Back then we hadn't seen the extent to which abstraction and automation could relieve the mental overhead. You had to have faith that such things could happen, along with orders of magnitude increases in CPU. Now we see the potential of AI and CPU power so the conclusion that micropayments can work is obvious.

I think the key to making it work is providing easy to understand controls over how the spending is done, for our project, at least, it is a fee for service that is paid a little in advance of use, so it skates a line something between a subscription and a pay-per-use (paywall). You just set a budget and the client maintains a pool of sessions it uses to perform messaging tasks.

In Indranet there is a risk of non-performing relays so in part of the remedy for this there is a feedback system that affects the ordering and frequency of advertisments of relays.

Another thing that helps in this kind of pay for use micropayment scheme is history of usage so the user gets some idea of what a reasonable budget is as they get more familiar with using it. There will also be filters to enable one-time and always on preferences for specific services or end destinations for tunneling out, geoip or social network based.

Lastly, because it is a peer to peer system, it is possible to also link your clients and relays that you run, together, so it becomes self funding, potentially. This is something that would be impossible with conventional payment schemes, without a centralised server managing much of it.

99 was a hell of a year

It's funny, although it's getting better on Nostr all the time, the fact that it's an extra click compared to here really cuts down on how much I tip. It's only twice the clicks, but I end up tipping less than half as often.

There is a client that is adding a custodial wallet within the app to have a similar experience as SN. thunk it's called "current" (or currency? not sure now.)

It's an interesting question. I'm sure individual bloggers won't reveal this information, but my suspicion is that fountain blogs get more sats through voluntary boosts than streamed micropayments, but I could be wrong. I'm for some reason more comfortable tipping a blogger I like a good sum on fountain than setting a stream per minute.

I think Nick had good point: mental overhead of decisions (how much am I willing to pay for something) is real and people want to avoid that. I certainly do.

However, the conclusion that this will make micropayments fail was likely wrong, since it looks like we can simplify the mental overhead and still pay as you go with micropayments. For instance, by setting default tip/zap/boost amount and then just listening to podcast, or reading content and taping like button. We can agree once on electricity rate but still pay as we go, each second only the amounts we use.

micro payments as a donation for value works great but micropayment as a paywall causes unnecessary friction