๐ Last week,
@TelliusData sponsored an incredible event in SF on the Future of AI Interfaces, organized by @deviyer_ and hosting a talk by
@karinanguyen from
@OpenAI .
It got me thinkingโchat wasnโt always the obvious AI interface. GPT existed via API long before ChatGPT took off. And one of the most best GPT interfaces yet? Canvasโinspired by coding IDEs. The next wave of interfaces wonโt be easy to predict, but there will be analogies to the types of interfaces we already have today.
So whatโs next?
๐น Operator โ Orchestrator
AI wonโt just respondโitโll anticipate, coordinate, and execute across systems & teams. Less input, more action.
๐ค Multiplayer AI
AI is mostly 1:1 today. The future? Collaborative AIโthink AI-assisted brainstorming, strategic planning, and problem-solving in real time with teams. Inspired by @deviyer_!
๐ Generative UIs
Interfaces will be fluid & adaptive, reshaping based on workflows, intent & context. Less rigid menus, more seamless experience.
๐ช The Best UI is No UI
Interfaces will fade into the background. AI will just happen, responding proactively without us needing to engage explicitly. Inspired by
@vinodiyengar!
๐ Voice is the Next Big AI Interface
Writing is dense, but voice is effortless & natural. AI interactions will feel like chatting with a colleague on a Zoom call.
And with all the (stupid) panic around DeepSeek โcheaper, faster, smarter models will only accelerate interface innovation. Each leap in model power unlocks entirely new ways to interact with AI. There's so much more to talk about here, I didn't even get to video ๐
@vidythatte @channel42_co