Will Sam Altman always win the OpenAI board fight in an AI agent simulation? | VentureBeat
“A new study showed ChatGPT achieved 90% accuracy in medical diagnosis, outperforming both human doctors (74%) and doctors using ChatGPT (76%) So much progress to be made for AI and healthcare. Really cool to already start seeing these results already
“Wait – is the new GPT-4o a smaller and less intelligent model? We have completed running our independent evals on OpenAI’s GPT-4o release yesterday and are consistently measuring materially lower eval scores than the August release of GPT-4o. GPT-4o (Nov) vs GPT-4o (Aug): ➤
“ChatGPT as universal translator 👇 I went through Seoul and Tokyo a few weeks ago doing the same thing. Last weekend in Las Vegas, I had an Arabic/English conversation about AI with a cabbie who was previously a dentist in Egypt. Voice mode isn’t perfect yet, but it’s still” / X
“One blindspot for AI reasoning engines like o1 is that they all appear to be trained on very traditional deductive problem solving. What would a model trained on induction or abduction do? What about one trained on free association? Expert heuristics? Randomized exquisite corpse?
Prompt Injecting Your Way To Shell: OpenAI’s Containerized ChatGPT Environment | 0din.ai
OpenAI releases new o1 reasoning model – The Verge
OpenAI blamed NYT for tech problem erasing evidence of copyright abuse – Ars Technica
Advancing red teaming with people and AI | OpenAI
Inside Elon Musk’s Quest to Beat OpenAI at Its Own Game – WSJ
“The irony of AI is we expected it to be awesome at math & be all cool logic (“your idea of love does not compute!”) Instead, AI is bad at math, just wants to write poems, & is all hot, weird simulated emotion. For example, if you make GPT-3.5 “anxious,” it changes its behavior!
OpenAI’s Sora video generator appears to have leaked | TechCrunch
OpenAI accused of trying to profit off AI model inspection in court – Ars Technica
OpenAI is funding research into ‘AI morality’ | TechCrunch
OpenAI moves to trademark its o1 ‘reasoning’ models | TechCrunch
OpenAI stops Sora video model access after artists leak in protest – The Washington Post
“I only saw one leaked Sora video that had a whole prompt and which was done in one shot (from the appropriately named @SirMrMeowmeow). Unfortunately could only download low resolution of it from X. I gave the same prompt to Runway & Kling, one shot. They struggled with mouths.
“The leaked Sora videos are well beyond Runway and even the major Chinese models like Kling (I have made hundreds of AI videos, so have some experience) – they aren’t perfect but very impressive. It is an order of magnitude better if these are representative of first attempts.” / X
OpenAI faces critical test as Chinese models close the gap in AI leadership | VentureBeat
Thomson Reuters’ CoCounsel redefines legal AI with OpenAI’s o1-mini model | VentureBeat
“OpenAI’s Sora video model appears to have leaked. And it’s even more advanced than the February demo. A group from the early beta testers just dropped access to what looks to be Sora’s ‘turbo’ variant on Hugging Face, citing concerns about OpenAI’s early access program and
OpenAI to present plans for U.S. AI strategy and an alliance to compete with China
Leave a Reply