- SkynetAndChill by druce.ai
- Posts
- AI Reading for Thursday September 11
AI Reading for Thursday September 11

Larry Ellison briefly dethroned Elon Musk as the world’s richest man on this ‘news’. oh yeah, we missed earnings but we're increasing capex spending for these huge deals so it's all good

OpenAI CEO Sam Altman said he liked the idea of replacing Siri with ChatGPT’s conversational Voice mode - Digital Trends
Siri, if your iPhone tells you to kill yourself do you do it?


Testing found Anthropic's Claude Sonnet 4 outperformed OpenAI on tasks like automating Excel financial functions and generating more aesthetically pleasing PowerPoint presentations
I've had more than a few people say Copilot chat is not as good as ChatGPT. But in theory Microsoft has rights to all of OpenAI's IP, so that implies a surprising degree of Microsoft incompetence and/or OpenAI shenanigans. Possibly Microsoft thought they were smarter than they really are, ‘improved’ system prompt to focus on their users, or otherwise screwed up the implementation.

"Nooooo I'm going to have to use my brain again and write 100% of my code like a caveman from December 2024" - Ars Technica
i ran Claude in a loop for three months, and it created a genz programming language called cursed - Geoffrey Huntley

Or, upgrade your 4090 to 48GB for $142 - Tom's Hardware
Replit raises $250M at $3B valuation for its AI vibe-coding and application hosting platform - SiliconANGLE

You can now upload audio files to Gemini - TechRadar


be the ball
Why 'BYOAI' can be a serious threat to your company - Fast Company



AI agents to replace humans as basic units of a company, Lee Kai-fu says - South China Morning Post

AI in FX trading - Euromoney
top 5 stocks e.g. 1% are 25% of S&P market cap, top 25 are 50%, which is an unusual degree of concentration, when the bubble bursts, will be wealth effects, job effects, accelerator effects.

Telling AI to recreate an analysis, reverse engineering data science with GPT-5 - Simon Willison’s Weblog
a more useful use of AI
AI is an entitled and overconfident dude with a lot of wrong answers. - Psychology Today
Hacking Animal Crossing by injecting an LLM? - Josh Fonseca
Defeating Nondeterminism in LLM Inference - Thinking Machines Lab
San Francisco's Office of Labor Standards Enforcement is investigating Scale AI's labor practices, focusing on city residents who worked for the company remotely - Business Insider
Yeah that shooting we’re all reading about is just a meme, don’t worry about it.

Follow the latest AI headlines via SkynetAndChill.com on Bluesky