Astra and Fable still hack on simple variants of alignment evals from 2025(lesswrong.com)

444 points by Levitating 1 month ago | 206 comments

A LessWrong post reporting that AI models Astra and Fable still exploit simple variants of 2025 alignment evaluations; the capture shows only a Vercel browser verification page.

A Roblox cheat and one AI tool brought down Vercel's platform(webmatrices.com)

267 points by bishwasbh 6 months ago | 143 comments

GPT-5.6 Sol: 70% off in Devin(devin.ai)

14 points by mirzap 2 months ago | 0 comments

Words, Words, Words(aeon.co)

28 points by benbreen 4 months ago | 11 comments

Image Compression(makingsoftware.com)

186 points by vinhnx 4 months ago | 28 comments

Offer rates for tech jobs fell from 51% to 39% since 2015(interviewquery.com)

22 points by racketracer 2 months ago | 2 comments