I feel like this could be really useful for making predictions, like stock prices or sports results. What do you guys think?
1 Like
I think you give them to much credit.
It’s more like “Hey, ChatGPT, read these supplied documents. Then tell me what order these events should be released, also include a web write-up for each and video announcement scripts.”
If they actually thought of anything’s themselves I would lose faith 
lol
1 Like
is it possible to share the o1 pro with family? My wife and kids?
1 Like
Reinforcement Fine-Tuning needs a grading function. What if the grading function needs to be a Human?
In my specific case, I’d like to use Reinforcement Fine-Tuning to train o1 on our internal “scripting language”. So the output is not something like a definite answer or value … the output is a script and can be different scripts that all work.
Any plans for a human grading system?
1 Like