The threat of preference modification

So YPlan. It’s a neat enough proposition. Basically, it’s a way of getting rid of spare tickets, or if it gets big enough, guaranteeing a minimum sale. It buys up what’s left. It asks new users some questions and runs a weighted Bayesian classifier to guess what they might accept and pushes recommendations to them. Whatever they take permits them to train the classifier further. They run an in-app credits system to make it easier to do giveaways and to let the app store handle payments.

This fills me with a sort of cybernetic horror. Recall this post from July and especially this quote:

maybe we could get the interface down to an iPhone app that would superimpose a bright white line over the camera’s view of the surrounding street just telling us where to walk and what to do and buy all day long. Wouldn’t that be a bit of a relief?

So that’s what they did. A magenta line running out of the office, with BUY popups. You might say this is a kind of hipster Luddism. All those years of self-fashioning and bedroom posters and obscure blogs and your girlfriend doing haughty walkouts from here and there, and these scum have an app. But this would be wrong. Ludd had a point, and the point was nothing to do with despising progress as such.

This is an example of a future beyond advertising. Advertising aims to influence you before you come to the point of needing to buy something. A different future tries to influence you at that point – my favourite example is a paint manufacturer whose mobile app matches the colour you point the camera at to the right paint and tells you where the nearest stockist is. This future, though, aims at something else.

The more college-smarty kind of economists like to talk about “desire modification”. If I could choose to like cheaper stuff I’d be richer! This is of course yet another failing of US education, with the homecoming jocks and nerd sessions and bull diamonds and whatnot, but that’s beside the point. A version of desire modification that might actually exist is preference automation, and here it is.

Rather than even having preferences that are influenced by advertising, why not hand them over entirely? Confirmation bias, the availability heuristic, and social proof will make sure you’re convinced pretty soon that whatever gets recommended is actually what you like. This means, of course, that preferences increasingly carry no information whatsoever, a serious blow for the entire notion of a market economy.

Whatever one says about “emotional labour”, it is difficult or perhaps impossible to deliver good service without taking the initiative, by definition unscriptable. Industry has long grasped this with regard to quality, as in Toyota production, or gone bust. Finance? Well, their managers either have no idea what they’re doing because they don’t or else they’re faking it.

In the future, at least in the UK, it’s even possible that we might end up enjoying more actual freedom at work than we do outside it. The consumer economy might be dominated by preference automation and spam that knows where your dog goes to school. The public sphere might be reduced to Tory badgering as replacement for public services or post-Blairite policing of social norms against talking in the street.