Astra and Fable still hack on simple variants of alignment evals from 2025Article URL: https://www.lesswrong.com/posts/munJKF7iWMsWJLAH2/astra-and-fable-still-hack-on-simple-variants-of-alignment Comments URL: https://news.ycombinator.com/item?id=49684393 Points: 10 # Comments: 0https://www.lesswrong.com/posts/munJKF7iWMsWJLAH2/astra-and-fable-still-hack-on-simple-variants-of-alignment#Tech #Technology #TechNews #AI #Gadgets #Software #Cybersecurity #Apple #Google #Microsoft #Startup #HackerNews [Hacker News]
Related
The world of Internet intimacies #negativepid #digitalInvestigations #OSINT #cybersecurity #AI #tech #onlineInvestigatio...
The world of Internet intimacies #negativepid #digitalInvestigations #OSINT #cybersecurity #AI #tech #onlineInvestigations #robotics #cyberpsychology #cybercrime https://negativepi...
OpenAI Offers Free ChatGPT Plus to Veterans and ServicemembersOpenAI reports more than 20,000 transitioning servicemembe...
OpenAI Offers Free ChatGPT Plus to Veterans and ServicemembersOpenAI reports more than 20,000 transitioning servicemembers and veterans have claimed a free year of ChatGPT Plus, wi...
A quick note about Next Horizon: I share links to science articles, and some of them feature AI-generated illustrations....
A quick note about Next Horizon: I share links to science articles, and some of them feature AI-generated illustrations.These images help visualize complex ideas. They're not prese...