> [Chomsky] has never said you could not statistically approximate a natural process, in fact he has explicitly said that you probably could Okay, but the arti

> [Chomsky] has never said you could not statistically approximate a natural process, in fact he has explicitly said that you probably could Okay, but the article has Rodney Brooks asserting that > I personally remember very heated arguments on the MIT campus between Chomsky followers and computer scientists working on statistical models in the early 2000’s. The arguments came up in computer science faculty hiring meetings. The Chomskians claimed that because of the need for universal grammar there would never be anything useful that came out of statistical approaches to language. ChatGPT has proved them wrong. Worth noting that in the early 2000s, Rodney Brooks was director of the MIT Computer Science and Artificial Intelligence Laboratory, so... you know, he might have some context on that debate. Although I'll concede he attributes this to Chomsky followers, not to Chomsky.