For most people, the negative utility of deciding whether or not to do a transaction is on the order of a buck or two (based on some barely-remembered research from the ’90s). Pricing transactions less than this amount is inefficient; you don’t get enough extra transaction volume to compensate for the lower price (regardless of the value of whatever you’re selling).
That feels a lot like allowing your computer to write blank checks, which is a tough sell for users. If it were me, I’d want to cap the payments at some affordable maximum level. The service would likely find ways to ensure that users almost always hit the cap, after which point the cap is basically a subscription fee.
For most people, the negative utility of deciding whether or not to do a transaction is on the order of a buck or two (based on some barely-remembered research from the ’90s). Pricing transactions less than this amount is inefficient; you don’t get enough extra transaction volume to compensate for the lower price (regardless of the value of whatever you’re selling).
This argument doesn’t apply to the Agoric computing case though, in which the microtransactions are being decided by the programs and not the human.
And for common kinds of online activity, should be cheap enough that users can ignore it.
That feels a lot like allowing your computer to write blank checks, which is a tough sell for users. If it were me, I’d want to cap the payments at some affordable maximum level. The service would likely find ways to ensure that users almost always hit the cap, after which point the cap is basically a subscription fee.
Most people do this for other utilities all the time though (like power)