Lately, I’ve been feeling a bit dazed while browsing task platforms, like I’m not playing on the blockchain but clocking in at work: connecting to which chain, how many interactions, how long the intervals are, how much balance is left in the wallet... Honestly, it feels like feeding data to “witch score” systems. The platforms are pretty tough too—if you don’t set a threshold, a bunch of scripts will overwhelm you; if you do set a threshold, you push normal users into becoming attendance experts.

When I look at on-chain data myself, I care more about causality: the more you template behaviors, the more it looks like the same group of addresses’ fingerprints, making it easier to be identified as a single type of user; conversely, doing fewer actions, making them more random, and not forcing steps just for scores costs less and feels more natural. Recently, new L1/L2 incentives immediately boost TVL, and old users complain about “mining, selling, and dumping,” which isn’t without reason—many interactions have no real demand backing them, leaving only a bunch of busywork for scores… That’s how it is for now—I don’t want to turn “grinding” into a second job anyway.