Bruce Schneier on smart contracts:
The basic problem is that the code is the ultimate authority — there is no adjudication protocol — so if there’s a vulnerability in the code, there is no recourse. And, of course, there are lots of vulnerabilities in code.
To me, this is reason enough never to use smart contracts for anything important. Human-based adjudication systems are not useless pre-Internet human baggage, they’re vital.
I’d love to see a smart contract platform that allows human adjudication somehow! Devil’s in the details, but still, seems promising. It would be directly counter to most of that camp’s ideology and dogma though, so I’m not holding my breath.
(News flash: society is built on human trust. Code won’t change that.)
The Web3/smart-contracts community is way ahead of you and Schneier here. I think (?) best-practice today is to have a multisig, time-locks, “emergency halt button” etc for human oversight in the early stages of deploying a major new contract. (Tweeps: correct me if I’m wrong!)
Ooh, interesting! I remember other big early exploit(s) handled like this, ie with rollbacks by enough stakeholders. Good human adjudication, but also seemed so difficult and exceptional that they maybe proved the rule. Great to hear that might be getting standardized!