We're living through a shift that most people haven't noticed yet. The bottleneck has moved from what AI can do to how we experience it. And whoever wins the interface wins that space.
TLDR
The Technical Foundation is Already Here
Builders have all the tools they need to build modern interfaces. If you look around, some of the hardest problems have been solved. GPT-4 and Gemini can already see, hear, and speak. Google Lens lets you point your camera at anything and ask questions with your voice. Project Astra showed us what happens when AI can interpret the world through smart glasses in real-time.
Claude can maintain conversations that span days without losing context. Magic.dev ingests entire codebases and helps you debug like a human collaborator who never forgets.
The models got good enough. Now the race is about making them feel inevitable.
Single Products Will Pioneer New Interfaces
I'm going to start with an example. Meet Comet by Perplexity. If you don't know, Comet is an AI-powered web browser that works like a personal assistant, helping you search, summarize, shop, schedule, and automate tasks directly in the browser.
Since launching Comet in early 2025, Perplexity's valuation has exploded from $14 billion to $18 billion - after raising a $100 million round.
Comet represents exactly what I mean by next-generation interfaces. Instead of bolting AI onto an existing browser, they reimagined what browsing could be when intelligence is built in from day one. This is how new interface paradigms emerge-one breakthrough product that makes the old way feel clunky.
Why Interface is a Growing Moat
When I open most apps on my phone, they still feel like they were built for the mobile era, not the AI era. Almost every app interface has no automation, limited voice control, or limited AI capabilities.
This creates a massive opportunity. Just like the web-to-mobile shift reset which apps dominated our lives, I predict AI will do the same thing. The difference is that this time, the interface is the product.
Speed becomes a solid moat. When the underlying technology becomes cheap and accessible, the companies that ship the best experience fastest will win. Amazon's Alexa proves this - it's not the smartest assistant, but it got the voice-first experience right early and captured the market.
Think about something as simple as measuring cups. There are dozens of options on Amazon—some with digital scales and screens, others made of plastic, glass, or wood. But the interface (how you actually use it) determines which one you buy and keep using. Just because a new product is released with a fancy interface, it doesn't mean it will be the best. Like who the hell is going to buy this product just because it's 'new' and 'fancy'?
The same principle applies to AI products. When everyone has access to similar models, the interface becomes the only real differentiator. The measuring cup with the best interface wins your kitchen counter, just like the AI product with the best interface wins your daily workflow.
The Three Pillars of Next-Gen Interfaces
Multimodal by Default
The future interface doesn't ask you to choose between text, voice, or images. Google's AI Mode in Search Labs lets you combine all three seamlessly. You can ask a question, upload a photo, follow up with your voice. It feels obvious once you've used it.
Memory That Actually Works
Interfaces that remember may feel like magic now but will be the standard in the near future. When you can say "And tomorrow?" after asking about today's weather, without repeating context, the experience transforms from robotic to human. Long-context models like Claude are making this the new baseline. Once people adopt this tech, they will not settle for anything less.
Voice as the Primary Interface
This is a hot topic right now. I would like to preface that these are my current thoughts. I fully expect my opinion will evolve over time as we see new products come out and the average person begins to integrate AI into their daily lives.
I believe text and voice will be the dominant interfaces for the next decade. Real-time voice is becoming the default because it's faster than typing and feels more natural than clicking on a keyboard. When we all get robots in our homes, voice will become even more dominant.
These three pillars work together to create interfaces that feel less like using software and more like collaborating with intelligence.
Why now?
Here's what makes this moment different: technical progress shifted the bottleneck from capability to usability. I spent the past 12 months asking "What can AI do?". Now I'm asking "How should AI feel?".
This is fundamentally a design problem, not a research problem.
Design problems favor fast, focused teams over big tech incumbents.
The companies that figure this out first will build better apps but more importantly, they'll define how an entire generation interacts with intelligence.
The Hidden Opportunity
Smart interfaces unlock revenue streams that were impossible before. Synthesia lets anyone create professional video content with AI avatars-no cameras, or actors, or expensive production. Over 60% of Fortune 100 companies already use it for training and marketing.
ElevenLabs started with voice synthesis and built an entire creator economy around it. Their voice library lets people monetize their vocal likeness. Their audiobook platform turns any writer into a publisher. Even music creators can now compose and sell AI-generated tracks cleared for commercial use.
These platforms prove something crucial: the interface the founders choose really does strongly determine the business model. When AI becomes accessible through the right experience, entirely new markets emerge.
What This Means for Builders
Distribution will matter more than ever.
I believe people will begin to fade out the hype around 'AI' as we're seeing it in the Super Bowl ads, youtube videos and it's plastered all over social media. Marketing campaigns will turn away from dropping buzz words like AI and instead focus on the experience their product provides.
When everyone has access to the same models, the interface becomes the only differentiator. Brand, taste, and user experience are the new moats. The companies that win will be those that make AI feel effortless and obvious.
Just like the new wave of indie hackers, the founders that understand the power of interface will be able to fundraise larger rounds and ultimately get acquired at higher valuations. And I think we'll see more getting acquired - check out the future of ai native startups.
What an exciting time to be alive.