Thread · 4 tweets · 4 May 2020

I don't have a fully-baked solution there either, but I've been thinking that part of the solution might be to make this behavioural data rivalrous (by law obviously, that can't be done technically). It would introduce friction to the scale of the big data hoarders.

Personal data should always be owned by the data subject, but we can influence who is a data controller of it (in the GDPR "determines the means and purposes" sense). Today you can have multiple controllers of the same data, and the cost of sharing is close to zero.

But the value extracted from the data is not intrinsic to it: it benefits from network effects with other data a given controller has. At the point of collection your behaviour might be shared with a dozen controllers, but the biggest reap the most value — in a vicious cycle.

Limiting controllership would heavily change the economics of data sharing, and would cause stringent limits to scale for major collectors — not a solution on its own, but I think it could be part of one.