Post

Yes, they figured out how to use cache timing to communicate across VMs—but humans figured this out a long time ago. The Spectre/Meltdown side channel attacks in 2018 were cache timing leaks. And they were widely written up. So the LLMs had a record of cache attacks in training data. 1/3

Spectre/Meltdown 2018 (It’s made it into undergrad class materials) web.mit.edu/6.033/2021/w...

Paul Kocher wrote about timing attacks like this back in 1996 paulkocher.com/doc/TimingAt...

Cc @prasad.bsky.social and @joshtpm.bsky.social who I saw amplify this. It’s not that it’s not impressive, it is. But these kind of timing (and maybe even temperature?! I vaguely remember this) sidechanneling effects across processes have been widely discussed and worked on by humans.

as I alluded to, it’s the same principle as American POWs in Hanoi tapping code to each other across a concrete wall regardless of how it was implemented in this specific case, the general type of solution is something we know about from human history and experience