Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
alok-g
3 months ago
|
parent
|
context
|
favorite
| on:
Goodhart’s law isn’t as useful as you might think ...
Why not define additional metrics around what you would spot check and then produce a consolidated figure of merit as the metric to optimize. Aren't we doing that with AI training and general optimization (including regularization).
Consider applying for YC's Spring batch! Applications are open till Feb 11.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: