Could a neural net be trained to run hashing algorithms more efficiently?
It looks like p=np problem
Why would you use a neural net to make monopoly money?
>>61836623
this.
There's no discernible correlation between previous correct solutions and future correct solutions.
So no
>>61837401
I think you are forgetting that it can learn how to do the problem better. It is running the same hash over and over so it learns how to do it better by design
>>61837557
but existing algorithms are already incredibly fast and deterministic
>>61837557
IQ2: the post
Dumb pajeet: the post
Current state of g: the post
>>61837598
That post also sums up Deep Learning fags in general
>>61836598
Yes, but it would be very retarded.
>>61836598
No, you're retarded. Neural networks are used in tasks where you can co-relate the right answer and the question and have a ton of data. They're also not faster than an optimized alghorithm that will do the same task - less runtime performance for less programmer time. The AI's input would be number of 0s required + the block and the output would be a number that when added to the bpock and hashed gives that many 0s at star/end. But since SHA is a crypto hash alghorithm there hasn't been found a way to reverse hashing, we can only try different numbers and see which will work., so the NN will just run the same shit a farm would be running but with overhead.