AI News

Show HN: Gemini 4 Argon plays StarCraft

I've refreshed StarSkirmish Bench, a benchmark where LLMs play StarCraft BW by writing code, with results from the flagship models of OpenAI, Anthropic, Google, and Muse

How to make System 1 LLMs show System 2 competencies?

Coconuts in Medieval Britain (2015)

Turn your regular JavaScript code into a neural net

A model where companies lease critical minerals instead of selling them

BeBox is a Haiku application that emulates Be Inc.'s BeBox

Call Center Jobs Are Declining, but 14,000 Calls Show Humans Still Outperform AI

"I'm a Lawyer." Said the White House Press Secretary. Florida Would Like a Word

Why We're Banning AI

Anthropic Update on Independent Evaluators

More on Anthropic's Scion ripping off Mononoke

Muse Spark Open, the Most Capable US Open‑Weight Model to Date

Market selloff, rising rates test AI data centre deal boom

Krea 3 – Frontier Image Model 4× faster than Nano Banana Pro

Neural Networks Are Already Clean

How to Train Your Robot

Simons Foundation Launches $15M Effort to Help Mathematicians Navigate AI Era

Using AI to reverse engineer hardware