I absolutely agree on the sweet spot but if you are like me who remember 1978 when I looked at an Apple IIe with a VisiCalc thinking “so what do I do with this next” you would probably understand why the HAT way of thinking is that how we start is not how we grow/evolve. We start by getting the data that we already have into our HATs. This gives people something to look at. Something that makes an abstract into a concrete. Something that, in doing, we learn, because the actual interacting with data brings a host of issues in all its concrete-ness that the abstract will never get.
We evolve by getting more original data into HATs (hence HAT value proposition is backend-as-a-service for apps. See our developers portal).
APIs are unstable (there’s a bunch of our HAT engineers who are fist bumping you now) and they do create dependencies but they serve 2 purposes for us. (1) they make real time on demand data owned by us REAL (2) they make data mobility direct from individuals POSSIBLE. That means personal data does not need to be hoarded by an app. It can be inquired, not acquired. And remember that the data goes into the HAT as subject access request so the ownership is real.
Now we can go to government and say that individuals can give data in real time like they can give money in real time. Until this is possible, the centralised systems will have to prevail, because coming down too hard on them is seen to stifle the market. But showing that an alternative exists ala “here’s a black swan” (sorry NT) is where we want to get to for policy. That changes the game.