The Interface Revolution

I caught myself making a curious choice the other day. When I needed technical help, I opened Claude. When I wanted life advice, I reached for ChatGPT. Same underlying sophistication, but I realized I've been seeing these AI companies' brands completely differently—Claude feels like a technically focused tool, while ChatGPT feels like a general consumer product. Something about their positioning made me choose differently without even thinking about it.

And I'm sure these companies know exactly how they come across. If anything, they seem to be leaning harder into these distinctions—Anthropic doubling down on safety research and technical rigor, OpenAI pushing ChatGPT as the AI for everyone.

We could spend time debating which approach wins long-term, but what's got me more curious lately is the strategic moves these companies are making that go way beyond their current positioning. These aren't just about today's products—they're bets on how we'll interact with AI years from now.

Jony Ive - Sam Altman Partnership

This brings me to something that happened recently which sparked my curiosity. Sam Altman and OpenAI partnered with Jony Ive (previously chief design officer at Apple) through the acquisition of his design collective, LoveFrom. We don't know what they're building yet, but to understand why this matters, let's think about what Jony Ive already accomplished with the iPhone.

The smartphone, and the iPhone specifically, completely transformed how we interact with the internet. Most of our online experience is now mobile-first. The interface of the smartphone has become foundational for the modern digital experiences we take for granted: scrolling through social feeds, swiping through apps, touch-based interactions that feel intuitive rather than technical. There's honestly no telling what the world would have looked like in the last 15 years without this interface innovation.

To me, the Ive-OpenAI partnership signals that Sam Altman thinks the next frontier for AI is not just building better models alone, but that he is thinking about fundamental innovations in the platforms and interfaces through which we will interact with AI as a technology.

And honestly? I think he is absolutely right.

A New De-facto Interface

Since the dawn of widely available LLM-powered tools, the default way most of us use the internet has now evolved from search bars to chatbots (which are really an interactive search bar of sorts that talks back to you). But we're already moving beyond this with the advent of AI agents that can browse the web, use software, and complete complex workflows without us doing the clicking and searching ourselves.

I really think that as AI becomes more capable and usage spreads to pretty much everyone, what will the next innovations in the de-facto techno-human interface look like? Because I think this will fundamentally change how we interact with the internet as we know it, much like the smartphone did.

If AI agents can handle all the execution work, what do we humans spend most of our effort doing?

My optimistic view is that we get to do more high-level thinking, more synthesis, more of what makes us uniquely human. But perhaps AI agents are not frictionless enough. The added complexity of having to configure and orchestrate their interactions is inherently a barrier to widespread use.

What's Next?

It is clear that there is innovation pressure on the evolution of AI interfaces. I am fascinated to see where they go next. Will we interact primarily through self-orchestrating agents, direct neural interfaces, or something we haven't imagined yet?

I honestly don't know. But I think the answers to this question—not just the model capabilities themselves—are going to shape AI innovation going forward.