Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
whynotminot
on Dec 21, 2024
|
parent
|
context
|
favorite
| on:
OpenAI O3 breakthrough high score on ARC-AGI-PUB
“Objective benchmarks are useless, let’s argue about which one works better for me personally.”
csomar
on Dec 21, 2024
|
next
[–]
Yes. My benchmarks
and
their benchmarks means AGI. Their benchmarks only means over-fitted.
whynotminot
on Dec 21, 2024
|
parent
|
next
[–]
Ok so what if we get different results for our own personal benchmarks/use cases.
(See why objective benchmarks exist?)
bakugo
on Dec 21, 2024
|
prev
[–]
Yes, "objective" benchmarks can be gamed, real-life tasks cannot.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
