The article switches between «publisher of a page» and «author of the content of a page» as if the former is never someone cropping out the signature of the latter out of a reposted drawing…
There is probably a limit on how little trustworthiness evaluation by a human needs to happen, which is in conflict with what the article’s author hoped to find.
… like the web “Payment Request API” which gives you a JavaScript API for generating a payment
Clarification: the Payment Request API is not a complete solution to processing payments: it’s just a way for the browser to unify collecting payment information. Like fancy password autofill, but for payment information, and in a way that also supports things like Apple Pay and Google Pay which can’t simply be autofilled with a constant piece of text (instead, they rely on challenge/response to produce payment information).
That is, it gives you a payment method, not a payment (a payment would actually move the money). Its designed to be used with some external service to form a complete payment.
So it makes sense that it doesn’t fit the author’s desire for a way to pay website authors, because that’s not what it is.
Every time ideas for funding come up, I’m reminded of Clay Shirky’s The Case Against Micropayments from December 19th, 2000.
The relevant part in this case is his focus on how much people ignore the psychological “decision cost” whenever the idea of money comes up and how that plays into the competitive advantage that aggregating services (eg. NetFlix, especially in its heyday) have.
Access to the entire browsing history feels like a privacy nightmare. Yes you receive a “complete” sample of which web pages a user has visited but, yeah this doesn’t seem great?
When I was at Microsoft, I proposed building a thing like this on top of the Confidential Consortium Framework. This is a toolkit that provides a distributed key-value store and append-only ledger that runs in trusted execution environments. The idea was that we could provide technical guarantees that only aggregate data would ever leave the system. You could build app-store integration and a browser plugin that would send an attention share (i.e. how long you spent reading / playing the thing) and the provider’s ID, signed with your client ID, and it would aggregate these. At the end of the month, it would calculate how much of your monthly subscription should go to each of the providers (either as a direct mapping from time, or with some weighting per article or a user-configurable thing to say ‘pay more for news than entertainment’ or similar) and emit a report showing how much each provider had received. You could also build this into open-source projects with dependencies, so that you could run some app and it could report what packages it depends on and some approximation of how much they’re used, and then time spent running the app could pay any library authors who opted in. Creators could also receive a notification of how much time was spent in their content, but not by whom (and not on a per-resource basis, which would let you put tracking strings in things and deanonymise).
The browser extension and client library would provide the current page / app with a signed attestation that they are being paid, which would let them automatically turn off ads if they were being paid directly, which would give users a nice model: pay and you don’t see ads.
The chief economist liked the idea in theory, but concluded that it wasn’t a thing Microsoft could do because companies like Google or Apple that control major app stores could roll it out trivially. If the main Android or iOS advertising frameworks have an option that is ‘no tracking and no ads if you’re part of our subscription scheme’ then they could probably get a few million subscribers overnight. Similarly, Google’s web ads could opt out of tracking and ad delivery entirely if they got the attestation from the browser (nonce embedded in the page is signed with the payment system’s private key and forwarded to the ad server, which responds ‘you’re paying for this, no ads for you!’ to the ad API request) and get any web site that uses Google Ads opted in automatically.
I think the author wants this: https://webmonetization.org
This looks interesting!
The article switches between «publisher of a page» and «author of the content of a page» as if the former is never someone cropping out the signature of the latter out of a reposted drawing…
There is probably a limit on how little trustworthiness evaluation by a human needs to happen, which is in conflict with what the article’s author hoped to find.
Clarification: the Payment Request API is not a complete solution to processing payments: it’s just a way for the browser to unify collecting payment information. Like fancy password autofill, but for payment information, and in a way that also supports things like Apple Pay and Google Pay which can’t simply be autofilled with a constant piece of text (instead, they rely on challenge/response to produce payment information).
That is, it gives you a payment method, not a payment (a payment would actually move the money). Its designed to be used with some external service to form a complete payment.
So it makes sense that it doesn’t fit the author’s desire for a way to pay website authors, because that’s not what it is.
Every time ideas for funding come up, I’m reminded of Clay Shirky’s The Case Against Micropayments from December 19th, 2000.
The relevant part in this case is his focus on how much people ignore the psychological “decision cost” whenever the idea of money comes up and how that plays into the competitive advantage that aggregating services (eg. NetFlix, especially in its heyday) have.
As unadopted as Brave’s basic attention token system is/was — among other barriers — the concept was so hopeful.
When I was at Microsoft, I proposed building a thing like this on top of the Confidential Consortium Framework. This is a toolkit that provides a distributed key-value store and append-only ledger that runs in trusted execution environments. The idea was that we could provide technical guarantees that only aggregate data would ever leave the system. You could build app-store integration and a browser plugin that would send an attention share (i.e. how long you spent reading / playing the thing) and the provider’s ID, signed with your client ID, and it would aggregate these. At the end of the month, it would calculate how much of your monthly subscription should go to each of the providers (either as a direct mapping from time, or with some weighting per article or a user-configurable thing to say ‘pay more for news than entertainment’ or similar) and emit a report showing how much each provider had received. You could also build this into open-source projects with dependencies, so that you could run some app and it could report what packages it depends on and some approximation of how much they’re used, and then time spent running the app could pay any library authors who opted in. Creators could also receive a notification of how much time was spent in their content, but not by whom (and not on a per-resource basis, which would let you put tracking strings in things and deanonymise).
The browser extension and client library would provide the current page / app with a signed attestation that they are being paid, which would let them automatically turn off ads if they were being paid directly, which would give users a nice model: pay and you don’t see ads.
The chief economist liked the idea in theory, but concluded that it wasn’t a thing Microsoft could do because companies like Google or Apple that control major app stores could roll it out trivially. If the main Android or iOS advertising frameworks have an option that is ‘no tracking and no ads if you’re part of our subscription scheme’ then they could probably get a few million subscribers overnight. Similarly, Google’s web ads could opt out of tracking and ad delivery entirely if they got the attestation from the browser (nonce embedded in the page is signed with the payment system’s private key and forwarded to the ad server, which responds ‘you’re paying for this, no ads for you!’ to the ad API request) and get any web site that uses Google Ads opted in automatically.