Wednesday, 3 August 2016

Robots stealing all the jobs? - Nationalisation is the Answer, not a Basic Income


In the last year there has been increased discussion about introducing a 'Universal Basic Income' (or UBI). A Universal Basic Income would be a replacement for welfare where everyone gets an amount of money, say £100 a week, from the government as a right, just for being a citizen of the country. Unlike current welfare there are no conditions to meet to get it, either of age, or income, or anything else, and everyone gets the same. It is much simpler than current welfare, and the idea is to give people enough money to live on, just, to free people from the fear of destitution regardless of circumstances.

It is a rather expensive idea, but one that has significance advantages in areas such as simplicity and reliability. Recently, a perhaps unexpected source of support has come in the shape of various internet and technology billionaires, millionaires, etc. These captains of cyber-industry are worried about what happens if robots take all our jobs. They think, as do some others, that the rapid development of artificial intelligence and robotics will, over the next few decades, rapidly put most people out of work, not just in manufacturing, agriculture or low skilled services, but even in previously safe white-collar professions like Law, Medicine, Accounting, or whatever. They fear this will lead to a society where an increasingly tiny group of capitalists own all the robots, and everyone else is unemployed and penniless.

(Credit: Shutterstock/Salon)
Funnily the main negative thing our app-overlords have noticed about this is that if we're reduced to a vast penniless underclass there will be nobody who can afford to buy the products their army of robots and AIs produce. Hence their support for a UBI, to re-close the circle between production, purchase and consumption. Coincidentally, apart from the UBI part, this is exactly what Karl Marx thought would happen to Capitalism in the late 19th Century, but quite obviously didn't. Hence the failure of Marxism.

Ignore for a moment the obvious problem with this vision: that technology has been wiping out jobs en masse since 1750 but we haven't run out of jobs yet. The 'Tech' crowd's response is still ludicrously complicated when you think about it. UBIs are horribly expensive, so this would require massive taxation to pay for it. But if most of the population are penniless then the tax burden will have to fall heavily on the people with the money i.e. the tiny number of hyper-capitalists. So these people would be paying massive taxes, so the money can be distributed to everyone, so they can buy stuff, so the tiny minority can make massive profits on their capital, a large proportion of which would go in massive taxes, so the money could be redistributed, and so on.

It's worth noting at this point that the contemporary world is historically unique in that our aristocratic elite is both socially and economically liberal. That means they all like to think of themselves as nice, progressive people, but they also love venture capital, initial public offerings, and, oh yes, hate taxes. So I'll let you guess how long this re-distributive scheme would last before the elite decide that serfdom isn't so unthinkable after all.

But anyway, back to our dystopian future. So almost everyone has lost their jobs because robots can produce and provide goods and services of all kinds cheaper than human beings can. This means there pretty much is no problem of scarcity anymore, because AI can do basically everything. In this case there is a simpler solution than UBI, and it's called Socialism.

If we ever reach the point where most jobs disappear and capital is concentrated among a tiny elite who own all the robots and software, then we should just nationalise the robots and the software. That cuts out the ludicrous circularity of the UBI scheme (giving money to people, so they can give it to plutocrats, who give it back to the people, so, the whole process can go round again).  This all seems designed purely to keep money and power concentrated within a tiny, pseudo-aristocratic elite after real competitive capitalism has ended. Just nationalise them and then, once the robots are in common ownership, just use them to make stuff and provide services and just credit everyone directly from the output thus cutting the fat-cats out of the loop entirely. Problem Solved.

Now, I'm a right-wing, conservative, free-marketeer, so why am I suggesting socialist revolution if things go a certain way? For the answer we have to go back to Marx. The main problem with Marx's prediction that socialist revolution was inevitable was that it was based on assumptions about how economic development would occur . . . that were all wrong. A pretty large flaw. Marx thought capital would be consistently concentrated in fewer and fewer hands while wages were driven down more and more, until the whole system collapsed under its own inequality. But even his own data in Das Kapital showed that wasn't happening, real wages for workers were rising, as was their wealth (albeit from a very low base) and by the time of the Russian Revolution history clearly wasn't panning out how Marx predicted. According to his own theory revolution should happen in the most advanced societies. According to Marx Russia was the last place revolution should occur. Even more contrary to Marx Capitalism has chugged along more-or-less happily for the last 130 years, spreading wealth and raising incomes dramatically. But should this process go dramatically into reverse in the future then Marx's solution might be the right idea after all.

Free-market capitalism relies on competition among agents to generate the best prices and direct consumption and investment to best deal with the problems of economic scarcity. In all western societies this private, competitive mode of operation is balanced with democratic state control in areas where free-market provision can't cope. But in the robots-steal-all-the-jobs scenario competition and scarcity have rapidly vanished. Software putting everyone out of work would imply a massive increase in productivity had been achieved, as machines do the jobs cheaper than humans could so there would be little danger in damaging further economic growth. Economic growth doesn't matter so much if everyone has plenty anyway, while the democratic argument, that we can't allow the economy to be dominated by a handful of Mark Zuckerbergs while everyone else sinks into poverty, becomes far stronger as capital is concentrated in fewer and fewer hands.

At this point we should just nationalise companies as they reach a certain size and stability and effective competition recedes (Google, Facebook, Microsoft, anyone?). This could potentially even just take the form of nationalising certain patents or pieces of intellectual property, in terms of software. The original owners should be reasonably compensated but control should  pass into democratic hands, cutting out the fat-cats, and ensuring general control of the economy remains with the wide public. The economy would generally remain private, and competition would remain because nationalisation would be piecemeal, rather than by whole economic sectors (e.g. the Steel Industry). Businesses would remain managed much as before, allowing employees to research and develop improvements to products, and the state could allow them to shrink and go bust if necessary, because there would be a pool of further businesses or inventions to be nationalised. This should largely avoid the problems faced by classical post-war 'mixed' economies. The main difference would be that once a company or piece of intellectual property had been sufficiently developed, instead of owners being able to just sit back and live off a stream of profits perpetually, they would be effectively bought out by the state in a one-off, reasonable payment.

The majority of people would enjoy comfortable existences doing some limited work but largely supported by a vast army of robots and AI. Thus we would fulfil Marx's dream and eventually, as technology developed further, glide gently beyond Socialism to a state of prosperous democratic control, communism in the sense Marx originally meant it. Bliss.

Or alternatively the soothsayers may be wrong. Mass unemployment may not rapidly spread over the next few decades due to a shortage of work to do, and ownership of capital may not become rapidly concentrated among a plutocratic elite. If that happens we should not embrace Socialism and should just carry on more or less as we do now. Sorry.

2 comments:

Kenny Cargill said...

The simple answer is that I doubt robots and AI will ever be good enough to completely replace workers, particularly white-collar professionals. I know skepticism is the easy way out, but I think it is warranted in this case. I am a professional translator, so I am familiar with all of the doom-mongering among my colleagues who fear being completely replaced by machine translation, for example. But natural language in so many places has simply too much built-in ambiguity for a machine to be able to come up with the right answer all the time. There are sentences that even humans can misinterpret, or interpret two ways, unless there is additional context. Therefore, with these sentences it is a priori impossible for the machine to know the right translation.

Machine learning depends on Bayesian analysis and other probabilistic reasoning to learn the right answers. Yet there are always bound to be situations where the probable answer is not the right one. Hence the need for the professional who can dig deeper and reason in order to come up with an original and unexpected solution the machine would never be able to think of.

Stephen Wigmore said...

Actually I agree with you. For reasons much like the ones you give I think the fear of robots rendering everyone unemployed is grossly overstated. Of course mechanisation will continue to replace certain jobs but it will create others, just as it has in vast numbers since the 18th Century.

But if it were to occur I think my point holds. . .

Post a comment