o3: The grand finale of AI in 2024 – by Nathan Lambert
https://www.interconnects.ai/p/openais-o3-the-2024-finale-of-ai
“Jared Friedman of Y Combinator says that while 2024 was the year you could have a natural-sounding phone call with AI, 2025 will be the year you can have a realistic video call with an AI
https://x.com/tsarnick/status/1874554916632727618
“i always wanted to write a six-word story. here it is: ___ near the singularity; unclear which side.”
https://x.com/sama/status/1875603249472139576
The Next Great Leap in AI Is Behind Schedule and Crazy Expensive
https://www.msn.com/en-us/money/other/the-next-great-leap-in-ai-is-behind-schedule-and-crazy-expensive/ar-AA1wfMCB
“This tweet is exactly what you would expect to see in a world where AI capabilities are growing: last years very hard math tests built to challenge AI are already getting solved, we need much harder tests. Feels like the background news story in the first scene of a scifi drama.” / X
https://x.com/emollick/status/1875186092141666419
“@doomslide @chaitinsgoose I do not acknowledge this “disagree” because I agree: there is. What we might really be disagreeing about is who’s being pedantic. I think Sonnet 3.5 is meaningfully a proto-AGI, ie a big boy AGI can be derived through some relatively trivial (if costly) improvements on Sonnet” / X
https://x.com/teortaxesTex/status/1876353300217598046
“This bit of Sam Altman’s newest post is similar in tone to a post by the CEO of Anthropic & what many (not all) researchers from every lab have been saying publicly and privately. You do not have to believe them, but I think they believe what they are saying, for what it worth.
https://x.com/emollick/status/1876119097660100636
“We’re going to be releasing ARC-AGI-2 towards late February. It’s been a long time coming — I first announced it in early 2022. Beyond that, we’re starting work on a next-generation AGI benchmark that completely departs from the 2019 ARC-AGI format. We’re very excited about it!” / X
https://x.com/fchollet/status/1877070012508180886
“very few people talking about the dark side of AGI” / X
https://x.com/vikhyatk/status/1877155735114563996
“A sequel to banger “what Wenfeng Liang wants” (can’t find due to X inability to do search on strings > 280 characters): What a Chinese AGI lab founder/ML researcher really looks like What DC China hawks think a Chinese AGI lab founder looks like
https://x.com/teortaxesTex/status/1877403009795182674
“AGI won’t be announced. You will hear of it by word of mouth – from friends and family. You will be convinced of it by personal experience as you give it increasingly difficult tasks 1 after the other and it solves each one. You will then stop and realize you are no longer just” / X
https://x.com/nrehiew_/status/1876088957404127592
