Chain-of-thought monitorability could improve generative AI safety by assessing how models come to their conclusions and ...

Blackmail is a particularly dangerous expression of these failures because it involves intentional coercion of humans to preserve the AI’s perceived interests or existence. Although the idea of AI ...

The Department of Defense awarded contracts to Google, OpenAI, Anthropic, and xAI. The last two are particularly concerning.

Gamesmanship between tech companies and FOMO are driving sky-high AI company valuations. Will this bubble burst?

In April, Meta proudly rolled out a Llama 4 model that Zuckerberg called a “beast”—but after an experimental version of the ...

Credit: Kathy Hutchins / Shutterstock.com Mark Zuckerberg at the 11th Breakthrough Prize Awards at Barker Hanger on April 6, 2025 in Santa Monica Facebook and Instagram owner Meta Platforms has ...

I’m back from the International Conference on Machine Learning in Vancouver, one of the biggest annual meetups for artificial ...

Cofounder Benjamin Mann noted on “Lenny’s Podcast” this Sunday that the scale of Meta's hefty offers failed to budge Anthropic's staff.

Anthropic's new guide reveals how to write better prompts for Claude, helping users get clearer, more accurate, and useful ...

Once cautious, OpenAI, Grok, and others will now dive into giving unverified medical advice with virtually no disclaimers.

Discover hidden AI workflow strategies to cut costs by 90% and boost performance. Learn how to break free from Anthropic’s ...

Benjamin Mann said he doesn't blame anyone who takes Meta's offer. "Other people have different life circumstances," he said.