill distribution, gaps, and opportunities while delivering personalized learning recommendations through Viva Learning. The 2024-2025 Copilot integration includes a dashboard tracking AI usage patterns and a benchmarking feature comparing individual AI usage against team and company averages, creating implicit pressure for adoption. As one observer noted: “If you were worried about your boss knowing that you avoid Copilot at all costs, it’s probably time to say hello to the AI companion.”

Notion AI demonstrates rapid enterprise adoption of self-learning capabilities, reaching $500M in annualized revenue by September 2024 with AI as the primary growth driver. Over 50% of enterprise customers now pay for AI features, up from 10-20% in early 2024. Notion Agents can execute multi-step tasks, work across pages and databases, and pull in context automatically, with memory pages allowing agents to learn user preferences for formatting and aesthetic choices. This creates context-aware suggestions based on workspace content and usage patterns, with behavioral triggers surfacing relevant information based on access patterns.

Modern behavioral analytics platforms demonstrate sophisticated implicit understanding of user needs. Amplitude’s “Signal” feature auto-detects significant behavior changes with predictive analytics identifying likely converters or churners at 87% accuracy. Behavioral cohort analysis reveals patterns like “users who create first automation within three days have 8.2x higher retention.” Heap’s automatic capture records every interaction without manual event setup, enabling retroactive analysis—you can analyze past behavior you didn’t anticipate needing to track. Session replay adds qualitative context showing why users behave in specific ways.

Academic research provides the foundation for these capabilities. Studies using EEG sensors, eye-tracking, and physiological signals demonstrate that implicit behavioral cues can reliably detect emotion and intent without explicit user reports, achieving approximately 70% accuracy in emotion recognition. Research published in ACM describes reinforcement learning-based frameworks for intelligent adaptation of user interfaces that learn from past adaptations to improve decision-making capabilities. However, self-learning systems face significant limitations: context blindness where behavioral data lacks the “why” behind actions, confirmation bias risk creating echo chambers by over-optimizing for observed patterns, and cold start problems where new users receive generic experiences until sufficient data accumulates.

Organizations moving away from traditional surveys report substantial benefits: 50-70% reduction in help desk call volume when behavioral analytics identify issues proactively, 30-40% decrease in survey deployment as passive listening systems mature, and 2.7x improvement ratio when research integrates into business decisions versus rarely incorporated. Microsoft, Adobe, Slack, and major telecom operators layer multiple insight sources including social listening, behavioral analytics, and support ticket analysis rather than relying on single NPS scores. Amazon pioneered this shift with recommendation engines as implicit research at scale, A/B testing infrastructure continuously optimizing without explicit user input, and behavioral cohort analysis identifying patterns across millions of users.

The timeline for adoption shows a clear progression. 2025-2026 represents the hybrid era where 60-70% of organizations use AI in some aspect of UX research, behavioral analytics becomes standard, traditional surveys decrease by 30-40% while remaining for validation, and human researchers focus on interpretation and strategic decisions. 2027-2028 brings predictive maturity where AI proactively identifies issues before users complain, self-learning systems handle interface adaptations automatically, survey volumes drop to 40-50% of 2024 levels, and the research role shifts heavily toward education and enablement. 2029-2030 sees autonomous insights where AI handles 70%+ of routine research activities end-to-end, continuous real-time insight generation becomes the baseline, and traditional surveys limit to specialized contexts like high-stakes decisions, novel domains, and vulnerable populations.

Critical limitations demand human oversight

Despite rapid AI advancement, critical limitations require continued human involvement in user research. The synthetic users controversy illustrates the boundaries of AI capabilities. Nielsen Norman Group assessed synthetic users in June 2024, concluding: “UX research needs real users. Synthetic users cannot replace the depth and empathy gained from studying and speaking with real people.” The research found synthetic users provide shallow or overly favorable feedback, generate long lists of needs without understanding priority, and miss the critical nuances where real participants share messy truths about abandonment and context changes.

The empathy problem persists across AI research tools. Maze’s 2025 research highlights that “AI excels at data processing and pattern recognition, [but] human researchers remain essential for empathy, critical thinking, stakeholder communication, and contextual understanding.” Nielsen Norman Group’s 2024 evaluation of AI-powered UX research tools found critical issues: tested tools were text-only and unable to analyze actual video of usability test sessions, marketing promises like “eliminate bias” or “analyze usability tests” proved inaccurate, transcript-only analysis misses participant confusion and UI misunderstandings, hallucination risk where AI confidently presents false information, and the requirement for constant verification—”if double-checking isn’t possible, don’t use it.”

The ethical concerns around passive behavioral tracking and AI-driven personalization create a surveillance paradox. Users may not understand the extent of behavioral tracking and data collection occurring without explicit permission. The line between personalization and surveillance blurs as adaptive systems create tension between user benefit and organizational control. Performance monitoring disguised as personalization reduces autonomy as systems increasingly guide work patterns, while data portability and ownership questions remain unresolved. As the UX Trends 2025 report warns: “Personalization has gotten so complex that it’s now out of human control, and can lead to echo chambers, warped perspectives, and consequences we’re unable to predict.”

Bias amplification represents another critical challenge. ML models trained on historical data inherit and amplify existing biases including gender bias and generalized stereotypes. AI systems trained primarily on Western internet data struggle with emerging markets and underrepresented populations. Without diverse training data, systems produce inaccurate insights for minority user groups. The lack of validation creates another problem: without explicit feedback as ground truth, how do we know if ML inferences are accurate? Most organizations still use periodic surveys to validate their behavioral models, acknowledging the limitations of purely implicit understanding.

The optimal path forward involves layered research ecosystems that strategically deploy behavioral analytics as the continuous foundation, AI-powered analysis for speed and scale, targeted surveys for validation and specific questions, qualitative research for depth and context, and human interpretation for strategy and stakeholder alignment. As Cheryl Couris from Cisco summarized: “AI is a co-pilot, not a replacement—using AI to augment research has helped us do more, faster.” The next 3-5 years will see dramatic changes in how UX research is conducted, but the profession will evolve toward strategic insight leadership rather than disappear—professionals who leverage AI tools while maintaining empathy, critical thinking, and contextual understanding that only humans provide.

When surveys work versus when they fail

Understanding when surveys are appropriate requires Nielsen Norman Group’s dimensional framework examining three key dimensions. Attitudinal versus behavioral: use surveys when you need to understand what people think or feel, use observational methods when you need to know what people actually do, remembering that “very often the two are quite different.” Qualitative versus quantitative: qualitative studies generate data by observing or hearing directly, quantitative studies gather data indirectly through instruments, and surveys are quantitative but can include qualitative elements. Context of use: whether the research occurs in scripted controlled settings, natural unscripted environments, without using the product, or with limited product forms.

Surveys are appropriate when you need to measure attitudes, satisfaction, or stated preferences at scale, gather demographic information, complement qualitative findings with quantitative data, track metrics over time like NPS or SUS, identify potential issues to investigate further, or work within limited resources requiring quick inexpensive insights. Surveys are not appropriate when you need to understand actual user workflows or behaviors, investigate usability or findability, understand why users behave certain ways, conduct early discovery without knowing what questions to ask, need detailed contextual information about problems, study behaviors that are hard to recall or count, or observe actual task performance.

For enterprise intranet research specifically, surveys should be secondary methods always paired with behavioral observation, directional tools used to identify areas needing deeper investigation, attitudinal measures focused on what surveys actually measure rather than claimed behaviors, critically analyzed and interpreted with full awareness of bias limitations, and properly timed by administering immediately after experiences rather than retrospectively. The data tells you how users feel about their workflows, not how they actually work.

Best practices for reducing bias when surveys must be used include distributing immediately after relevant experience, using surveys in connection with observational methods, emphasizing confidentiality for sensitive topics, providing response ranges rather than exact numbers, using semantic-differential scales rather than Likert scales, randomizing question and response order when appropriate, keeping surveys as short as possible, pilot testing with 4+ rounds before deployment, using validated instruments like SUS or SEQ when appropriate, and always pairing with behavioral performance metrics.

For large multinational organizations, the critical insight is that survey-based research alone systematically misses contextual factors determining success or failure. With 50,000+ employees across multiple countries and diverse roles from field workers to engineers to executives, surveys will overestimate satisfaction by 10-15 percentage points through non-response bias, miss safety-critical workflow nuances that field studies would reveal, and fail to capture the multilingual and cross-cultural usage patterns essential for successful intranet adoption. Investment in comprehensive multi-method research (estimated $150,000-300,000 for enterprise-scale projects) is justified by potential productivity savings of $15M+ annually and the critical nature of effective communication in safety-sensitive operations across global industrial environments.

Structuring case studies about vendor collaborations

For large multinational organizations presenting intranet redesign projects with vendor collaborations like Flying Bisons, a process-focused methodology case study offers an effective structure that handles confidentiality constraints while demonstrating research rigor and critical reflection. The key is framing expectations early: “This case study focuses on the research methodology and collaborative process for an ongoing project. As the project is still in progress and results are confidential, this reflection examines our approach and decision-making.”

The vendor collaboration should be framed with clear role definition early in the case study: “My role as internal UX researcher was to lead user research strategy and ensure alignment with organizational culture, while collaborating with Flying Bisons who provided specialized survey design expertise and technical implementation.” Use language that includes collaborators while clarifying internal leadership: “In collaboration with Flying Bisons, our internal SAIL lab led the research approach, with specific responsibility for stakeholder management, bilingual adaptation, and cultural appropriateness.”

Balance contributions by highlighting your unique value: “While the vendor provided survey design best practices, I ensured questions aligned with our unique organizational culture, multilingual needs, and the specific workflows of field workers, engineers, and administrators across global operations.” Show critical thinking about vendor recommendations: “The vendor proposed a standard 15-minute survey, but given our context of high survey fatigue and mobile field workers with limited connectivity, we adapted it to a focused 7-minute mobile-optimized format with offline capability.”

The recommended structure follows a process journey narrative: Context and challenge describing the research team mission and intranet redesign needs without revealing sensitive information. Research philosophy and approach explaining why survey research was selected and the guiding principles of multilingual accessibility, inclusive sampling, and cultural appropriateness. Methodology design process detailing the survey design journey from scoping through collaborative design with vendor, multilingual adaptation, pilot testing, and refinement. Vendor collaboration model describing how the partnership was structured, division of responsibilities, decision-making framework, and what was learned about effective collaboration. Implementation and adaptation covering challenges encountered, solutions implemented, mid-course adjustments, and real-time data quality management. Critical reflections discussing what’s working well, unexpected challenges, what would be done differently next time, and skills being built. Looking ahead describing next phases without revealing results and broader implications for the organization’s research practice.

For handling confidentiality constraints, use percentages rather than absolutes: instead of “survey received 1,247 responses,” say “survey achieved 47% response rate” or “participation exceeded our target by 20%.” Generalize specifics: instead of “3,500 daily active users across 12 departments,” say “a large-scale enterprise intranet serving thousands of employees.” Focus on transferable insights emphasizing methodology and process over proprietary details, discussing challenges and solutions at conceptual level, and sharing frameworks applicable elsewhere. What you can share includes research methodologies used, your role and responsibilities, general problem statement and goals, design process and iterations, challenges faced and solutions, decision-making rationale, skills and tools employed, and general outcomes. What to avoid includes specific financial metrics, competitive advantages, proprietary methodologies, actual user data or screenshots with identifying information, internal politics, unreleased features, specific vendor pricing, and contract details.

The critical reflection framework for ongoing projects should address structured questions: What did we set out to do, including initial objectives and assumptions? What have we done so far with methods and decisions made? What’s working well in effective processes and valuable insights emerging? What challenges have we encountered methodologically and logistically? What would we do differently with alternative approaches and lessons learned? What questions remain with uncertainties to resolve and future research directions? This reflection demonstrates professional maturity and continuous learning—exactly what makes an excellent UX researcher capable of adapting methodology based on real-world constraints and emerging insights.

Practical recommendations for enterprise research teams

The research synthesis reveals a clear strategic direction for enterprise UX research at organizations like Aramco and similar global corporations. Near-term priorities should focus on implementing hybrid approaches that layer behavioral analytics onto existing survey research rather than replacing wholesale, establishing continuous listening infrastructure through Voice of Customer platforms unifying multiple feedback sources, upskilling research teams with data science and ML interpretation capabilities, establishing ethical guidelines with clear policies on AI use and data privacy, and maintaining human touchpoints preserving qualitative research for context and validation.

For the current intranet redesign project, critically reflect on survey limitations in your article revision by explicitly acknowledging that surveys capture only attitudinal data about perceptions and preferences, not the behavioral reality of actual workflows. Frame survey findings as directional insights requiring validation through observational methods. Discuss the specific biases affecting your survey: non-response bias likely overestimating satisfaction by 10-15%, recall bias affecting accuracy of workflow descriptions, and social-desirability bias especially in organizational contexts where employees may fear consequences of negative feedback. Position the survey as one component of a comprehensive research strategy rather than the primary source of truth.

The vendor collaboration with Flying Bisons should be presented as a strategic partnership where external expertise in survey design methodology complemented internal knowledge of organizational culture, multilingual needs, and diverse user populations across global operations. Highlight specific adaptations made: adjusting survey length for mobile field workers, ensuring cultural appropriateness for multilingual respondents across different regions, sampling across all employee segments from field operations to engineering to administration, and maintaining internal control over research questions to align with business objectives. Demonstrate critical thinking by discussing what was learned about the limitations of chosen methodologies and what alternative or complementary methods would be employed in future phases.

Position enterprise research teams as forward-looking by discussing AI-powered alternatives and self-learning systems as the future direction. Reference specific tools under evaluation: Looppanel for interview analysis in future qualitative phases, Quantum Metric or FullStory for continuous behavioral analytics post-launch, Microsoft Clarity as a budget-conscious option for session replay, and adaptive intranet search solutions like Glean or Guru for the redesigned platform. Discuss how future research will shift from periodic surveys to continuous implicit feedback through usage analytics, behavioral pattern detection, and adaptive personalization learning from actual user interactions without explicit surveys.

The article revision should emphasize that the next 3-5 years will see enterprise research teams evolving from periodic survey-based research to continuous AI-augmented insight generation, from measuring attitudes to tracking actual behaviors, from asking users what they want to observing what they actually do, and from reactive problem-solving to proactive issue detection through predictive analytics. However, maintain the critical perspective that AI augments rather than replaces human researchers—the research team’s role will shift toward strategic insight leadership, ethical AI oversight, stakeholder communication, and the empathy and contextual understanding that only humans provide. This positions research organizations as both critically reflective about current methodology and strategically positioned for the future of user research in large enterprise contexts.

The comprehensive research plan for large multinational enterprises should include parallel discovery tracks combining quantitative baseline surveys with qualitative field studies and contextual inquiry, information architecture design using card sorting and tree testing validated with 75-100+ participants, iterative design validation through three rounds of usability testing with 8 participants each, continuous post-launch monitoring through analytics dashboards and behavioral pattern detection, and periodic targeted surveys only for validation and measuring specific attitudinal constructs where surveys excel. Budget realistically at $150,000-300,000 for comprehensive research programs at enterprise scale, recognizing this investment is justified by potential $15M+ annual productivity savings and the critical importance of effective communication in safety-sensitive operations across global organizations.

Conclusion

The future of enterprise user research requires a fundamental reconceptualization of the researcher’s role. Traditional surveys will persist but occupy a narrower niche—measuring attitudes and tracking satisfaction while acknowledging inherent limitations. The center of gravity shifts toward behavioral analytics, AI-powered insight generation, and self-learning systems that understand users through what they do rather than what they say. For organizations like Aramco and similar global enterprises, success requires embracing this transition while maintaining the human elements of empathy, critical thinking, and contextual understanding that technology cannot replicate.

The evidence is unambiguous that surveys alone are insufficient for enterprise intranet redesign, systematically overestimating behavior frequency by 2x, predicting only 25% of actual usability from stated preferences, and missing the contextual workflow details essential for effective design. Yet surveys retain value for specific purposes when properly deployed, validated against behavioral data, and interpreted with full awareness of limitations. The optimal approach layers multiple methods: behavioral analytics as the continuous foundation, AI-powered analysis for speed and scale, targeted surveys for validation, qualitative research for depth, and human interpretation for strategy.

As AI capabilities mature over the next 3-5 years, the question becomes not “will AI replace user researchers?” but rather “how can we design AI systems that amplify human understanding without losing the essential human connection that makes experiences truly meaningful?” The answer lies in strategic insight leadership—professionals who leverage AI tools to handle routine analysis while focusing their uniquely human capabilities on empathy, ethical oversight, stakeholder communication, and the critical thinking that transforms data into actionable wisdom. This is the future toward which enterprise research teams at organizations like Aramco and similar global corporations should position themselves: critically reflective about current methodology, strategically invested in AI-powered capabilities, and confidently human in the value they provide.

About the Author

I work at the intersection of design, technology, and human experience—crafting intelligent systems that amplify human capability rather than replace it. As a Digital Experience Design Architect, my practice is grounded in a belief that the most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and create.

My approach combines rigorous methodology with creative vision. I question assumptions, challenge conventional wisdom, and seek patterns that others might miss. Whether exploring user research methodologies, designing enterprise systems, architecting digital experiences, or examining broader societal challenges, I maintain a critical lens that asks not just “what works” but “why it works” and “for whom does it work best.”

Each article I write reflects this philosophy: technology should expand our creative horizons, design should serve genuine human needs, and innovation should be tempered with wisdom about its implications. I write to share insights, provoke thought, and invite others into conversations about how we can build a future where human creativity and technological capability work in genuine partnership.

For those eager to explore further:

Subscribe to User First Insight for perspectives on design, technology, and human experience in enterprise contexts. For broader explorations of sustainability, global politics, and societal challenges, follow Black & White where I examine clear perspectives on the issues that matter and practical ways to solve them. My book Unfinished: Notes on Designing Experience in a World That Never Stops Changing offers deeper exploration of design philosophy in an age of constant transformation. Connect with me on LinkedIn for professional conversations, follow my writing on Medium for additional insights and case studies, and visit haiderali.co and stayunfinished.com to see how these ideas manifest in practice.

This is more than content—it’s an invitation to question, to evolve, and to reimagine what becomes possible when we approach both technology and society with critical thinking and thoughtful action. The tools keep changing, the challenges keep evolving, but the mission remains constant: creating experiences and solutions that genuinely improve how people live, work, and create.

The digital landscape is a vast ocean of potential, and I’ve learned that true innovation requires more than just technical skill—it demands a unique perspective, a willingness to challenge the status quo, and the courage to embrace continuous transformation.

The Genesis of a Digital Manifesto

My journey with StayUnfinished began not as a website, but as a philosophy—a belief that technology should amplify human potential rather than replace human creativity. Working at the Saudi Accelerated Innovation Lab (SAIL) in Aramco, I discovered that innovation is less about the tools we use and more about the mindset we cultivate.

A Framework for Technological Adaptation

Through years of experience, I’ve developed a three-stage approach to meaningful technological innovation:

1. Curiosity Stage: Embracing the Unknown Technological innovation starts with radical curiosity. This isn’t about blindly adopting every new tool, but about creating a mindset of open exploration. At SAIL, we learned to approach new technologies with wonder rather than skepticism. Each tool, each platform becomes a window into new possibilities.

2. Integration Stage: Strategic Implementation Innovation requires thoughtful implementation. When I chose Astro for StayUnfinished, it wasn’t a random selection but a strategic decision. The platform embodies a philosophy of minimal complexity and maximum impact. Its architecture represents a new way of thinking about web development:

   • Zero Client-Side JavaScript by Default: Creating lightning-fast pages

   • Hybrid Rendering: Intelligent component loading

   • Markdown-First Authoring: Focusing on content over configuration

3. Amplification Stage: Scaling Insights True innovation isn’t about keeping knowledge private—it’s about sharing, documenting, and creating reproducible methodologies. StayUnfinished became more than a website; it transformed into a living case study of technological adaptation.

A Historical Lens: The Caravel of Digital Exploration

During my time at SAIL, I often used historical analogies to illustrate technological transformation. The Portuguese caravel serves as a powerful metaphor—a breakthrough naval technology that redefined global exploration in the 15th century.

Imagine traditional shipbuilders resistant to change, clinging to existing design methodologies. Meanwhile, Portuguese navigators saw beyond immediate limitations. By combining square and lateen sails, they created a vessel that wasn’t just an improvement—it was a quantum leap in maritime technology. This vessel didn’t just travel faster; it expanded the boundaries of human exploration.

The parallel with today’s technological landscape is striking. AI, rapid development tools, and adaptive frameworks are our modern caravels—not replacements for human creativity, but sophisticated amplification tools that extend our collective capabilities.

Practical Wisdom for Aspiring Innovators

For those feeling overwhelmed by technological change, here are strategic approaches to navigate the digital landscape:

1. Cultivate a Learner’s Mindset

   • Dedicate 5-10 hours monthly to exploring emerging technologies

   • Follow thought leaders across diverse domains

   • Participate in cross-disciplinary communities and workshops

2. Build a Personal Innovation Toolkit

   • Curate adaptive tools that align with your creative process

   • Develop lightweight, reusable project templates

   • Create personal documentation systems that capture your learning journey

3. Develop a Network of Diverse Perspectives

   • Connect with professionals outside your immediate field

   • Attend conferences that challenge your existing understanding

   • Engage in knowledge exchange platforms that promote cross-pollination of ideas

The Philosophical Underpinning

Technology is more than efficiency—it’s a lens for reimagining human potential. We’ve evolved from stone tools to moon landings, from manual labor to AI-assisted creation. Each technological wave doesn’t replace human judgment; it elevates our collective capability.

StayUnfinished: More Than a Website

This project represents a living manifesto about continuous learning, technological empowerment, and the art of staying perpetually curious. It’s a testament to the idea that our work is never truly complete—it’s always evolving, always expanding.

About the Author

In a world of rapid technological transformation, I’ve positioned myself as a bridge between human creativity and technological potential. As a Digital Experience Design Architect, my mission extends beyond coding or design—I craft intelligent systems that amplify human capability.

My work explores the nuanced intersection of design, technology, and human experience. Each project is more than a product—it’s a philosophy, a statement about how we can leverage technology to expand our creative horizons.

For those eager to dive deeper:

• Subscribe to User First Insight

• Explore Black & White

• Connect on LinkedIn

• Follow my writing on Medium

• Visit haiderali.co

• Explore stayunfinished.com

This isn’t just a website or a book—it’s an invitation to reimagine what’s possible when human creativity meets technological innovation.

The site is live, the DNS is propagating, and the future is loading—one millisecond at a time.

Every time you interact with an AI system—whether you’re asking a conversational interface to summarize a document, watching a recommendation algorithm surface relevant content, or relying on automated systems to flag potential risks—you’re standing on scaffolding built nearly seventy years ago by a man who never witnessed any of these implementations.

In 1956, when “computer” still meant a person who computed rather than a machine, John McCarthy gathered researchers at Dartmouth College and made a proposition that seemed almost absurd: “Every aspect of learning or any other feature of intelligence can, in principle, be so precisely described that a machine can be made to simulate it.”

That sentence didn’t just name a field. It set a direction that would eventually lead to the entire ecosystem of artificial intelligence we interact with today—from the large language models that power conversational interfaces like ChatGPT, Claude, and Gemini, to the computer vision systems that help diagnose medical images, to the predictive algorithms that optimize supply chains. ChatGPT and similar chat interfaces are simply one visible layer—a frontend that makes it easy for humans to communicate with these underlying AI systems. The real intelligence lives in the models, the architectures, and the computational approaches that McCarthy helped establish.

The Invisible Architecture We Still Build Upon

McCarthy’s influence extends far beyond coining the term “Artificial Intelligence.” He constructed the conceptual and technical infrastructure that makes modern AI possible, though most designers and developers today wouldn’t recognize his fingerprints on their work.

Consider LISP, the programming language he created in the late 1950s. While few people write LISP today, its core philosophy—symbolic reasoning, recursive thinking, treating code as data—shaped how we conceptualize intelligent systems. When you design a chatbot that reasons through nested conditional logic, you’re following mental models McCarthy established decades ago.

Then there’s time-sharing, his vision of multiple users accessing a single powerful computer simultaneously. In 1960, this was revolutionary. Today, we call it cloud computing, and it’s the foundation on which every generative AI model runs. The architecture that lets millions of people query GPT-4 simultaneously? McCarthy sketched that pattern when mainframes filled entire rooms.

But perhaps his most enduring contribution was philosophical: he saw AI not as replacement but as augmentation. Machines assisting humans in reasoning, planning, and decision-making. This human-centered view of automation feels remarkably prescient when we’re now grappling with how to design AI that collaborates rather than dictates.

What McCarthy’s Legacy Means for Modern Design Practice

I’ve spent seventeen years moving from visual design to digital experience architecture, and I’ve noticed something curious. Tools change constantly—the design software I used in 2007 is obsolete, the prototyping methods have evolved three times over—but the underlying patterns of systems thinking remain remarkably stable. McCarthy’s work reveals three of these durable patterns that matter as much today as they did in 1956.

First, naming crystallizes thinking. When McCarthy called it “Artificial Intelligence,” he gave researchers and builders a shared frame of reference. This wasn’t just semantics; it was strategic clarity. In my projects, I’ve learned that naming the problem space crisply creates alignment faster than any tool choice. The difference between “we need AI” and “we need to reduce cognitive load in approval workflows by surfacing risk signals early” is the difference between wandering and shipping.

Second, design for collaborative scale from the start. McCarthy’s time-sharing wasn’t merely technical innovation; it was a collaboration pattern. Modern AI products must consider multi-user environments where people and models co-create. Who sees what? Who can override the AI’s suggestion? How do teams review and refine outputs collectively? These aren’t implementation details—they’re foundational design decisions.

Third, treat intelligent systems as capable colleagues, not oracles. The most successful AI features I’ve architected behave like good teammates: they explain their reasoning, accept feedback, adapt to corrections, and visibly improve over time. This is fundamentally a design challenge. An AI that generates a perfect output but can’t explain its logic is less useful than one that produces good work and shows its reasoning.

From SAIL to SAILs: Testing Ideas in the Real World

McCarthy didn’t just theorize about artificial intelligence from a distance. In 1963, he helped establish SAIL—the Stanford Artificial Intelligence Laboratory—where abstract ideas collided with the messy reality of systems, tools, and actual people using them. SAIL became legendary not because it produced perfect systems, but because it created an environment where researchers could fail fast, learn deeply, and refine their thinking through direct contact with problems.

That experimental spirit finds a direct descendant in my current work at Aramco’s SAIL, the Saudi Accelerated Innovation Lab. The name is no accident—it’s an intentional nod to Stanford’s legacy and an aspiration to build the same culture of rigorous experimentation for Saudi Arabia. Like its Stanford predecessor, our lab operates as a proving ground where we validate what genuinely improves decisions and reduces friction, rather than chasing whatever feels novel this quarter.

Our work at SAIL operates through a decentralized model where our Digital Experience Design Architect team collaborates across organizational boundaries—with designers embedded in business units, external vendors building our tools, and internal product teams shipping features. We don’t wait for problems to arrive at our door; we take initiative, moving between discovery research, rapid prototyping, and AI-assisted analysis to spot opportunities and build innovative solutions before they become urgent requests.

A typical week might involve working with a procurement team to understand workflow friction, prototyping an AI-assisted approval interface with an external vendor, and collaborating with an internal product team to instrument decision-clarity metrics across their existing tools. The work is about turning fuzzy problems—”approvals take too long” or “we miss critical risks”—into testable flows that we can measure, refine, and scale across the enterprise.

What makes this feel connected to McCarthy’s legacy is our insistence on certain practices. We measure usability not through subjective satisfaction scores but through task success rates, time-to-clarity metrics, and error-recovery patterns. When users stumble, we want to know precisely where and why. We design prototypes that directly inform governance policies, ensuring that questions about roles, permissions, and audit trails get answered during design rather than after deployment. We build AI-assisted research tools that provide summaries with full citations and visible rationale, always asking “Why this recommendation?” and creating correction loops so the system learns from its mistakes.

Perhaps most importantly, we design within enterprise constraints from day one. Security boundaries, data governance requirements, and on-premises integration needs aren’t obstacles we work around—they’re design parameters we work within. This constraint-aware approach prevents the common tragedy where a brilliant prototype dies in the gap between lab and production.

The goal remains what McCarthy championed seven decades ago: augmentation, not replacement. We’re building systems that explain their reasoning, adapt based on feedback, and scale responsibly across teams. Systems that make expertise more accessible without pretending to replace the judgment that comes from years of experience in a domain.

From Philosophy to Practice: A Working Framework

When I design AI-assisted experiences, whether in our SAILs lab or in broader consulting engagements, I use what I think of as the McCarthy Method—a framework inspired by his approach to breaking down intelligence into describable, machine-simulatable components.

I start by defining what decision or task we’re actually trying to clarify or accelerate. Not “add AI to the dashboard” but “help product managers identify which customer requests signal market shifts versus individual edge cases.” Specificity forces clarity.

Then I decompose the problem. What inputs, constraints, and context does the model need to do quality work? If a human expert would need customer segment data, usage patterns, and competitive intelligence to make this call, the AI needs structured access to the same information.

Next comes dialogue design. How will the system explain its output? Not just what it concluded, but why? How will users correct it when it’s wrong? This correction loop isn’t a nice-to-have—it’s how the system learns what “good” means in your specific context.

Governance follows naturally. What’s the review loop? When must a human make the final call? What’s the audit trail? These questions feel bureaucratic until something goes wrong, and then they’re the only questions that matter.

Finally, evolution. Where will the system learn over time? Are we refining prompts, adding examples, or updating policies? Making this explicit prevents the common trap where AI features slowly degrade because no one owns their continued improvement.

This framework turns AI from an inscrutable black box into a transparent collaborator. It also keeps teams aligned on intent, not just interface—a distinction that saves weeks of rework.

The Unfinished Nature of Intelligence

In my book Unfinished: Notes on Designing Experience in a World That Never Stops Changing, I explore the tension between our desire for permanence and the reality of continuous evolution. AI embodies this tension perfectly. Like design itself, intelligence is never finished. It’s a living system that learns, forgets, and relearns.

McCarthy understood this. He didn’t present AI as a solved problem in 1956; he presented it as a research program that would unfold over decades. That patience, that comfort with incompleteness, is something we’ve lost in our rush to ship. The best AI products I’ve seen embrace their unfinished nature. They ship with clear limitations, obvious feedback mechanisms, and visible improvement over time.

Imagination, McCarthy reminds us, is a design material as real as code or pixels. The systems that endure are those that remain adaptable, explainable, and fundamentally centered on human needs—even as the underlying technology transforms.

Five Moves You Can Make This Quarter

If you’re designing or building AI-powered products, here are five concrete actions that embody McCarthy’s principles:

Ship an explanation pattern for every AI output. Add “Why this recommendation?” to your interface. Users don’t need to understand transformer architecture, but they deserve to know why the system suggested prioritizing bug A over bug B. Clarity earns trust faster than accuracy alone.

Create a genuine correction loop. When users edit an AI-generated summary or adjust an automated decision, capture that feedback as training data. Design the interaction so corrections feel natural, not like filing a bug report. “Accept with changes” should be as easy as “Accept.”

Measure outcomes, not outputs. Stop counting how many AI suggestions users accepted. Start measuring time to clarity, rework avoided, and friction removed. These are the metrics users actually feel, and they’ll tell you whether your AI is truly helpful or just novel.

Build a versioned prompt library for your team’s top workflows. When someone crafts a prompt that consistently produces quality results, capture it. Version it. Share it. This scales quality across your team and creates a learning artifact that improves over time.

Draft a single-page ethics and limitations document. Be explicit about data boundaries, known model weaknesses, and when humans must intervene. This isn’t legal coverage—it’s design honesty. It also prevents painful conversations six months from now.

The Long View

Before ChatGPT could respond in seconds, John McCarthy spent years asking better questions. He died in 2011, never witnessing the explosion of generative AI that now dominates our industry. Yet his fingerprints are everywhere.

Every time we discuss designing intelligence, crafting human-machine harmony, or building systems that augment rather than replace human capability, we’re walking a path he sketched more than half a century ago. We’re standing on scaffolding he built before most of us were born.

That might be the most beautiful form of design: work that outlives its designer, infrastructure that enables futures the creator never saw, questions that remain relevant across generations of answers.

The tools McCarthy used are obsolete. The problems he identified are still here. And the methods he pioneered—clear problem definition, collaborative architecture, human-centered automation—remain as relevant as ever.

Perhaps that’s the real lesson. In a field obsessed with disruption and novelty, the most powerful contributions are those that establish enduring patterns. McCarthy didn’t just predict the future. He drafted its architecture.

Haider Ali is a Digital Experience Design Architect exploring the intersection of design, technology, and human experience. His work focuses on building AI-powered systems that augment human capability rather than replace human judgment.

For more writing on design systems and intelligent interfaces, subscribe to User First Insight or explore Black & White. Connect on LinkedIn, follow on Medium, or visit haiderali.co.

After years of designing and observing digital experiences, I began collecting thoughts that never quite felt “done.” Those ideas turned into Unfinished — a book about designing in motion, about clarity in change, and about learning to see design as a living process.

📘 Read or download here → https://leanpub.com/unfinished-design

You can use:

• How to design for evolving systems, not static screens.

• The hidden patterns behind user trust and perception.

• How AI is reshaping the craft of design.

• Notes from real projects on clarity, constraints, and change.

I’d love your thoughts and reflections — this book is intentionally unfinished, because design never truly stops changing.

Over the years, I’ve shifted through a long list of tools: Illustrator, XD, Sketch, Figma, and countless plugins. Each one promised better speed, precision, collaboration, and creativity. And to be fair, each one did move the needle forward in some way. But at the heart of it, the pattern remained the same. We were designing with tools, not through intelligence.

Now, with the dawn of AI, we’re standing at a new crossroads. Many designers are still using AI tools the same way they used their legacy software, treating them as an upgrade to Figma or Photoshop. But I believe that’s the wrong lens entirely.

Our end objective isn’t to design an interface in Figma. It’s to create products that people can use effortlessly, products that solve problems, reduce friction, and deliver value faster than ever before. That shift in thinking changes everything.

The Workflow Shift

Over the past few months, I’ve been rethinking how I work, critically evaluating every step of my design workflow to remove what’s unnecessary and amplify what truly matters.

Instead of spending hours pushing pixels or fine-tuning layouts, I now design within intelligent environments. Tools like Cursor, Visual Studio Code, Windsurf, Builder.io, and IBM Watson are more than design tools. They’re AI-integrated environments that think with you, not for you.

My current choice is Windsurf, which has quietly transformed how I create, test, and iterate on ideas. I’ll share a detailed breakdown of why I prefer it in another post soon, but for now, what matters is this: the environment you work in shapes the outcomes you can achieve.

Why I Believe This Is the Future of UX

AI has effectively closed the skills gap. I no longer need to master a programming language to bring my ideas to life. The code barrier that once divided designers and developers is fading away.

Now, the most valuable skill isn’t learning syntax or shortcuts. It’s learning how to think. How to ask better questions. How to translate user needs into intelligent systems that can self-optimize and adapt.

That’s where the true value of a designer lies today: in crafting intent and experience, not just interfaces.

A Glimpse into the Future

In the last couple of months alone, I’ve built several internal AI agents to make my own workflow more productive. Each one replaced hours of manual effort with intelligent automation and freed me to focus on creative strategy and user value.

Right now, as I write this piece, I’m also working on a new enterprise-grade application designed to help employees generate polished presentations in minutes using a few well-written prompts.

Think about the scale of that. In large organizations, employees spend hundreds of collective hours every week just building slides: aligning shapes, adjusting layouts, rephrasing titles. If AI can turn that effort into a five-minute process, the productivity impact is enormous, not just in saved time, but in reclaimed focus.

From Design to Direction

This is why I believe designers should stop thinking of AI as a tool and start seeing it as a collaborator, one that reshapes how we think, not just how we work.

Our job isn’t to master the next tool. It’s to rethink the workflow itself. Because the future of UX won’t be about how beautifully we can design. It’ll be about how intelligently we can create.

In my next piece, I’ll break down how tools like Windsurf and Cursor can become part of a next-gen design workflow, one that blends creative intent with intelligent execution.

About User First Insight: This publication explores the intersection of user experience, artificial intelligence, and enterprise design. Subscribe to receive insights on building products that truly serve people.

What’s your experience with AI in design? Reply to this email or leave a comment—I’d love to hear how you’re navigating this shift.

As someone who works at the intersection of design, technology, and human experience, I’m constantly drawn to stories of systemic failure—not because I enjoy watching things fall apart, but because these moments of crisis reveal something profound about how we create and maintain complex systems. Dr. Ken Berry’s journey from conventional physician to dietary rebel offers unexpected insights for anyone designing experiences in our rapidly changing world.

The Architect Who Followed the Blueprint

Berry’s story begins exactly where mine does every morning: following established guidelines that seem unquestionable. As a family physician, he dispensed the American Diabetes Association’s official advice with complete confidence—low fat, high whole grains, abundant fruits and vegetables. The institutional framework was clear, the research seemingly conclusive, the path forward well-documented.

Then came his own health crisis. At 297 pounds with prediabetes and fatty liver disease, Berry did what any good practitioner does: he followed his own professional advice. The ADA guidelines. The expert consensus. The established playbook.

His health deteriorated.

In my book Unfinished: Notes on Designing Experience in a World That Never Stops Changing, I explore how our most dangerous assumptions are the ones we inherit without examination. Berry’s realization—”I didn’t know a damn thing about human nutrition”—mirrors moments I’ve witnessed in design organizations where perfectly logical systems produce persistently poor outcomes.

Pattern Recognition vs. Pattern Acceptance

What fascinates me about Berry’s pivot isn’t just that he changed his approach—it’s how he changed it. He didn’t abandon methodology; he questioned which methodology deserved trust. He moved from pattern acceptance (following established protocols) to pattern recognition (observing what actually worked).

This distinction matters profoundly in my work as a Digital Experience Design Architect. We’re constantly presented with frameworks: design systems, research methodologies, implementation strategies. The question isn’t whether to use frameworks—they’re essential for scaling insight—but whether we’re willing to interrogate them when they fail to serve genuine human needs.

Berry’s experimental approach—trying Atkins, exploring paleo principles, testing ketogenic eating, eventually discovering carnivore—reflects a research methodology that designers should recognize: rapid iteration based on measurable outcomes. Within three months, he had real data. Not theoretical constructs, but lived results.

The Proper Human Diet Spectrum: Lessons in Personalization

What Berry now calls the “Proper Human Diet Spectrum”—a range of carbohydrate restriction tailored to individual metabolic needs—offers a powerful metaphor for experience design. There’s no universal solution. The framework must flex to accommodate vastly different user contexts.

This mirrors challenges I navigate daily in large-scale organizational design. When architecting digital experiences for thousands of users with different roles, technical capabilities, and workflow contexts, the question isn’t “what’s the one right answer?” It’s “what’s the spectrum of viable approaches, and how do we match individuals to the right point on that continuum?”

Berry uses comprehensive lab work and continuous glucose monitoring to teach patients how different foods impact their specific metabolism. He’s not just treating disease—he’s creating a feedback system that builds user agency. This is experience design at its most fundamental: giving people tools to understand their own patterns and make informed choices.

Risk Communication and Motivation Design

One aspect of Berry’s approach particularly resonates with my work in complex enterprise environments: how he communicates risk. Rather than abstract statistics, he shows patients real-life images of diabetic complications. This isn’t fearmongering—it’s making invisible consequences tangible.

We face similar challenges when advocating for design changes in large organizations. Telling stakeholders that “poor UX will impact conversion rates” often falls flat. Showing them actual user sessions—the confusion, the frustration, the abandoned workflows—creates visceral understanding that drives action.

Berry’s warning against “false choices”—replacing junk food with equally problematic “healthy” alternatives—parallels what I call “cosmetic innovation” in digital design. Swapping one confusing interface for a differently confusing interface doesn’t solve the underlying problem. We need systemic rethinking, not surface-level redesign.

Challenging the Cholesterol Consensus: When Proxy Metrics Mislead

Berry’s stance on saturated fat and cholesterol offers crucial insight for anyone working with data and metrics. The medical establishment focused intensely on LDL cholesterol as a primary cardiovascular risk factor, while Berry argues that metabolic syndrome itself poses far greater danger.

This is a proxy metric problem that designers know intimately. We obsess over click-through rates while ignoring task completion time. We celebrate engagement metrics while users grow more frustrated with each interaction. We optimize for what’s easily measurable rather than what genuinely matters.

The question Berry asks—”Are we measuring the right thing?”—should haunt every dashboard we build, every KPI we establish, every success metric we celebrate. Sometimes the most important outcomes resist easy quantification.

Beyond Diabetes: Emergent Properties in Complex Systems

What makes Berry’s work particularly relevant to experience designers is his documentation of unexpected improvements beyond metabolic health: better mental health outcomes, reduced autoimmune symptoms, improved neurological function. These weren’t predicted by the initial framework—they emerged from system-wide changes.

This mirrors my observation that truly effective experience design often produces benefits beyond initial success criteria. When we redesign a workflow to be more intuitive, we don’t just improve task completion—we reduce cognitive load, decrease stress, improve job satisfaction, sometimes even shift organizational culture.

In Unfinished, I argue that design in our era of constant change requires embracing emergence rather than demanding complete predictability. Berry’s “reverse education” prediction—patients educating doctors through their personal success—reflects a user-driven innovation model that resonates deeply with how I see technology adoption unfolding in enterprise contexts.

Systemic Change Through Individual Agency

Berry co-founded the American Diabetes Society not through top-down reform but by empowering individual practitioners and patients to challenge mainstream protocols. This grassroots approach to systemic change mirrors what I’ve observed in large organizations: lasting transformation rarely comes from executive mandates alone. It emerges when individual contributors gain enough agency and evidence to challenge inherited assumptions.

This is why my work focuses on creating systems that amplify human capability rather than constraining it. Whether it’s continuous glucose monitoring giving patients real-time metabolic feedback or thoughtfully designed internal tools giving employees autonomy over their workflows, the principle remains constant: information plus agency drives better outcomes than prescription alone.

The Unfinished Nature of Knowledge

Berry’s journey from confident conventional practitioner to questioning explorer captures something essential about intellectual honesty in any field. His willingness to admit “I didn’t know a damn thing” required more courage than maintaining the appearance of expertise.

In my practice spanning enterprise architecture, user research, and system design, I’ve learned that the most dangerous designer is the one who’s certain they understand the problem completely. Our world changes too rapidly, our systems grow too complex, our users too diverse for any single framework to capture full truth.

This is why I titled my book Unfinished. Not because I lack commitment or rigor, but because I recognize that design, like Berry’s understanding of human nutrition, must remain open to revision. The moment we declare something “finished”—whether it’s a dietary protocol, a design system, or a research methodology—we close ourselves to the very feedback that could improve our work.

Practical Implications for Design Practitioners

What should we take from Berry’s journey into our design practice?

Question inherited frameworks. Not cynically, but methodically. When established guidelines produce poor outcomes despite faithful implementation, the problem may be systemic rather than individual.

Build feedback systems that reveal truth. Berry’s use of continuous glucose monitoring creates visceral, immediate understanding. What are the equivalent tools in your practice that make invisible patterns visible?

Personalize rather than standardize blindly. Berry’s spectrum approach acknowledges that different individuals require different interventions. How do your design systems accommodate genuine user diversity rather than forcing conformity?

Measure what matters, not what’s convenient. If your success metrics don’t align with actual user wellbeing, you’re optimizing the wrong things.

Empower users with agency. The best systems teach people to understand their own patterns and make informed choices. Are your designs building user capability or creating dependency?

Embrace productive uncertainty. Berry’s willingness to experiment, iterate, and revise based on evidence reflects intellectual humility that every designer should cultivate.

The Larger Pattern

Berry’s story represents more than one doctor’s dietary discoveries. It’s a case study in how complex systems evolve, how inherited knowledge sometimes misleads, and how genuine innovation often emerges from questioning rather than accepting established wisdom.

As I’ve argued throughout my writing on technology, design, and human experience: our most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and thrive. Whether we’re designing dietary protocols or digital experiences, the fundamental challenge remains identical—creating systems that serve genuine human needs rather than perpetuating inherited assumptions.

The world keeps changing. Our frameworks must evolve accordingly. The practitioner who admits uncertainty and pursues evidence ultimately serves users better than the expert who defends outdated consensus.

About the Author

I work at the intersection of design, technology, and human experience—crafting intelligent systems that amplify human capability rather than replace it. As a Digital Experience Design Architect, my practice is grounded in a belief that the most meaningful innovations emerge not from technology alone, but from deeply understanding how people think, work, and create.

My approach combines rigorous methodology with creative vision. I question assumptions, challenge conventional wisdom, and seek patterns that others might miss. Whether exploring user research methodologies, designing enterprise systems, architecting digital experiences, or examining broader societal challenges, I maintain a critical lens that asks not just “what works” but “why it works” and “for whom does it work best.”

Each article I write reflects this philosophy: technology should expand our creative horizons, design should serve genuine human needs, and innovation should be tempered with wisdom about its implications.

For those eager to explore further:

Subscribe to User First Insight for ongoing perspectives on design, technology, and human experience in enterprise contexts. My book Unfinished: Notes on Designing Experience in a World That Never Stops Changing offers deeper exploration of design philosophy in an age of constant transformation.

For broader discussions on societal issues, sustainability, and global politics, explore Black & White Perspective — clear perspectives on the issues that matter, and practical ways to solve them.

Connect with me on LinkedIn for professional conversations, or visit haiderali.co and stayunfinished.com to see how these ideas manifest in practice.