The new science of “emergent misalignment”
(news.ycombinator.com)
5341.
5342.
5343.
OpenAI relaxes GPT-5 rate limit, promises to improve the personality
(bleepingcomputer.com)
5344.
Bluesky: Updated Terms and Policies
(news.ycombinator.com)
5345.
All Souls exam questions and the limits of machine reasoning
(news.ycombinator.com)
5346.
5347.
Google releases pint-size Gemma open AI model
(arstechnica.com)
5348.
5349.
Gemma 3 270M: Compact model for hyper-efficient AI
(news.ycombinator.com)
5350.
5351.
Why LLMs can't really build software
(news.ycombinator.com)
5352.
5353.
Gemma 3 270M: The compact model for hyper-efficient AI
(news.ycombinator.com)
5354.
Is chain-of-thought AI reasoning a mirage?
(news.ycombinator.com)
5355.
5356.
Why LLMs Can't Build Software
(news.ycombinator.com)
5357.
What's the strongest AI model you can train on a laptop in five minutes?
(news.ycombinator.com)
5358.
Upcoming DeepSeek AI model failed to train using Huawei’s chips
(arstechnica.com)
5359.
Mbodi AI (YC X25) Is Hiring a Founding Research Engineer (Robotics)
(news.ycombinator.com)
5360.
What to expect at Apple's iPhone 17 event
(engadget.com)
5361.
5362.
Hooray! ChatGPT Plus brings back legacy models alongside an updated GPT-5 experience
(androidauthority.com)
5363.
5364.
5365.
I tested GPT-5 and now I get why the Internet hates it. Is it time to ditch ChatGPT?
(androidauthority.com)
5367.
5368.
5369.