AGI - AI Dynamics

LeCun and Schmidhuber debate current AI approaches

By

–

14 March 2023 17h18

Yann LeCun's epic interview with ZDNet from the fall has been updated with replies by Jürgen Schmidhuber. Meta's AI guru LeCun: Most of today's AI approaches will never lead to true intelligence https://
zdnet.com/article/metas-
ai-guru-lecun-most-of-todays-ai-approaches-will-never-lead-to-true-intelligence/
… @ylecun @SchmidhuberAI #AI #ArtificialIntelligence

→ View original post on X — @tiernanraytech

14 March 2023

Government AI Action Addresses Concerns for Millions

By

@bobgourley

–

14 March 2023 13h34

I was seriously nervous, not for me but for millions of others, and deeply appreciate the hard work you and many others did to help the government take action.

→ View original post on X — @bobgourley

14 March 2023

AGI Meaningless Without Basic Physical Fitness

By

@alexalbert__

–

14 March 2023 2h20

plz do what good is AGI if people can't even bench their bodyweight

→ View original post on X — @alexalbert__

14 March 2023

ChatGPT hype and AGI speculation: corporate responsibility needed

By

@willknight

–

13 March 2023 20h10

Seems like hype around ChatGPT is distorting things (wild speculation about what it will do and even talk of AGI). Microsoft and OpenAI could do more to keep that in check.

→ View original post on X — @willknight

13 March 2023

Ten Predictions for Generative AI Future Trends

By

@ninadschick

–

13 March 2023 18h04

https://
ninaschick.substack.com/p/10-predictio
ns-for-generative-ai
…

→ View original post on X — @ninadschick

13 March 2023

How AI Models Conform to Leading Questions in Responses

By

@willknight

–

13 March 2023 14h29

I don't think you're being misleading at all. What I find interesting though, is how much the model seems to want to find answers that fit with leading parts of the question. If you'd asked "what role microsoft didn't play", presumably it would've run with that.

→ View original post on X — @willknight

13 March 2023

LLMs Override Semantic Priors Through In-Context Learning at Scale

By

@dair_ai

–

12 March 2023 15h08

5/ LLMs do In-Context Learning Differently – shows that with scale, LLMs can override semantic priors when presented with enough flipped labels; these models can also perform well when replacing targets with semantically-unrelated targets.

→ View original post on X — @dair_ai

12 March 2023

Old man and scientist discuss alignment of leaked pre-trained model

By

@goodside

–

11 March 2023 8h10

[In a world… fully aligned…]
TEENAGER: “Prompt engineer?” What’s that?
OLD MAN: Just words, kid. Just words. […one final task… remains…]
SCIENTIST: The leaked model is pre-trained! There’s no RLHF! Not even instruct!
OLD MAN: *Dons e/acc shades* Let’s think step by step.

→ View original post on X — @goodside

11 March 2023