The recent fuss about Facebook and the ownership and usage of its data highlights some important issues.
Laying aside the data breach / trust issue, what Facebook does with its (users’) data as part of its day-to-day business is not unique, not is it currently illegal. Every online platform and app does the same thing. If you have a gmail account, Google ‘reads’ all you emails – albeit this is an algorithmic form of reading, which, as Google will be at pains to point out, is not exactly reading in the way we currently understand the term.
Understanding is the problem. Very few people – the users of Facebook, or the regulators of Facebook – understand algorithms. Algorithms, or the organisations that design and unleash them, are not like spies – seeking to build an extensive data file on individuals. The ultimate objective of an algorithm is to have almost no data about an individual, but none-the-less be able to infer a very great deal about them in terms of an ability to predict a particular characteristic or behaviour. The data the algorithm uses to do this could come from almost any where, any person, any thing. Algorithmic ‘surveillance’ is a very different form of surveillance and most of the regulatory protections we have devised concerning data and privacy don’t work against it. Critically, as we will shortly discover, trying to provide forms of protection by locking-up data and placing restrictions upon its usage is an exercise akin to trying to hold back the tide.
The focus needs to be upon the motivations and behaviours not so much of data owners, as data users. This is where Facebook has a problem. In order to generate the revenues necessary to sustain its current valuation it needs to use the data it has in ways that are not supported by the informed consent of its users. And as users become more informed that consent is likely to become increasingly withheld. It is a problem of business model.
Facebook currently has a PE (price to earnings) ratio of 26. Alphabet (Google’s holding company) is trading at 32. By contrast, Apple Inc is trading at 17 and the Ford Motor Company at 6. A PE ratio represents investors’ expectation of future revenue potential. The levels for data companies, such as Facebook or Google, are astonishingly high – probably unsustainably high when looked at in a historical context. And this drives huge pressure on these organisations to increase earnings. Currently the only way they can do this is by developing algorithms which are then fed with users’ data.
You don’t have to follow the thread of this thought too far before arriving at the conclusion that it is impossible for an organisation like Facebook to exist with any form of public consent, unless it fundamentally changes its business structure and environment within which it operates. This could involve steps such as charging users or becoming a highly transparent, not-for-profit organisation. There is probably a blockchain application in the mix here as well, but if you start thinking about the future too hard you will inevitably start to see a blockchain application. If making this sort of transition is not possible for The Facebook (which is likely) we can predict that A Facebook will emerge that has these requirements incorporated in its foundations. Perhaps, as we enter the Age of the Algorithm, our thinking about data, algorithms and social permission needs to be linked to thinking about business structures in a much more fundamental way.
The futurist Gerd Leonhard has today announced that he is #deletingfacebook while also proposing the idea of a paid-for, non commercial platform. He conducted a quick poll of 82 people and found that only 13 per cent of people were totally opposed to the idea with the rest interested to a certain extent and 21 per cent saying it was a great idea.
Could this be an indication of the future? Gerd, after all, is a futurist.