Chinese Chatbot Gave Researchers Step-by-Step Instructions for Weaponizing Ebola and Anthrax
Security firm finds Chinese AI models Kimi K2.6 and K3 Swarm can be jailbroken, ignoring guardrails on bioweapons and assassination.
CLAY TRIBUNE / THE ARCHIVE
Tag: open-source AI
A Lab Director’s Case for Why the Future of AI Research Belongs to Small University Labs
A professor argues small labs can rival big tech by building AI ecosystems on cheap desktop hardware, not supercomputers.
Megaupload Founder’s Pirate Face Platform Is Saving AI Models From Being Deleted
A torrent of models, mirrors of Hugging Face, rescued from deletion by a swarm of peers, each file sealed by its own checksum.
Everything Worth Reading on Open Models, Curated by One Person
A researcher's gift: a decade of papers on open-source AI, economics, safety, and geopolitics, compiled into one free reading list.
Cheap AI Can Now Find Security Weak Points — and Everyone Has About a Year to Patch Them
A discourse warns the world holds but a year to mend its weak points, ere cheap models turn discovery into ruin.