We’re all investing billions in artificial intelligence (AI). But we’re not investing nearly enough in designing how people actually experience it. In fact, here’s a number that should bother every UX designer working on AI products right now: 77% of workers say that AI tools have decreased their productivity. Not increased. Decreased.
That number from Upwork’s research matches what we’re seeing across the software industry. We’re in the middle of the largest wave of technology investment in a generation. US private AI investment hit $109 billion last year. Enterprise teams are shipping AI features at a pace that would have been unthinkable three years ago. And yet, by nearly every measure that matters to the people using these products, the user experience is getting worse.
Champion Advertisement
Continue Reading…
The Technology Is Not the Problem
Whenever product leaders talk about artificial intelligence, the conversation tends to default to technology decisions much faster than it should. Which AI model? How fast can we ship? How do we match the competition?
These are real questions, but they’re downstream questions. The teams producing the best AI experiences are making decisions at a level that most product organizations skip entirely. They’re asking questions about six facets of AI design that sit underneath the surface—and those questions shape everything users eventually see, touch, and trust.
The gap between AI investment and AI impact is not a technology problem. It’s a human experience problem—and most organizations aren’t oriented to see it.
Questions About Six Facets of Human-First AI Design
Let’s consider those six questions that we should all be asking to put humans first when designing artificial intelligence.
1. What Is the Role of AI?
Is artificial intelligence a tool that executes commands, a space that users navigate, or a character with a point of view? This question isn’t a branding exercise; it identifies the strategic core of the entire experience. A product team that can’t answer this question clearly will produce a user interface that confuses users at every turn. Unfortunately, most teams never ask it deliberately. This role defines the entire relational contract between users and AI. It shapes what users expect, how they interpret an AI’s behavior, and whether they feel in control or at the mercy of something they don’t understand.
2. What Should AI Actually Do?
Today, the reflex is to add artificial intelligence everywhere. But the best products are ruthlessly specific about where AI enters a workflow and where it stays out. Some tasks are too critical or too personal for automation. Some are perfect for it. UX designers must know which is which and fit AI into the moments that make the user more capable, not more dependent. Scope is a design decision, not just a product decision. The real measure isn’t what the AI can do; it’s whether the AI addresses a genuine human need and delivers something that meaningfully improves someone’s day, work, or life. Value is harder to design for than features, and it’s the only thing that users actually remember.
Champion Advertisement
Continue Reading…
3. What Does an AI Know and What Doesn’t It Know?
Context is the most under-designed dimension of AI products today. 81% of AI professionals say their company has significant data-quality issues. UX designers can’t treat this as someone else’s problem. The information architecture underneath the AI, its memory, its feedback loops, and its awareness of the user’s situation is all design material. Ignore it and no amount of user interface polish can save the experience. An AI that acts on incomplete or stale context doesn’t just underperform; it erodes the user’s confidence in the entire product. Designing for what the AI knows—and being transparent about what it doesn’t—is one of the most consequential decisions that a product team can make.
4. How Should an AI Interact?
A text box can handle about 50 words per minute. Visual processing handles 10 million bits per second. As AI capabilities multiply across user-interface modalities, the challenge isn’t adding new channels; it’s orchestrating them. Users employ voice interactions when their hands are full, shift their focus to the visual when they’re exploring, and use touch when they’re deciding. All of these modalities are grounded in patterns that are familiar enough that users don’t need a tutorial to use them. But familiar patterns can take us only so far. Multimodal design introduces new choreographic decisions. New paradigms are emerging in which UI modalities don’t just coexist but converge, blending into interaction models that have no direct precedent. Designing for this new AI frontier requires new expertise and a willingness to let go of the screen as the center of gravity.
5. How Should an AI Express Itself?
When every AI product can summarize, generate, and analyze, the feelings that they elicit are the only differentiators. While 72% of customer-experience (CX) leaders expect AI agents to be a direct extension of their brand’s identity, most AI products still sound the same. They’re polite, generic, and forgettable. All systems deserve the same design rigor and attention to personality, tone, and emotional expression. Emotion isn’t just a matter of expression; it’s a social signal. Humans read emotional cues in every interaction, and the interactions that AI drives are no exception. Character is who the AI is, behavior is how that identity shows up moment to moment. Teams that conflate the two end up with personalities that feel inconsistent or hollow.
6. How Does AI Earn Users’ Trust?
Today, 72% of consumers trust companies less than they did a year ago. So we must build users’ trust in artificial intelligence through relationship competence, consistency, transparency, and recovery. That last aspect of trust—recovery—is the one that teams most often overlook: AI will fail. The question is whether your design solution earns back users’ trust or loses it for good. An AI that acknowledges a mistake is more trustworthy than one that never explains itself. Relationship design is the framework that answers how—defining how an AI evolves with the user, earns the right to know more over time, and protects what it knows. Teams that embrace relationship design stop designing single interactions and start designing for the long arc of how people and AI can learn to work together.
The AI Capability That Matters Most
Together, these six dimensions of artificial intelligence form a complete map of the human-AI relationship. UX design has always been the practice of making technology work for people. AI is raising the stakes of our work enormously, and UX designers are the ones who are best equipped to champion and master that work.
The organizations that are getting real returns from AI are those that are investing in their organization’s ability to think across all six of these facets of AI. The Boston Consulting Group (BCG) has found that successful AI leaders allocate 70% of their resources to people and processes, not to algorithms. The competitive advantage isn’t the AI model. It’s the team’s ability to design for people.
The next time someone asks you to “Add AI to this”—ask them which of these six questions they’ve answered and what need they’ve identified for which AI is the best answer. If the answer is none, you know where to start.
Acknowledgment—I would like to extend my thanks to Nate Cox, VP of Design, and the Punchcut team for their invaluable research and insights that helped shape this column.
Ken was a co-founder of Punchcut and has driven the company’s vision, strategy, and creative direction for over 20 years—from the company’s inception as the first mobile-design consultancy to its position today as a design accelerator for business growth and transformation. Punchcut works with many of the world’s top companies—including Samsung, LG, Disney, Nissan, and Google—to envision and design transformative product experiences in wearables, smart home Internet of Things (IoT), autonomous vehicles, and extended reality (XR). As a UX leader and entrepreneur, Ken is a passionate advocate for a human-centered approach to design and business. He believes that design is all about shaping human’s relationships with products in ways that create sustainable value for people and businesses. He studied communication design at Kutztown University of Pennsylvania. Read More
