And that's why we should aim to learn *physical* generative models and causal chains!
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Pattern recognition functionality is possible with just individual experiences. While principled reasoning requires acknowledging communities' customs for proper functionality. Or else?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
the question remaining is: how to do the latter?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think that a successful models of "reasoning" would be like the first successful models of differential calculus. > Data -> Intuition --> Reasoning ---> Meta-Reasoning ---->... > Position -> Velocity --> Acceleration ---> Jerk ---->...
-
The value of reasoning is not that you need less data (you almost certainly need more to reason than to intuit), but that you can model an exponentially broader input distribution.
- Show replies
New conversation -
-
-
If I saw a problem I'd never seen before I would simply try to reformulate it as a problem I have seen before.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So underfitting vs. overfitting?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Eg to counter one’s own or others’ associative intuitions (preventing human automation bias)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Finally, some sense. :)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So you're telling us not to overfit? Cool.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.