Sam Altman got a lot of flak from ML folks for that, despite this having been OpenAI’s mission from the start. I believe that he really intends to do it, yet at the moment, most of the OpenAI folks seem to still think that most progress can be made in ML.
Funny quote below, but who in the ML community called BS on Sam Altman when he said re: OpenAI in July that "he and his team of researchers hope to build artificial general intelligence, or A.G.I., a machine that can do anything the human brain can do"? https://www.nytimes.com/2019/07/22/technology/open-ai-microsoft.html …https://twitter.com/Plinz/status/1194403722563993600 …
-
-
-
i am just trying to show that you some pretty visible, well funded teams with lots of employees certainly claim that they are pursuing AGI; DeepMind has also often portrayed itself as trying to solve intelligence. you made it sound like I was making it up.
- 3 more replies
New conversation -
-
-
I get why this is annoying/silly to see stuff like this, but I (and I think many others) have a rule of thumb: never determine a researcher's opinions based on what they say to the popular press. Only use what they actually put their name on in scientific conferences/journals.
-
i gave pointers to well cited articles on transformer nets in the scientific literature (arxiv, by google, on BERT).
- 1 more reply
New conversation -
-
-
If all he said was they hope...
- 1 more reply
New conversation -
-
-
I mean, it feels weird to call BS on people's hopes. Now, if he claimed to have *achieved* AGI, if be right there with you.
-
Yeah
@GaryMarcus, tbh most AI researchers (as far as I can tell) just roll their eyes at Altman etc. and keep working. As I've pointed out before, most focus on research and not Sci comm to public, so they don't spend their days explaining their work is not AGI, they just do it.
End of conversation
New conversation -
-
-
ML twitter was filled with tweets mocking the AGI hype around the announcement. Some examples from people I follow: https://twitter.com/egrefen/status/1153398849953775617?s=20 … https://twitter.com/ylecun/status/1153515608379547648?s=20 … https://twitter.com/fchollet/status/1153517318825111554?s=20 … https://twitter.com/rupspace/status/1153534253025525760?s=20 …https://twitter.com/Smerity/status/1153364705777311745?s=20 …
-
FWIW - googling: "general intelligence" site:https://papers.nips.cc/paper/ gives 61 results from the last couple of years of NeurIPS. Most appear to be citations of journals/conferences with that phase in the name. "function approximation" gets 585. There's a big disconnect between...
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.