Nowadays you can run arbitrary verified computation at 100-300MHz using so-called zkvm. State of the art are probably risczero and succinct.

Practically speaking, write any program you want, compile it to riscv, imagine to run it on a pretty fast microcontroller, and in addition to the result you get a proof of correct execution.

I’d say it’s pretty practical, all major unlocks happened like in the past 3-4y and of course there’s a ton of research still happening.

This is the slow/generic version. For specific problems (aka dedicated circuits) it can be much faster.

There's a big asymmetry where creating a proof is usually dramatically more expensive than verifying it (which is often extremely cheap!). That might be OK for some applications but still rule out other applications.