Recently reviewed a few DAO proposals—on the surface they’re “optimizing,” but in reality they’re just moving voting power/allocation to some little inner circle by a tiny amount. Put plainly, what proposals should be scrutinized isn’t the slogans, but the incentives: whoever has sustainable “chips,” whoever can decide the budget, and whoever can cast one vote to shut others down. A lot of people only look at “good for the ecosystem,” but I’d rather dig into the attachments and the implementation rules—the smaller the font, the more critical it is.

That AI Agent wave also feels pretty similar. Everyone is hyping how smart automated interactions are, but when it actually lands, it still comes down to old problems like permissions, private keys, and whitelists. If security hasn’t been tightened properly, then talking about scale is just window dressing.

Personally, I trust data more. Intuition is too easy to be led around by narratives, while data at least lets you review and revisit what happened. If the rules are wrong, revise them—don’t try to muscle through it on emotion alone.