What would be a better metaphor for such "chains of functions"?
-
-
-
Or we can replace chain=networks functions=neural. Neural networks works great as a metaphor to work "similar" to human brain
- Show replies
New conversation -
-
-
So Functionchains? Re. Blockchains. Should propagate this.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
LeCun is promoting “differentiable computing” as a new monicker to replace DL and NN. I like it
-
"Differentiable programming" is a lot better, but to be honest that seems a lot more general than what we do in deep learning. Deep learning is a very very small subset of what differentiable programming could be
- Show replies
New conversation -
-
-
I can understand your pendulum swing but this is like saying living beings are a sad misnomer; they are neither living nor being; just bunch of proteins by happenstance forming stable and replicating molecules. Definitely accurate but ...
-
Because mechanisms may be high school level but emergent properties can be complex and powerful
- Show replies
New conversation -
-
-
Yes, and regression is an awful name for fitting a function to data. But the reason we call is regression is for historical reasons (inspired by "regressing to the mean"). Same thing with neural networks.
-
Historical reasons didn't keep cycles per second (CPS) from becoming Hertz or centigrade from becoming Celsius, even though the new names for both of those things were much less descriptive. We have to fight back.
End of conversation
New conversation -
-
-
Remember the man in the Bourgeois Gentilhomme who's delighted to find he's been speaking 'prose' his whole life? I've been writing IF-statements in Excel my whole career - imagine how delighted I was to find out I was making AI ;)
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.