Download this as a PDF
In 2015, in a post called ‘Against ‘scale”‘, I quoted W. Brian Arthur (2009, p. 22, in David Harvey, 2014, p. 96) on the ‘combinatorial evolution’ of technologies:
slowly over time, many technologies form from an initial few, and more complex ones form using simpler ones as components. The overall collection of technologies bootstraps itself upward from the few to the many and the simple to the complex.
A staggering and chilling case of this is the UK data science company Cambridge Analytica (CA) which combines “25 years’ experience in behavioral change, pioneering data science, and cutting-edge technology to offer unparalleled audience insight and engagement services and products.”
One of their products is called ‘Validity’:
Greater scale, more certainty, better predictions, better decisions: to paraphrase Primal Scream, what a beautiful future!
CA’s marketing copy might seem blandly scientific or perhaps ‘metricophilic‘ but it masks actual practices which are truly shocking and are about to start biting very very hard in the everyday lives of millions of people. If that sounds like hyperbole (and it wouldn’t be the first time I’ve been guilty of that!), please let me know in a comment but first watch the video below of CA’s CEO Alexander Nix speaking about “the power of big data in global elections” and his company’s “revolutionary approach to audience targeting, data modeling, and psychographic profiling.” Then, for a recap and some additional background, read this translation of an article by Hannes Grassegger and Mikael Krogerus: ‘Trump knows you better than you know yourself‘.
This is the metricophilic, sociopathic and cold cutting edge of technology and politics today. Grassegger and Krogerus write that it is
not at all the case, as is so often claimed, that statisticians lost this election because their polls were so faulty. The opposite is true: statisticians won this election. It was just certain statisticians, the ones using the new method. It is a cruel irony of history that Trump, such a detractor of science, won the election thanks to science.
Some more of my thoughts on Nix’s talk and Grassegger & Krogerus:
- I don’t know whose idea it was to play CCR‘s ‘Bad Moon Rising’ over the intro but that’s a pretty sick joke.
2. It’s good that Grassegger and Krogerus (in translation at least) say that ‘Trump won the election thanks to science’ rather than using a determinist phrasing like ‘science/statistics won the election for Trump’.
3. Notice data broking firm Acxiom’s logo at the 5:37 mark and the mention of CA’s ability to “match offline data [purchased from Acxiom and others] to cookies to drive digital advertising.” I wrote in September 2015 about cookies and Acxiom in a post about the ‘student data blueprint‘ and urged educators to consider the risks of this ‘data blueprint’ to their students, especially ‘refugees or those who come from countries with authoritarian governments who might be very interested to know what their citizen is up to, e.g. China, Saudi, Venezuela, Thailand, or Russia.’ At the time, I wondered whether I was drawing too long a bow. Now, with Trump in power, I feel like the worst case scenario is materialising; what new uses could he make of CA’s “revolutionary approach to audience targeting, data modeling, and psychographic profiling”? Are we going to see what happens when ‘scientism’ is taken to its logical conclusions in every aspect of our lives, including in Australia? In the meantime, are we in education just going to continue ‘sleepwalking through our mediations with technology‘ (Selwyn, 2014, paraphrasing Winner, 2004) while paying lip service to ‘digital literacy’?
4. Notice the references to ‘psychometrics’ in the Grassegger and Krogerus article: “Psychometrics, sometimes also known as psychography, is a scientific attempt to “measure” the personality of a person” or, as Denny Borsboom puts is, ‘measure the mind’. It’s had a massive influence on education, assessment, language assessment and debates around reliability and validity. That’s where I’m planning to head in my next post.
pedagogablog 
hi Kyle
thanks for these posts, these issues need to be examined as much as possible
regarding companies like CA we need to separate some things:
if we are saying they somehow are a major factor in Trump and Brexit then this could be argued it is hyperbole if we listen to a Brexit mover such as Dominic Cummins (https://dominiccummings.wordpress.com/2017/01/09/on-the-referendum-21-branching-histories-of-the-2016-referendum-and-the-frogs-before-the-storm-2/) who make the reasonable point that many after-the-fact explanations are simply that.
issues such as third party buyng and selling of data can be tackled by governments if people organise around such issues
i don’t see that such “big data” approaches as any more onminous than other approaches given the context of the ideological dominance of neo-liberalism in which such approaches thrive
not sure if above made sense, put it another way “Don’t PANIC” as Douglas Adams would remind us : )
ta
mura
Hi Mura,
Thanks, as always, for reading, sharing, commenting and challenging me to think more!
You make a very valid point about ‘branching histories’ and I may have been too eager to accept the version of events which places Cambridge Analytica at the centre of things. Whatever else they’re capable of, Nix and his crew are clearly very good at self-promotion. Even so, I would say that the story is at the very least technically plausible and that the worst abuses of the technology are yet to come.
Governments – certainly the Australian government – have been terrible at tackling issues like…well, pretty much anything to do with the internet and I think the past is the best indicator of the future here. In other words, we have no reason to be optimistic about governments’ ability or even their willingness to address issues of online privacy and security. In fact, the Australian government is aggressively pursuing a data linkage policy without bothering to make a public case for it and the first projects resulting from this work have been a total disgrace.
I think ‘big data approaches’ are more ominous because they strengthen the coercive powers of governments (and other agents). Perhaps you’re right about not panicking; it’s too late anyway 😉
Cheers,
Kyle.