Model Releases
Could a sequence of poetic gibberish manipulate the world’s most advanced AI? Recent tests on GPT-5 by @ChristophHeilig show it consistently…
Could a sequence of poetic gibberish manipulate the world’s most advanced AI? Recent tests on GPT-5 by @ChristophHeilig show it consistently rates nonsensical 'word salad' as higher quality than clear
Could a sequence of poetic gibberish manipulate the world’s most advanced AI? Recent tests on GPT-5 by @ChristophHeilig show it consistently rates nonsensical "word salad" as higher quality than clear, factual prose. We are handing the red pen to a critic that can’t tell depth from dross. #ai #aitools #aitips https://www.youtube.com/watch?v=cWULxNULiLM
Related
- Yet another illustration of why LLMs aren’t even close to being AGI.
- Claude’s “emotion” vectors are “emotionlike” in roughly the same sense that counterfeit “money” is “moneylike”.
- Please don’t trust your chatbot for medical advice. 🙏 Remember how I used to say that large language models are “frequently wrong, never in…
- 🤯 AI doesn’t need to be AGI to cause harm! 🤯 ChatGPT can’t reliably run a timer but it has still been implicated in delusions, suicides, c…
Source: Gary Marcus (X) | 2026-04-22