Bruce Sterling @bruces on Nostr: *This "superhuman" rhetoric really gets in the way here. LLMs don't break because ...
*This "superhuman" rhetoric really gets in the way here. LLMs don't break because they're more-human or less-human, they break because they are Large Language Models.
https://arstechnica.com/ai/2024/07/superhuman-go-ais-still-have-trouble-defending-against-these-simple-exploits/
https://arstechnica.com/ai/2024/07/superhuman-go-ais-still-have-trouble-defending-against-these-simple-exploits/