we are quietly shipping, anon sneak peek at LurkOS, our provable operating system sending 100k transactions between actors yes, this is all intra-actor and without proofs, but this is just the start.

Lurk does something similar with similar results. All data and code is commitment based, including proof I/O (which is then opened and committed to at each step of computation), so with content-addressed data storage you can generate proofs of computation over arbitrarily sized datasets stored in arbitrary locations. Even private data. Although we use Poseidon to get const sized commitments, this makes me wonder if there is some PCS method to use as a commitment scheme for proofs over large datasets that would perform better than Poseidon.

A little view behind the curtain, anon. You can now write provable JavaScript with Lurk. This shows why Lurk is so powerful. It took just 5 weeks to make it possible to compile JS into a JS representation of Lurk that can either be interpreted in place as JS or passed through Lurk for proving. Onwards to on-boarding millions of devs to ZK.