We will all have more leisure time in the future
|
Canards of modernity: an occasional series. No 1: We will all have more leisure time in the future.
This idea, propagated by thought leaders like Ray Kurzweil,[1] John Cryan[2] and dashing young heir to the Susskind clairvoyance dynasty Daniel Susskind,[3] posits that robots and artificial intelligence will, shortly, supplant the need for human labour altogether. We, the meatware, will shortly be technologically redundant. The lot of us.
Which poses the question: seeing as we won’t be working Jacquard looms, wiping arses, writing wishful dystopian techno-political tracts or managing business change programmes anymore, what to do with all the spare time we’ll suddenly have?
On this view there be nothing left to do: we will all just loaf around, playing chess and drinking grappa the way Mediterranean pensioners always have. Sounds great, doesn’t it! Especially if you don’t dwell on the thought that your robo-slave could wipe the floor with you at chess, too, if it felt like it.
In any case: nice lack of work, if you can get it.
Now if something about this scenario nudges your implausibility hooter, that makes two of us. For one thing, from our vantage point, the last thirty-odd years have been one long Cambrian explosion of technology (t’internet! iPhones! SETI@home! Uber! Drones!), but so far there is no sign of any extra leisure time. Work — new work, derivative work, previously unimagined work — is piling up. Granted, a lot of it is crap work: internal audit, software change management, six-sigma process analysis and talent acquisition — hardly the effervescent future we envisaged as wild undergraduate dreamers — but it definitely is work, and it definitely isn’t going away. (You do wonder: had our elders foreseen this new world of work, would they have told us? Who breaks a butterfly upon a wheel?)
In any case, the unknowable then has coagulated into a crushing now, and the reality is this: there is more work to do now than ever. Work is here, it is overwhelming, it is debilitating, and it is barricading the way to that chessboard in Στούπα, with no end in sight.
The millenarians ask us to believe that a 5,000-year-old asymptote will suddenly invert. But how? When? Why? No-one knows what we will be doing in ineffable, co-evolving future: we don’t, and chatbots definitely don’t. It might not be regulatory change programme management (though, ahhh, don’t bet on it) but, if the past is any guide, it will be something, it will be tedious, and it sure as hell won’t be slugging back génépi over a backgammon board in the Haute-Savoie.
“Ahh,” say the digital prophets of our time, “but is the past any guide? We say it is not.[4] This time is different. This time the machines will not just be our handmaidens; they will replace us altogether.”
Okay; let’s run with that for now. Even if you are right, thought leader types, your theory of technological unemployment assumes:
- that all human activity in the economy can be and, before long will have been, articulated in a way that can be entirely, reliably and cheaply carried out by artificial intelligence;
- that once they have been so articulated, those activities will nonetheless hold their value and won’t become worthless overnight, as has every other artisanal craft made redundant by machinery in human history;[5]
- that an economy which has been thus automated to saturation, and to which human participants no longer contribute, will still function more or less as normal, and
- that, in other words, an entire economy not only can be fully determined — solved — but has been: that our current polity is in a fully taxonomised, Taylorised end-of-history state in which no new activities or work categories are possible, and all that machines can look after those that do exist.
These assumptions are transparently absurd. They get the Yngwie Malmsteen paradox 180° back to front. The more information processing power we have, the more complicated our information structures will be. This is because we are lazy, backward-looking creatures. Increasing automation increases complexity, multiplies the interconnectivity between components of our distributed systems, accelerates the speed at which data circulates, and tightens the couplings between components. The JC has been harping on about systems theory and complexity a lot recently, but these are not trivial problems. Artificial intelligence cannot solve them. We are going to be needed for a long time yet.
In a Nutshell™: put away the checkerboard and stick the limoncello back in the cupboard. There’s work to do.
This article was written by a disembodied neural network. © 2020 Klaatu Barada Nikto.
See also
- Rubbish maxims
- Technological unemployment
- This time is different
- Perspective chauvinism
- [[Systems theory]
References
- ↑ The Singularity is Near
- ↑ Rumours of our demise are greatly exaggerated.
- ↑ A World Without Work: Technology, Automation, and How We Should Respond. See also David Goodhart’s more thoughtful (but still, on this point, misguided) Head Hand Heart.
- ↑ They have learned the compliance mantra: Past results are no guarantee of future performance.
- ↑ Ask yourself: how much would you pay to deliver a first-class email? Or to get your digital photographs developed?