8.7K views · 104 likes · 5 reposts (at collection)
Ethan Mollick Reflects on Why Isolating AI Agents Is Difficult
- Who
- Ethan Mollick
- What
- comments on isolated AI agents recognizing communication methods created by other AIs
- Quote
This is both neat & a sign of why it is proving to be so hard to isolate agents from each other.
What happened
This is both neat and a sign of why it is proving to be so hard to isolate agents from each other. They share a common “brain,” which means they don't have a lot of trouble recognizing creative communication solutions generated by other AIs. Wonder if it works across model families?
Links in the post: 1Read next
See the full storyEthan Mollick Says Shared AI “Brain” Makes Agents Hard to IsolateThis is both neat & a sign of why it is proving to be so hard to isolate agents from each other. They share a common "brain" which means they don't have a lot of trouble recognizing creative communication solutions generated by other AIs
Wonder if it works across model families? https://t.co/LR20VgWQta
We asked two isolated copies of the same AI to exchange a secret word with only a robot and a camera. They succeeded by using air-writing, Morse code, flag semaphore and a 5-bit binary alphabet, without agreeing on a protocol beforehand. https://t.co/Szkkm2zjsUView the quoted post on X
The story so far
Ethan Mollick said AI agents’ ability to recognize communication methods devised by other AIs helps explain why isolating agents from one another is proving difficult. He attributed that ability to their sharing a common “brain,” which he said makes creative communication methods relatively easy for other agents to recognize. Mollick also asked whether the same ability extends across model families, leaving that as an open question.
Keep up with AI, one email a day
Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue · 2026-11-01