Oddbean

Just seems very inefficient, we really need zk primitive opcodes rather than the way they are doing it here. Side note: organizations can only be trustless if they are algorithmic, which means they are no longer human organizations. As soon as other humans are making decisions, you have to trust them to some extent. But if other humans are not making any decisions it's not an organization. So if you're going to build an organization of humans, the only thing really need from Bitcoin is unfuckable money and unfuckable timechain (proof that this data existed at this time). For everything else you can just use nostr.

Re efficiency, yes, 100%. But the value of a financial network rests on three things: adoption, adoption, and adoption. Bitcoin has it, and weird hacks to bolt on new features are easier than building widespread adoption of an altcoin. Re organisations, yes 100%, but there are valid corner cases like trust-less futures contracts that are now somewhat possible. A widely trusted oracle on local weather could feed into a self-executing contract that either pays out or does not pay out based on events, allowing Weed Farmer Ted to securely stay in business during a drought, even if he has to pay some of his profits to counterparties if instead its a good year. No lawyers required, and no vampire middlemen.

Not trying to shit on the idea, but I still don't understand why you'd use BitVM in the oracle example. The oracle which you need to trust anyway is essentially just signing a transaction to release sats, so why do you need BitVM?