AI Reading for Thursday March 13
OpenAI calls for US government to codify 'fair use' for AI training. - TechCrunch
seems to be missing the part about what the creators get in return, whether they get any say
OpenAI asks White House for relief from a hodgepodge of state AI rules - Bloomberg
Why Sam Altman is smiling: ChatGPT has suddenly kicked into a higher gear - Business Insider
On benchmarks, others have caught up but the overall product is more functional and fully developed. Agents, deep research, fast multimodal voice assistants, things going to go super mainstream fast.
Adobe can’t monetize AI for creatives if there are no creatives left - SiliconANGLE
Ushering In a New Era of Business Process Innovation with AI and ML - ACM
Google introduces AI models for robots - Google DeepMind
Google DeepMind’s new AI models help robots perform zero-shot physical tasks, without training - The Verge
Should AI get trained to 'quiet quit' unpleasant chats? - Ars Technica
Anthropic's MCP standard lets any AI talk to any API - Hacker Noon
If it can compile C, then it can run anything, make a linux kernel running chatgpt.
Station F is Paris's answer to YCombinator - FT
Snapchat adds AI filters that no one asked for - The Verge
Cursor is cursed AF - Ars Technica
Using freedom of information laws to get chat records reveals UK tech secretary uses ChatGPT for policy advice - New Scientist
AI can save HR days of time in onboarding. - Business Insider
Nvidia won the AI training race, but inference is still anyone's game - The Register
Amodei says Chinese spies are targeting AI secrets - TechCrunch
Home improvement stores launch AI for DIY - Marketplace
Grok seeking truth, spitting facts: Dems perform better on the economy. - HuffPost
(some spiky bits to get shaved off in the next fine-tuning)
I got a great idea for an AI create-a-SaaS SaaS - Reddit
Follow the latest AI headlines via SkynetAndChill.com on Bluesky