[YeGoblynQueenne]

So, I understand this blog post is about something else completely (the internet argument started by Yoav Goldberg on Medium, reportedly) but for me the really interesting part is the historical information in it. I wish Fernando Pereira could find the time to expound a bit on all those parenthetical notes in his blog post, perhaps even write a short book on the history of AI.

AI is kind of a strange beast like that: it's gone through a few very different phases and it's difficult for one person to understand all of them equally well. Which of course makes it even harder to avoid reinventing wheels and repeating mistakes. A bit of history would do us all a world of good.

Btw, I'm getting the feeling most people here will probably hear of Fernando Pereira for the fist time but he has a very long career in AI and NLP. He was a prominent symbolicist, with some important contributions to logic programming (he was one of the co-founders of Quintus, the company that sold the first commercial Prolog, along with Warren, Byrd and others). Then he turned to statistical AI and now he's a VP at Google (a.k.a. the den of the connectionists, if I may be so bold). He's probably one of the few computer scientists around who understands both symbolic and statistical AI in equal measures. If anyone is qualified to talk about their relative merits, that's him.

(and if I sound like a bit of a fangirl- that is because I basically am. Pereira is one of my logic programming heroes and a great teacher to me, albeit unbeknownst to him :)