Design Is the Future of AI
Why OpenAI’s $6.5B bet on Jony Ive signals that design is the next great battleground for AI.
Exactly 20 years ago, Steve Jobs gave a commencement address at Stanford that rewired how a generation thought about technology. I’ll never forget the first time I watched this speech. It completely changed the way I approached my career, my life, and what kind of impact I wanted to make.
He told a story about taking a calligraphy class—seemingly useless at the time—that ended up shaping the design of the original Macintosh. That same attention to detail, to typography, to how things feel, arguably kicked off the age of human-centered computing.
Jobs wasn’t celebrating design as decoration. He was reminding us that great products come from caring about every detail of how people experience technology. That lesson feels more urgent today than ever.
Because if AI is to become the next universal interface—the way we interact with knowledge, tools, creativity, and even ourselves—then it won’t be the size of the model that matters.
It will be the design.
The Age of Models Is Ending. The Age of Interfaces Is Just Beginning.
Artificial intelligence is at an inflection point. For the last few years, breakthroughs in large language models have dominated the headlines. Model performance leapt forward. Parameters scaled. Benchmarks were crushed. And yet quietly—and increasingly loudly—the conversation is shifting. Not to the next algorithmic leap, but to something far more human: design.
We’re entering a new era. One that doesn’t ask, “What can this model do?” but instead, “How do people actually use this?” It’s a shift from capabilities to experience. From novelty to everyday utility.
And it may define who wins the next decade of AI.
The OpenAI–Jony Ive Bet
OpenAI’s $6.5 billion acquisition of io, the device startup that CEO Sam Altman and famed Apple designer Jony Ive have quietly been working on for two years is more than just a headline—it’s a thesis.
With Ive’s design firm, LoveFrom, now heading up creative and design work, Altman is assembling a vertically integrated stack: frontier models, consumer software, custom silicon, and now, physical hardware.
The goal isn’t just better performance. It’s distribution. Trust. Ubiquity. And none of those things are possible without design.
Jony Ive helped define the look and feel of personal computing. From the iMac to the iPhone, his design ethos made powerful technology intuitive, joyful—even personal.
That’s what OpenAI is betting on: that AI’s next chapter won’t be about prompts in a browser, but about cognitive hardware. A new category of object built from scratch to integrate seamlessly into your life, not your taskbar.
Design as Distribution
Model performance used to be the differentiator. Now, it’s commoditized. GPT-4o, Claude Opus, Gemini—they’re all impressive. And largely interchangeable.
What separates winners going forward will be the same thing that separated the iPhone from every other smartphone: design, distribution, and ecosystem.
OpenAI’s advantage isn’t just its models. It’s the mindshare it captured through a simple, well-designed chat interface. It’s the trust it built by making AI feel helpful, not intimidating. It’s the shift from a technology you have to learn, to a companion that learns you.
And that’s the design principle that matters most now: contextual awareness. AI that listens. Remembers. Anticipates. Not a search engine replacement, but a thought partner. A layer that blends into your workflow—whether that’s drafting a document, researching a trip, or thinking through a product strategy.
From GUI to Multi-Modal
Every generation has had its interface moment:
The graphical user interface (GUI) turned personal computers from machines into companions.
Multi-touch screens turned smartphones into extensions of our bodies.
Now, the question is: what’s the native interface for AI?
So far, we’ve settled for the lowest common denominator: prompts typed into a box. But that’s not how intelligence works in the real world. Real intelligence listens, observes, remembers, adapts. It’s less like a command line, more like a conversation. Or a presence.
That’s what Jony Ive and Sam Altman seem to be building: not just AI, but technology that removes friction. Devices that aren’t just tools—but companions. Not something you boot up, but something that’s already with you.
Cognitive Hardware: The Battle for the Endpoint
OpenAI’s acquisition isn’t your traditional roll-up acquihire. It’s a vertical integration play. Altman is assembling a stack: models, chips, applications, and now, hardware. It’s the same blueprint Apple used to redefine personal computing. Google is doing the same with Gemini and Pixel. Microsoft’s partnership with Anthropic for GitHub Copilot signals it’s willing to go multi-model, provided the software layer stays under their control.
What’s at stake? The endpoint—the device, interface, or object that mediates the user’s relationship with AI.
Whoever owns the endpoint owns the experience. They set the defaults. They define the feedback loops. They decide what’s tracked, remembered, and suggested. Think about what Safari is to Apple, what Google Search is to Chrome, or what Alexa was supposed to be for Amazon.
Right now, AI lives in the cloud, accessed through a browser, app, or API. That works—but it’s clunky. If you want to ask ChatGPT a follow-up question about a conversation from last week, you have to dig through chat logs, re-explain yourself, and hope it remembers the right context. It’s not intelligent. It’s patient.
What Altman and Ive are betting on is a fundamentally different interaction paradigm—one where AI is not something you open, but something that’s already with you. A device that knows your context, understands your preferences, and dissolves friction the way the iPhone dissolved the line between desktop and mobile computing.
We’re talking about cognitive hardware—ambient, persistent, and aware. A companion, not a tool. It’s not just form factor innovation. It’s an attempt to redefine what it means to “use” AI in the first place.
Design becomes the tip of the spear. Not just industrial design, but product architecture, UX flows, and behavioral integration. The key question isn’t, “Can we build this model?” but, “Will someone use it in the middle of their real-life messiness—while commuting, cooking, working, thinking?”
Lessons from Apple, Redux
AI will not be defined by the largest model or fastest inference time. It will be defined by who makes it useful to the most people, most intuitively.
That’s not a research problem. That’s a design problem.
Remember: Apple didn’t win because it had better specs. It won because it cared about the seams—the transitions, the animations, the feel and look of the device. It won because it made technology personal.
OpenAI’s move echoes that. The ChatGPT app was a breakthrough not because of its raw power, but because it made AI feel useful, friendly, and intuitive. But there’s still steps you have to go through—laptop, browser, login, create a prompt.
The future? Something ambient. Something new. Something designed from scratch for human-technology interaction. That’s what Ive’s specialized in and proven his ability to do time and time again over his long career.
What This Means for Product Leaders
Design is no longer a nice-to-have in AI. It is the differentiator. The market is no longer constrained by model performance—everyone has access to excellent LLMs. What matters now is who can wrap those capabilities in a product that feels inevitable.
So here’s the brief for product leaders:
If you're still shipping AI features inside modals and calling it innovation, you're already behind.
Treat design not as polish, but as the substrate of your product strategy.
Assume the interface is going to collapse. AI-native UX isn’t a prompt box—it’s adaptive, contextual, and increasingly invisible.
The future isn’t one model winning. It’s whoever gets distribution through design, builds trust through interaction, and owns the endpoint where cognition happens. That might be your app. Or it might be someone else's device.
If you’re building in AI, the question to obsess over is no longer “What can it do?” It’s: “What does it feel like to use every day?”
Because the AI that wins won’t be the smartest. It’ll be the one people never think twice about using.