It could work with a heavy efficiency penalty.

https://en.m.wikipedia.org/wiki/Homomorphic_encryption

Though I don't think gradient descent will select for something like this.

Far more likely is to obfuscate how it works so while it's not encrypted we learn little with the tools we have, and would have an extremely difficult time verifying something is absent.

Homomorphic encryption

>Homomorphic encryption is a form of encryption that allows computations to be performed on encrypted data without first having to decrypt it. The resulting computations are left in an encrypted form which, when decrypted, result in an output that is identical to that produced had the operations been performed on the unencrypted data. Homomorphic encryption can be used for privacy-preserving outsourced storage and computation. This allows data to be encrypted and out-sourced to commercial cloud environments for processing, all while encrypted.

^([ )^(F.A.Q)^( | )^(Opt Out)^( | )^(Opt Out Of Subreddit)^( | )^(GitHub)^( ] Downvote to remove | v1.5)

If it really had human like intelligence the encryption key would be 123, so it would be easy to fix.

You're not making a lot of sense. It's not clear you understand what software is, what an AI is, what encryption is. Why do you think encrypting an AI would affect it running? Do you have some specific encryption in mind?

​

Are you just high?

Intelligence is compression

But like yeah if you zipped the NPY files storing the weight matrices if most likely would not run