Thoughts on algorithmic regulation

Over the weekend, the Observer published Evgeny Morozov’s essay “The rise of data and the death of politics“. It is an impassioned plea about the rise of data and technology to solve political issues and the rise of Tim O’Reilly’s coined phrase “algorithmic regulation”. I am aware of the slants on both sides: both have very different views of the glass. I do wonder if they are coming at the same thing from different angles.

O’Reilly puts out his principles of regulation that is essentially tweaked like an algorithm to improve the outcome in his essay, “Open Data and Algorithmic Regulation“. His apparent focus is on how, as a set of societies, can achieve better legal outcomes for that society using open data and tweaking them when they do not work.

Through using Open Data and sharing, it would appear the argument goes, society can develop tighter ways of dealing with problems. Apps can be built on top of this. David Eaves discusses this briefly with the San Francisco parking meters changing prices based upon their usage. Of course there are ways that this might be potentially abused for gain such as MonkeyParking’s app which I caught up on via Pandodaily and Techcrunch but there does appear to be a public reaction to this as well as civic pressure to cease and desist. One might also see this as a validation of a part of Morozov’s position that technologists will abuse the process.

Andrew Leonard’s article “One code to rule them all: How big data could help the 1 percent and hurt the little guy” also looks at the issue of the algorithm being incorrect and requiring tweaking. He very firmly sides with Morozov and from the evidence, one can see why. In this case, YouTube are not using public data but the argument is similar.

If one looks at the Open Definition, there is no block on creating paid for applications using an Open Definition compliant license. As well as releasing data, companies are consuming data to make better products or visualisations. In part the data flow might be more towards consumption than producing but I do wonder if this is part of a societal shift that needs to take place and perhaps gradually is making that move.

Technology and open data have roles to play in this change but as creators we need to be aware of the moral and ethical choices that must be attached, in my view, to social data and political choices. As David Eaves points out in his post, open data is political. It will become more so as items are built with it and choices mades with it. We do not all think alike politically so there is a rub along that will have to take place.

I do not believe that we should blindly try and create everything in a ruch of joyousness. I do think that mistakes will be made and the approaches currently used will be tempered in time and change. However we do need to think about how we can make the use of data and technology work. The ongoing scandals of what certain intelligence services shows what might happen when the moral and ethical aspects are ignored.