Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Retric
13 days ago
|
parent
|
context
|
favorite
| on:
The "confident idiot" problem: Why AI needs hard r...
It’s not about thinking, it’s about what they are trained to do. You could train a LLM to always respond to every prompt by repeating the prompt in Spanish, but that’s not the desired behavior.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: