Hacker News

points by nailer 3 days ago | 0 comments

> > But does Terra 6.1 beat Sol 6?

> Run your own evals then if you have this much of a need to "correctly" choose.

No. OpenAI want my money, it's on them to communicate properly.

What would you want to see instead? Not sure about you but my needs are specific enough that I don't think OAI has evals for them, and I wouldn't want to give them that data even if they wanted to run it for me.

[–] nailer 3 days ago

“ChatGPT 6.1 Max” would be simple.

[–] austhrow743 3 days ago

Then you're going to have ChatGPT 6.1 Max Low and ChatGPT 6.1 Max Max.

[–] nailer 2 days ago

No you're not.

[–] austhrow743 2 days ago

You're suggesting renaming the effort level as well?

[–] ssl-3 3 days ago

How does one more-properly articulate "it depends"?

[–] nailer 2 days ago

"New model, middle version" is already communicated well by Apple.

[–] ssl-3 2 days ago

That's a computer, wherein: More is better.

But here we have LLMs, wherein: It depends. The smaller models can in fact outperform the bigger ones, and get some stuff done cheaper, better, and faster.

If the question is "Which one is best?" then the answer is "It depends on what the fuck you're doing."

(If that answer is found to be unfulfilling, then let it serve as an indication that a person may be better-served by improving their questions.)

[–] 0-_-0 3 days ago

On what?