See e.g. https://
arxiv.org/abs/2212.09993 "Our experiments reveal that while powerful deep models offer reasonable performances on puzzles that they are trained on, they are not better than random accuracy when analyzed for generalization."
ETHICS
-
Deep Models Fail to Generalize on Puzzles Beyond Training Data
By
–
-
GPT models fail at genuine generalization on novel problems
By
–
Exactly. Except with the opposite conclusion. Every study so far that tries to test GPT-N for actual generalization has found that it scores no better than random on genuinely new problems — brand new coding problems in particular. This is why it can't do ARC either.
-
Training Data Alone Doesn’t Ensure Quality AI Performance
By
–
Even a hashtable can pass the bar exam giving enough training data. But you probably don't want to be represented by a hashtable.
-
Skill Acquisition Efficiency as a True Measure of Intelligence
By
–
Skill, on its own, is not a sign of intelligence. But skill acquisition efficiency over arbitrary skillsets is. And that looks very different in the context of human test takers vs. machines.
-
General Intelligence vs Specialized Machine Design in Chess
By
–
And so they could have used these 30 years to become great at *any* other human skill. They possess general intelligence. But you cannot make the same intelligence assumption when you see a machine *designed* to play chess, or one that has memorized one billion chess games.
-
Scientists Successfully Implant Synthetic Embryo in Primate Uterus
By
–
[FLASHER]
Première : des scientifiques implantent un embryon synthétique dans l'utérus d'un singe https://
buff.ly/435W4CJ v/ @techreview Cette prouesse, rendue possible par les cellules souches, ouvre la voie à des animaux créés artificiellement #BioTech
#FlashTweet -
AI Webinar Summary: Key Insights on Current Developments
By
–
Note:
– this is my summary of the webinar and may include word choice errors or oversimplifications
– these are not tweeted in the order they were said
– video can be rewatched on YT if you want to deep dive – I’m not a journalist – this thread does not mean I agree with them -
GPT-4+ Benefits Healthcare Education Without Addressing Trade-offs
By
–
Andrew said GPT-4+ could do a lot for healthcare and education, doesn’t want to slow down the “good” of this research. Both seemed to mention the “good” and “bad” sides but no tradeoffs shared, just that not having the “good” side feels like we’re missing out on something big.
-
AI as solution to social problems: safety and ethics debate
By
–
Andrew: “I don’t see any realistic threat of AI escape” Yann says he’s not involved with Meta policy work but that progress in AI fixes a lot of social media issues, like taking down hate speech Yann says AI is part of the solution, not the problem.
-
AI Transforms Healthcare: The Doctor’s New Digital Partner
By
–
The #AI will see you now. https://
wsj.com/articles/the-a
i-will-see-you-now-5f8fba14?st=5w8ouicou9jocux
… via @WSJ