4 ms·

Why is this considered the "prehistory" of AI? It's the actual history of AI. . .

bbor 3y ago
Heh I came in for the same point but decided to give the author the benefit of the attention-grabbing-headline, so to speak. AI started in earnest in 1950 with the rest of CS IMO, so technically this is “extremely early AI history”, but you can see how they were basically just positing stuff at this point. I mean “we didn’t use the word planet back then” really threw me for a loop. Also these days “prehistory of AI” means “pre-2010”, according to the LLM industry!
- lsh123 3y ago
  AI history starts from Newton gradient descent ;)
  - uberdru 3y ago
    More like Plato. But cybernetics and systems theory seem more apt models, and, most interestingly, they derive as much from anthropology as from math. . . .
- uberdru 3y ago
  I don't know. It likely goes much further back. The main problem is that there is no commonly accepted definition of "intelligence". So to add the modifier "artificial" (an equally fraught term) is just to muddle the topic. "Artificial Intelligence" is in the end just a metaphor, one that folds quickly under scrutiny.
- ghaff 3y ago
  The field is usually dated to a 1956 workshop at Dartmouth College. https://en.wikipedia.org/wiki/Dartmouth_workshop https://en.wikipedia.org/wiki/Dartmouth_workshop Obviously it was building on earlier work.
  - bbor 3y ago
    Great point! I cite Turing so much that I feel like I have to start it with “Can Machines Think?”, but that’s definitely the more historically valid answer. It certainly is if you ask Dartmouth lol, they have a whole page for it. On the topic of “incredible women historians of AI”, this article by Grace Solomonoff was posted here a while back and blew my mind. Would highly recommend for anyone interested in the minds that started this whole kerfuffle. https://spectrum.ieee.org/dartmouth-ai-workshop https://spectrum.ieee.org/dartmouth-ai-workshop
    - ghaff 3y ago
      Mostly coincidentally--although there were one or two overlapping participants--Cognitive Science is usually dated to an MIT conference a few months later. (Although I don't think it was called that yet.)

bbor 3y ago

Heh I came in for the same point but decided to give the author the benefit of the attention-grabbing-headline, so to speak. AI started in earnest in 1950 with the rest of CS IMO, so technically this is “extremely early AI history”, but you can see how they were basically just positing stuff at this point. I mean “we didn’t use the word planet back then” really threw me for a loop. Also these days “prehistory of AI” means “pre-2010”, according to the LLM industry!

lsh123 3y ago
AI history starts from Newton gradient descent ;)
- uberdru 3y ago
  More like Plato. But cybernetics and systems theory seem more apt models, and, most interestingly, they derive as much from anthropology as from math. . . .
uberdru 3y ago
I don't know. It likely goes much further back. The main problem is that there is no commonly accepted definition of "intelligence". So to add the modifier "artificial" (an equally fraught term) is just to muddle the topic. "Artificial Intelligence" is in the end just a metaphor, one that folds quickly under scrutiny.
ghaff 3y ago
The field is usually dated to a 1956 workshop at Dartmouth College. https://en.wikipedia.org/wiki/Dartmouth_workshop https://en.wikipedia.org/wiki/Dartmouth_workshop Obviously it was building on earlier work.
- bbor 3y ago
  Great point! I cite Turing so much that I feel like I have to start it with “Can Machines Think?”, but that’s definitely the more historically valid answer. It certainly is if you ask Dartmouth lol, they have a whole page for it. On the topic of “incredible women historians of AI”, this article by Grace Solomonoff was posted here a while back and blew my mind. Would highly recommend for anyone interested in the minds that started this whole kerfuffle. https://spectrum.ieee.org/dartmouth-ai-workshop https://spectrum.ieee.org/dartmouth-ai-workshop
  - ghaff 3y ago
    Mostly coincidentally--although there were one or two overlapping participants--Cognitive Science is usually dated to an MIT conference a few months later. (Although I don't think it was called that yet.)