Explain why differentiable computing isn't the future of computer science
>protip: you can't
Knock it off OP, there is no one that disputes this.
/thread
>>57159706
qft
yeah it will be the future, why even create this thread
>>57160219
So he can reference it in the future
>>57159694
Sure it will be used, but it probably won't be understood at a certain point.
We will never understand our brains but we'll be able to simulate one. The result will be extermination of the human species
>>57159694
Long tail misbehavior, constraint violation, lack of predictability, liability.
>>57160466
Average "muh neural networks" poster right here.