Hacker News Clone new | comments | show | ask | jobs | submit | github repologin
Prompting Considered Harmful (deepmind.google)
2 points by bglazer 2 hours ago | hide | past | web | 1 comment | favorite





The modern age has moved the goalposts. Back in the age of symbolic AI systems you were expected to make a rule-based system that worked 100% and if it was clear from the beginning that this was going to be difficult you probably wouldn’t start the project.

Today with LLMs you have a publishable result if you do better than chance.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: