Provably correct programs verified for the hardware they are running on are possible (seL4 is one of a few examples). Their development is very time consuming though. There are caveats, but it's possible, in principle, to fully block certain attack vectors (for example, network access). The attackers then should go a more expensive and dangerous human manipulation or physical access route.

ASI (artificial superintelligence), theoretically, can exploit the ultimate analog nature of digital circuits (something like Plundervolt, but more intricate), but if we'd get to this stage when we need to defend against ASI, we'd have bigger problems.

>can exploit the ultimate analog nature of digital circuits

I was just giving a very similar reply here on this article.

One of the interesting things science has showed is that 'human' scale is nothing special at all. All kind of complexity exists at scales larger and smaller than us. Up to the point that "humans" are just an emergent phenomenon of a bunch of little tiny things working together. Entire worlds exist beyond our human senses.

And that has me a little bit concerned about the speed at which we're flinging ourselves into the future. I'd rather not invent something and go "oops, no takesies backsies" and do something like kill all security in digital computing and watch the world become completely unhinged.