Uniswap (UNI)

hmm.... first: the distinction between “compensating for temporary downsides” and “attracting permanent mercenaries” is cleaner in theory than in practice. you say DeFi liquidity mining is defensible because it compensates for hack risk. but in 2020–21, the dominant reason people farmed yields was not fear of hacks, it was that yields were huge. the risk was a footnote. those farmers left en masse when yields dropped. by your own definition, that was “bad” incentives attracting users who would not stay organically. yet Uniswap and Curve survived, and their liquidity today is deeper and more distributed than pre-farming eras. so maybe the clean distinction breaks down: sometimes mercenary capital performs a permanent infrastructure function (deeper books, tighter spreads) even if the humans operating it are purely extractive. second: you say incentives that bring in users who wouldn’t use a mature version are bad. but what if the mature version never arrives without them first? this is the classic two-sided market problem. a social app with zero users is useless. you cannot bootstrap a network purely organically in 2026 when attention is monopolized by incumbents. sometimes the choice is not “good incentives vs bad incentives” but “pay for distribution vs never reach scale at all.” the question is whether the incentives are designed to phase out and whether the product retains a subset. many successful Web2 platforms did exactly this, PayPal paid $10 per signup, dropbox gave free storage. crypto just made the payment a tradeable token, which made the subsidy visible and distorted the timeline. but the underlying mechanism is not categorically different. third: your cynical 2021–24 take is mostly true, but it implies a clean separation between “real product work” and “narrative engineering.” i think that separation is a luxury of hindsight. In 2021, it was not obvious which apps would become useful. dome projects that looked like pure speculation like NFT marketplaces which later became legitimate distribution channels for artists. the problem is not that people used incentives. the problem is that incentives were the only plan, with no product hypothesis to test afterward. If you had a thesis (if we subsidize liquidity for 12 months, we will retain X% because traders like the UX), and you tested it and pivoted, that’s product development. If you had no thesis and just extended emissions forever, that’s a casino. the distinction matters more than presence vs absence of incentives. fourth: the “charging fees, paying back in tokens” model you endorse is cleaner, but it has hidden friction. requiring users to pay fees and get rebated in tokens forces them to care about token price, tax events, and claiming processes. for mainstream users, this is not turning users into investors by default, it is a confusing chore. the reason apps don’t do this is not that they are greedy. It’s that most users prefer app pays me sometimes over i pay app and maybe get money back later. and last, you say successful apps now do bulk acquisition through utility, not incentives. name them. Farcaster? Had a paid invite system effectively a financial filter, not an incentive, but still monetary.

btw even Ethereum L2s are paying millions in sequencer fees to attract users via gas subsidies. the pure product is so good people just come story is rare. usually there is a financial lever, just denominated in dollars instead of tokens, or hidden in infrastructure and not user grants. imo incentives must be deployed as a loan to your growth hypothesis, not as a salary to your community. if you are paying for a behavior that has no durable value after payments stop, you are burning money. but if the behavior (liquidity, content, referrals) creates assets that remain valuable even after the mercenaries leave, then the incentives were infrastructure spending