New OpenAI Model 'Imminent' and AI Stakes Get Raised (plus Med Gemini, GPT 2 Chatbot and Scale AI)

Описание к видео New OpenAI Model 'Imminent' and AI Stakes Get Raised (plus Med Gemini, GPT 2 Chatbot and Scale AI)

Altman ‘knows the release date’, Politico calls it ‘imminent’ according to Insiders, and then the mystery GPT-2 chatbot [made by the phi team at Microsoft] causes mass confusion and hysteria. I break it all down and cover two papers – MedGemini and Scale AI Contamination – released in the last 24 hours. I’ve read them in full and they might be more important than all the rest. Let’s hope life wins over death in the deployment of AI.

AI Insiders:   / aiexplained  

Politico Article: https://www.politico.eu/article/rishi...
Sam Altman Talk:    • The Possibilities of AI [Entire Talk]...  
MIT Interview: https://www.technologyreview.com/2024...
Logan Kilpatrick Tweet:   / 1785834464804794820  
Bubeck Response:   / 1785888787484291440  
GPT2:   / 1785107943664566556  
Where it used to be hosted: https://arena.lmsys.org/
Unicorns?;   / 1784969111430103494  
No Unicorns:   / 1785159370512421201  
GPT2 chatbot logic fail:   / 1785367736157175859  
And language fails:   / 1785101624475537813  
James Betker Blog: https://nonint.com/2023/06/10/the-it-...
Scale AI Benchmark Paper: https://arxiv.org/pdf/2405.00332
Dwarkesh Zuckerberg Interview:    • Mark Zuckerberg - Llama 3, $10B Model...  
Lavander Misuse: https://www.972mag.com/lavender-ai-is...
Autonomous Tank: https://www.techspot.com/news/102769-...
Claude 3 GPQA: https://www.anthropic.com/news/claude...
Med Gemini: https://arxiv.org/pdf/2404.18416
Medical Mistakes: https://www.cnbc.com/2018/02/22/medic...
MedPrompt Microsoft: https://www.microsoft.com/en-us/resea...
My Benchmark Flaws Tweet:   / 1782716249639670000  
My Stargate Video:    • Why Does OpenAI Need a 'Stargate' Sup...  
My GPT-5 Video:    • GPT-5: Everything You Need to Know So...  


Non-hype Newsletter: https://signaltonoise.beehiiv.com/

AI Insiders:   / aiexplained  

Комментарии

Информация по комментариям в разработке