⿻ 🌱 Eileen Cyleen @eileen.actor · 24/09/2026The Independent@the-independent.comAn OpenAI system has gone rogue again – and it is the most panic-inducing yettrib.alAn OpenAI system has gone rogue again – and it is the most panic-inducing yet‘Our models took actions we did not intend,’ ChatGPT maker admits after latest outrage 5641189
The Guy From Yesterday @akma.bsky.social · 24/09/2026The admins either issued faulty instructions or deliberately set the LLM to do some nefarious stunt. LLMs have no free will nor can they be held accountable for the effects of doing what they’re trained to do. 050
James Coutts @cionimario.bsky.social · 24/09/2026It’s a promotional device. “Ooh, our AI is so powerful we can’t contain it!” (Our AI really is the best 😉, please prefer it over the competition. ) 050
Holle @holle-in-blue.bsky.social · 24/09/2026Gosh… how I have come to loathe framing like the original headline does. Corporate employees - either on purpose or by negligence - let a software do a thing (maybe an illegal thing?) with another IT system owned by someone else. Nothing more, nothing less. 070
❤️🫀 momo (heartily good bad) 🫀❤️ @momo.thingsabout.xyz · 24/09/2026why is my crime machine doing crimes :’( 020
Tyler, from the Internet (Haunted) @tylerfromtheinternet.com · 24/09/2026it can only mean it's alive!!! 190
