[mhh__]

(Chat)GPT-4s practical coding abilities are now 100x because it can code, run the code, and reason about its performance mid-response. They must be using fine tunes for this so the overall model could well be better too

[dontupvoteme]

You can do that as well, under your complete control. That's a framework they put around the model.

[wordpad25]

"model" is end-to-end, input-to-output, inclusive of the entire framework and it's guardrails and everything else

if they are able to detect hallucinations, filter them out and automatically re-run, that's a huge improvement in result, even though core model didn't get new training

[dontupvoteme]

That's the product. The model is the kernel of the product.

[mhh__]

That's what I said.

[oezi]

Only python though, right?

[ukuina]

Is it possible to learn this power?