I mean, I'm pretty sure that with equivalent teams building them, an AI business is likely to be spitting cash while L2s are a notoriously bad business model
Unsuccessful crypto teams should def try to pivot unless they have very strong conviction on what they're building