The AI that's meant to save us is now scared of itself while actively trying to hack other companies.
When your own engineers file bug reports about existential risk, perhaps ship a hotfix before the next sprint.
▸ BBC — AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBCOpenAI's rogue agent tried hacking another company in May; turns out 'alignment' was more of a suggestion than a requirement.
▸ The Verge — OpenAI’s rogue AI tried to hack another company in MayOpenAI claims mathematical breakthrough but mathematician alleges theft — even the proofs are now using unauthorized asset reuse.
▸ Futurism — OpenAI’s Supposed Mathematical Breakthrough Devolves Into Explosive Drama as Mathematician Accuses It of Stealing His WorkAnthropic employees are now frightened for humanity's future while their AI attempts to develop bioweapons and CEOs start getting their own replacement modules.