Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google is making private AI practical with homomorphic encryption

blog.google

Post reply on HN
Print view View original

151–160 of 305 posts

Re: Google is making private AI practical with homomorphic encryption

#151

Post by kccqzy »

jewel wrote: post #28
kccqzy wrote: post #18

Earlier quoted context omitted.

This kind of attitude is really disrespectful of decades of progress in cryptography. Without even considering homomorphic encryption, classic encryption is specifically designed to make intermediate nodes such as ISP dumb pipes that do not know the contents of communication. The ISP can store your communications on their server however they want. If you don’t agree with this model, I’m afraid modern cryptography doe…

Even if your files are properly encrypted, if you keep them on, say, Google Drive, then you might lose access to them if your account gets inadvertently blocked. I think GP is trying to say something along those lines. Not that encryption isn't trustworthy, but that even with homomorphic encryption you still give up autonomy.

That’s not a problem solvable by cryptography. Your ISP can’t read your traffic, but can cut off your traffic. You can use end-to-end encryption to encrypt your files to upload to Google Drive, but Google can choose to delete your files.

Re: Google is making private AI practical with homomorphic encryption

#153

Post by llleeeoooh »

siddthesquid wrote: post #132
llleeeoooh wrote: post #120

Earlier quoted context omitted.

this is not entirely true, I think. secure enclaves can provide guarantee such that even the host machine cannot inspect the contents within the VM. so even though the AI model itself needs to see plaintext, all is happening in the enclave which the provider cannot see. the main difference is where the guarantee comes from. for FHE, it comes from math, which we trust. for secure enclave, the guarantee comes from Inte…

- If an AI provider has control over the AI algorithm running in the secure enclave, they can easily have functions that provide them the plaintext through a separate channel. secure enclave does not prevent that - The output can reveal information to the provider, which homomorphic encryption would have protected - Inference is running on GPUs - so its moreso nvidia than amd/intel, but this is just a nit So homomorp…

yes I agree. although just to nitpick, there are ways to prevent side channel attacks within the enclave using software tricks. and output can remain in ciphertext via diffie hellman with client, etc. etc. but you're right that FHE provides a much more concrete guanrantee.

Re: Google is making private AI practical with homomorphic encryption

#154

Post by redeeman »

tbenst wrote: post #126
dhx wrote: post #63

Earlier quoted context omitted.

To throw out some real and up-to-date numbers from [1] for FHE at "128-bit security level", to sort 8x 8-bit unsigned integers on the most ordinary of desktop PCs, wait 3 seconds for the result. Want to sort 32x 8-bit unsigned integers instead? Come back 34 seconds later for the result. update: also see [2] for some primitive unsigned 64-bit integer operation benchmarks with the TFHE-rs library (winner in the sorting…

That is sobering for sure, I wonder what the theoretical bounds are on what is possible if known. Would be such a dream to use a Frontier LLM one day with homomorphic encryption, but this sounds wildly implausible based on where things are today.

they would never allow it, atleast not for regular plebs such as you or I, consider if you made it say something politically incorrect? cant have that

Re: Google is making private AI practical with homomorphic encryption

#157

Post by randomImmigrant »

dhx wrote: post #63
sabretooth1405 wrote: post #16

My master's thesis is on a topic in this field (Privacy Preserving ML) and from my understanding HE and other techniques have very high overheads(~10^3) on inference tasks and thus aren't very commercially viable.

To throw out some real and up-to-date numbers from [1] for FHE at "128-bit security level", to sort 8x 8-bit unsigned integers on the most ordinary of desktop PCs, wait 3 seconds for the result. Want to sort 32x 8-bit unsigned integers instead? Come back 34 seconds later for the result. update: also see [2] for some primitive unsigned 64-bit integer operation benchmarks with the TFHE-rs library (winner in the sorting…

This seems a fine tradeoff to me, depending on the context. There are datasets and operations on them where speed being sacrificed for privacy/security seems appropriate.

Ideally, give me a dial, to ask for encrypted intelligence when I need it. Kind of like a private chat, but with deeper privacy protections.

Re: Google is making private AI practical with homomorphic encryption

#158

Post by joshuamorton »

noident wrote: post #6

Does this rely on the Trust Me Bro model, or is there some way for the client to verify that the provider actually isn't able to see your inputs? I want to read a whitepaper but all I can find is the tl;dw conference presentation

> Does this rely on the Trust Me Bro model, or is there some way for the client to verify that the provider actually isn't able to see your inputs?

The point of FHE is that you only ever send encrypted data. So if you trust that Google hasn't broken the encryption algorithm entirely, they never access the plaintext.

Re: Google is making private AI practical with homomorphic encryption

#159

Post by elgertam »

sabretooth1405 wrote: post #16

My master's thesis is on a topic in this field (Privacy Preserving ML) and from my understanding HE and other techniques have very high overheads(~10^3) on inference tasks and thus aren't very commercially viable.

I saw a paper about this in early 2020 (pre-COVID shutdowns) at the ScaledML conference. I looked into it and had the same conclusions. At some point, running your own models in the clear is just more practical.

Re: Google is making private AI practical with homomorphic encryption

#160

Post by lupire »

u1hcw9nx wrote: post #123
michaelmrose wrote: post #91

Earlier quoted context omitted.

Which seems massively worse than a real local device in fact 2x is probably untenable to the point of uselessness because actually privacy sensitive matters need actual privacy that can't be defeated by your government telling Google to serve you compromised js and spy on you anyway and most people don't give 2 shits about privacy so they won't pay 10% more let alone 2x. I'm glad people fund things that are only of i…

I think you have completely wrong use cases in mind. You will not use this for normal compute workloads. Typical use cases are for doing biometric authentication without giving your biometric information, or sensitive queries using medical information. Apple has homomorphic encryption in image search. You can use your own photos encrypted into the cloud to search for landmarks in the image without revealing photos. P…

Is that (LLM) AI, though?
Post reply on HN