Pinned
Bobbin's Threadbare
@bobbinth
blockchains, zero-knowledge proofs, virtual machines. working on @0xMiden
Joined September 2018
Self-custody (not having to trust anyone) is one of the core promises of crypto. But doing it securely is extremely difficult - at lest if you don't have the right setup. IMO, the right setup includes privacy, programmability, and layered security. More in the the article.- I would say that secure self-custody needs the following: - It must be a multisig. - It must be a "smart wallet" with spending limits (e.g., a hot key for convenience, but with limited spending capabilities). - It must be private. Any kind of entropy issues become much more
- Most importan innovations by decade (my opinion): 1900 - 1909: radio 1910 - 1919: automobile 1920 - 1929: airplane 1930 - 1939: antibiotics 1940 - 1949: nuclear fission 1950 - 1959: transistor 1960 - 1969: plastics 1970 - 1979: public key cryptography 1980 - 1989: personal
- A valid point - but not a reason to be pessimistic about ZK as there are ways to address this: - Run with training wheels for a while and double-check all proofs (e.g., in a TEE). - Formally verify the constraints. - Use multiple distinct proving systems to prove the same thing.This is generally one of the problems with zk. Exploits become impossible to detect. This is one of the reasons I was always skeptical of zk rollups which only publish state changes. A defi hack on such a chain would be insanely hard to debug, because the only thing that's
