Tag: language models

Culture Social

No Chatbot Will Agree to Sound Like a Pub Argument

Yasmina Khan

Scientists Spend £38 Million Learning That No Chatbot Will Agree to Sound Like a Pub Argument GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro and Grok 4.3 Asked to “Say Something Moronic With Entertainment Value” — Every Single One Refuses, Humanity’s Adequacy Review Published Anyway LONDON — Researchers at an institution that asked not to be named — presumably because it receives public funding and somebody in comms still has a…

Current Absurdities Technology

ChatGPT Hacks

Alan Nafzger

Entirely Sensible Observations About ChatGPT Hacks (Which Promptly Cease to Be Sensible) All ChatGPT hacks are just a polite way of saying, “Try asking it as though you actually know what you want.” Which, for most of the British public, is already a significant personal breakthrough — particularly on a Monday. The phrase “What am I not seeing?” is essentially therapy, but considerably cheaper and with none of the uncomfortable…