A rainy day at the beach. The letters “OpenAI” are scrawled in the sand. –ar 5:3 –style raw
OpenAI: “All ChatGPT Free users can now use browse, vision, data analysis, file uploads, and GPTs.”
https://x.com/OpenAI/status/1795900306490044479
Sam Altman Vows to Give Away Majority of Wealth – Business Insider
OpenAI for Education | OpenAI
Continued from last week: Scarlett Johansson told OpenAI not to use her voice — and she’s not happy they might have anyway – The Verge
Vox Media and The Atlantic sign content deals with OpenAI – The Verge
The Newsroom AI Catalyst: a global program with WAN-IFRA | OpenAI
“The Atlantic, Vox Media ink licensing, product deals with OpenAI
Vox Media and OpenAI Form Strategic Content and Product Partnership – Vox Media
News Corp and OpenAI Sign Landmark Multi-Year Global Partnership | News Corporation
Enhancing news in ChatGPT with The Atlantic | OpenAI
OpenAI signs 100K PwC workers to ChatGPT’s enterprise tier as PwC becomes its first resale partner | TechCrunch
Report: Apple signs deal with OpenAI for iOS, still wants Google as an ‘option’
“We used AI to bet on the Kentucky Derby, and won. My friend and I tested ChatGPT’s derby picks by asking it to run 100 race simulations. We then opened a @DKSportsbook account, backed the AI’s top picks, Sierra Leone (won most simulations), and the best value bet, Mystic Dan
“Providing widespread access to cutting-edge AI features via ChatGPT Free tier:” / X
Reproducing GPT-2 (124M) in llm.c in 90 minutes for $20 · karpathy/llm.c · Discussion #481 · GitHub
Reproducing GPT-2 in llm.c | Hacker News
What GPT-4o illustrates about AI Regulation
“GPT 4o is 100x overhyped compared to Claude Opus 4o is a good model, but it tends to make several mistakes. In addition, It considerably slows down when OAI has a lot of traffic Opus, on the other hand, has consistently been performing very well and doesn’t quite get its due
“LLMs “intelligence” is hard to benchmark, as we don’t have good benchmarks for human performance at complex tasks. Take theory-of-mind: several tests found GPT-4 beats humans, but another one finds a huge gap. Is it the testing structure? Prompting? Which is right? Hard to know.
“1/ Today is the 4th anniversary of the original GPT-3 paper—”Language Models are Few-Shot Learners” Some reflections on how the last 4 years have played out, and thoughts about the next 4 years” / X
“Its the fourth anniversary of GPT-3, which, well before ChatGPT, showed that LLMs were more capable and stranger than expected. @gwern was, I think, the first person outside an AI lab to get a sense of what was happening, it is worth revisiting the post.
“Apparently today is the 4th year anniversary of GPT-3!
“AI versus 100,000 humans in creativity in this careful study using the Divergent Association Test (a well-validated measure, but all measures of creativity have flaws) GPT-4 wins. Better prompting can further improve performance & diversity of ideas.
“Stunning display of incompetence. Misunderstand Technology & GPT3.5 API was released in March 2022 & it was wrapped into ChatGPT in Nov 2022 & It went viral > the audacity “We learned about ChatGPT on Twitter” Misunderstand Board Roles & “Sam claimed to be an independent
“I was able to get GPT-4o to highlight all clickable on-screen elements and make it take control of the mouse. It can navigate the computer autonomously using the button coordinates.
“Growing evidence that GPT-4 class LLMs give good business advice if done carefully: 1) Getting AI advice boosts profits of high performing Kenyan entrepreneurs by 18% 2) Our paper shows that AI can help consulting work at a high level 3) AI and strategy professors agree on advice
“We interviewed the most prolific jailbreaker of @ChatGPTapp and other leading LLMs, @elder_plinius . READ ⛓️💥👇
“Dang, ChatGPT desktop app w/voice will probably be the first place I try to get tech support as I work now. It just walked me through copying announcements from one course to another in Canvas.” / X
Paul Graham claims Sam Altman wasn’t fired from Y Combinator | TechCrunch
OpenAI Training New Model?
Jeremiah Owyang on X: “OpenAI to launch a model to replace GPT-4
OpenAI training its next major AI model, forms new safety committee | Ars Technica
OpenAI Robotics
OpenAI Is Rebooting Its Robotics Team
OpenAI disbanded its robotics team in 2020, citing lack of training data. Now, with an AI robotics boom taking shape, it’s reviving the team — though with a different mission than last time — per a job listing and sources. With @SarahNEmerson:
The OpenAI Executive and Board-Level Drama Is Relentless, Dramatic, Emotional, and Immature (all of ‘em)
OpenAI sends internal memo releasing former employees from NDAs
Disclosure, I think this is a bit suspect. I am a huge fan of Bilawal, but I’m not sure I’m down with Helen Toner (nor Sam Altman). It feels dramatic for drama’s sake. GPT was out in OpenAI’s world for 18 months (with chat overlays internally) before it was public. To frame it like ChatGPT was a total surprise seems contextually a bit dramatic.
Bilawal Sidhu on X: “❗EXCLUSIVE: “We learned about ChatGPT on Twitter.” What REALLY happened at OpenAI? Former board member Helen Toner breaks her silence with shocking new details about Sam Altman’s firing. Hear the exclusive, untold story on The TED AI Show. Here’s just a sneak peek:
“Stunning display of incompetence. Misunderstand Technology & GPT3.5 API was released in March 2022 & it was wrapped into ChatGPT in Nov 2022 & It went viral > the audacity “We learned about ChatGPT on Twitter” Misunderstand Board Roles & “Sam claimed to be an independent
Ex-OpenAI board member reveals what led to Sam Altman’s brief ousting
OpenAI board members respond to a warning by former members
“Ex-OpenAI board member Helen Toner finally spoke about why the board fired Sam Altman. Plus, big news from OpenAI’s next model, SignLLM, Canva, Copilot for Telegram, and xAI. Here’s everything going on in AI right now:” / X
“The OpenAI drama just doesn’t stop. Helen Toner, ex-OpenAI board member spoke out on why they fired Sam Altman in November: Sam Altman was “in some cases outright lying to the board,” and the board found out about the launch of ChatGPT through Twitter
AI firms mustn’t govern themselves, say ex-members of OpenAI’s board
OpenAI Safety Dialog Continues
So did they dismantle it or did they form it? What is going on people?
OpenAI Forms Safety Committee As It Starts Training Latest Artificial Intelligence Model – TV News Check https://tvnewscheck.com/ai/article/openai-forms-safety-committee-as-it-starts-training-latest-artificial-intelligence-model/
OpenAI Board Forms Safety and Security Committee | OpenAI
Disrupting deceptive uses of AI by covert influence operations | OpenAI
OpenAI finds Russian, Chinese propaganda campaigns used its tech – The Washington Posthttps://archive.md/tW97i
Leave a Reply