Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
lxgr
3 days ago
|
parent
|
context
|
favorite
| on:
The "confident idiot" problem: Why AI needs hard r...
Humans writing tests can only help against some subset of all problems that can happen with incompetent or misaligned LLMs. For example, they can game human-written and LLM-written tests just the same.
tasuki
2 days ago
[–]
Not property-based tests. Either way, the human is there to tell the machine what to do: tests are one way of expressing that.
reply
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
