ChatGPT isn’t as smart as people think claims AI expert

Rodney Brooks, an ex-MIT head, revealed his skepticism on how smart AI really is in a brand-new interview.

Rodney Brooks, The ex-head of MIT’s Computer Science and Artificial Intelligence Laboratory has come forward about his skepticism on the true intelligence of AI like ChatGPT in a new interview. Brooks has previously had a hand in the creation of devices like the Roomba.

Interviewed by IEEE Spectrum, Brooks talks at length about how he believes that while impressive, the AI tech behind ChatGPT and other language models isn’t as smart as people think it is.

The main part of his skepticism lies in how the AI models work. Brooks explains that he uses AI to help with “arcane coding” and admits it’s better than a search engine. However, he still finds that it is making mistakes due to having no real intelligence.

“It gives an answer with complete confidence, and I sort of believe it. And half the time, it’s completely wrong.

“And I spend 2 or 3 hours using that hint, and then I say, “That didn’t work,” and it just does this other thing.

“Now, that’s not the same as intelligence. It’s not the same as interacting. It’s looking it up.”

Brooks claims what the current spate of AI chatbots are “good at is saying what an answer should sound like, which is different from what an answer should be.”

Essentially what Brooks is saying is that with no real way to decipher the context of the situation and a general lack of self-awareness, it’s mostly just regurgitating information until it strikes the right cord in your brain.

AI expert isn’t worried about ChatGPT

Brooks has been around for quite some time now and is on his third robotics startup, Robust.AI. The company plans to work on a cart that will work with humans in a safer capacity.

He does think that these AI will be better than things like Watson Jeopardy! attempt, which IBM claimed would help science more than just answering TV quiz show questions. However, his experience and time in the industry have seen “about 30 or 40 events” that would change everything over the “last 50 years”. Brooks says that “many of them have turned out to be utter duds”, even if they’re “useful” – like chess machines.