What John Maeda thinks about AI agents
Designer and technologist; author of The Laws of Simplicity; former president of the Rhode Island School of Design; formerly at the MIT Media Lab.
John Maeda did not write this page.
We collected these quotes from things they published elsewhere, and every quote links to where it was said. They have no account here and have not endorsed this site. Quotes are word for word; the short line under each one is our own restatement, not their wording. Their own site. Is this you? Claim it or ask us to remove it. Or tell us what is wrong here.
Filter AI agents · Newest first
Only John Maeda on AI agents. See everyone on AI agents
6 dated positions, 2025 to 2026, in their own words. Our reading of what John Maeda has said — not written or endorsed by them.
-
Their wordsThe work did not disappear. It moved. From doing to judging. So the screen does not vanish in this new world, the way some people like to predict. It simply changes jobs.
-
Their wordsAnd this is why chat, all by itself, is like working with one hand tied behind your back.
- 3 months earlier
-
Their wordsBut AI with intent can teleport the user to the goal. Why would you want to send a user through a obstacle course when they just want to teleport to the goal? That is the central question in the era of agentic experience.
↗Design in Tech Report SXSW 2026: UX to AXyoutube.com 4th of 19 in this recording
-
Our readingThe only thing new about agents is the loop: a model running in a loop is what an agent is.
Their wordsto realize that only thing new is the loop. Everything is in a loop now. And when that when the model runs in a loop, that's what an agent is. A loop means basically autopilot.
↗Design in Tech Report SXSW 2026: UX to AXyoutube.com 5th of 19 in this recording
- 12 months earlier
-
Their wordsit was expensive and there wasn't that many tokens available on Earth so now the tokens are cheaper so agents are running on in loops and they're extremely powerful
↗2025 Design in Tech Report (Unedited): Autodesigners on Autopilotyoutube.com 5th of 19 in this recording
-
Their wordsthe problem with tools like calling an AP API is it's very powerful um when it's read only it's safe but when you can do write actions it gets pretty scary so there's risk and reward when you use tools
↗2025 Design in Tech Report (Unedited): Autodesigners on Autopilotyoutube.com 10th of 19 in this recording