new | show | ask | jobs Github

mbowcut2 5 days ago

I'm not surprised. People really thought the models just kept getting better and better?

▲

segmondy 5 days ago | parent | next [-]

The models are getting better and better.

▲

giveita 5 days ago | parent [-]

That's expected. No one will release a worse model.

▲

sodality2 5 days ago | parent [-]

Not a cheaper one, or better in some ways, or lower latency, etc?

	▲	giveita 5 days ago \| parent [-]
		They do that too but right now it is an arms race as well.

▲

guerrilla 5 days ago | parent | prev | next [-]

Maybe. How would I know?

▲

jMyles 5 days ago | parent | prev [-]

...even if the agent did "cheat", I think that having the capacity to figure out that it was being evaluated, find the repo containing the logic of that evaluation, and find the expected solution to the problem it faced... is "better" than anything that the models were able to do a couple years ago.