Burning small amounts of a token can be a powerful incentive tool, for two reasons:
- Requiring token burn before doing some network processing is an effective anti-Sybil mechanism, which can be used to ensure that a particular action has a configurable amount of “cost” (by requiring that a certain amount of the token is burned).
- Burning tokens translates demand for particular services (whatever it is that participants are willing to burn tokens for) into demand for the network token as a whole. Requiring that a particular token is burned is a way for service providers to support the network as a whole (while also likely taking some fees themselves).
By “network processing”, I mean any service provided on the network, such as:
- storing data
- relaying messages
- solving / matching intents
- ordering transactions
- serving historical data and indices
- etc.
One can imagine a mechanism where payment for services includes (a) a burn component and (b) a payment (which goes to the service provider). It’s of course always possible that service providers try to outcompete by not requiring any burn, but the incentive equilibrium ultimately lies with the users.
Intents have a potentially powerful equilibrium (at non-zero burn), because users will typically want their intents to be compatible with existing liquidity. Suppose that most solvers only accept swap intents with 0.01% burn – most users will probably be willing to include that in order to get their intents processed. Two sides of a swap intent can also each require 0.01% burn from the other. Using \le instead of = for the burn check can allow a more gradual negotiation of what the required amount should be.
It’s easy to extend this mechanism to many tokens (including at once), there’s no reason in particular it has to be associated with just one. In general, perhaps this can be a way for users and service providers to continuously vote on which networks they want to support.