2 ms·
Isn't this at the level now where it can sort of self improve. My guess is that they will just use it to improve the model and the cost they are showing per eva
by attentionmech 2y ago
Isn't this at the level now where it can sort of self improve. My guess is that they will just use it to improve the model and the cost they are showing per evaluation will go down drastically.
So, next step in reasoning is open world reasoning now?
- dyauspitr 2y agoI don’t believe so. If it’s at the point where you could just plug it into a bunch of camera feeds around the world and it could only filter out a useful training set for itself out of that data then we truly would have AGI. I don’t think it’s there yet.
- attentionmech 2y agomay be for some sub-domains like math and code it can do that since the verification process can be done / relatively tractable