As increasing numbers of jobs look to be under threat from automation, the time has come to think about the distribution of wealth in a world without work. In the first part of this series on the future of humanity and technology, we make the case for the widespread private ownership of software and robotics – and the new institutions we’ll need to help us achieve this.
The ancient board game, Go, is widely considered to be the most complex two-player strategy game in existence. With more possible Go-board positions than there are atoms in the visible universe, players who hope to win cannot rely on their quantitative analytical abilities alone. Instead, the world’s best Go players approach the game with a more creative disposition, employing their intuition and qualitative judgment. In many cases, professional Go players will say that a given move simply “felt right” – and in this sense, it’s a very human game.
The fact, then, that Lee Sedol, a 9 dan professional Go player, was beaten by Google’s AlphaGo should be met with awe and some disquiet. While computers have, for some time, been able to beat Chess grandmasters through virtue of their immense processing power, cracking Go was an altogether different prospect. As a game that is a much creative as it is analytical, Go was thought to be immune type of “brute force” processing AI that laid the groundwork for Gary Kasparov’s watershed loss to Deep Blue in 1997. Crucially, this meant that in order to master the ancient game, Google had to create a computer capable of replicating those most human of attributes: intuition and understanding.
In the ongoing, and no-longer lopsided, intellectual battle between humans and computers, AlphaGo’s victory over Sedol constituted a breakthrough moment. Google’s Deep Mind, the architects of AlphaGo, have now comprehensively demonstrated that computers can surpass humans in activities that were previously thought to be uniquely suited to our organic minds. An inescapable consequence of this is that there might actually be very few activities that cannot be automated and performed by computers.
The robots are coming, and nobody’s safe.
The triumph of AlphaGo should serve as a wake-up call to professionals who see themselves as unlikely to be impacted by the coming wave of job automation. In time, professions like medicine and law will be radically transformed by computer programs that excel at pattern-recognition and decision-making. Indeed, the question of whether professionals and information-workers will ever be displaced by technology has itself been displaced by questions about timing. Those who are informed on the topic now talk almost exclusively about when, rather than if, such a process of mass-automation in the professional labour market will take place.
Regrettably, the conversation surrounding automation continues to be obscured by our collective failure to distinguish between the economic landscape of yesterday and that of tomorrow. Even in the face of advancements like IBM’s Watson (a computer system which is already replacing workers in the accounting and insurance industries, and has its sights set on the medical and financial sectors) most of us continue to consider automation as something akin to globalisation. In other words, we tend to think of automation as a phenomenon that will exclusively affect individuals from a particular socio-economic bracket – one that has already been ravaged by the wave of outsourcing that came with globalisation. This is economic ignorance at its most destructive.
The uncomfortable truth is that there is almost no correlation between one’s socio-economic status and their vulnerability to displacement. All that really matters is the extent to which one’s chosen occupation involves routine. To provide a real-world example, an expert radiologist, trained to excel in pattern recognition, is far more susceptible to automation than a PA who is charged with handling a wide range of (often very simple) tasks as and when they arise.
Our reluctance to recognise the scale of the coming wave of job automation has left us ill-prepared to confront it. The conspicuous public policy vacuum is gradually being filled by one idea that is as simple as it is popular. This, of course, is the concept of the Universal Basic Income. In short, UBI is a scheme that would involve allocating every citizen a basic stipend capable of providing for some form of existence. But there’s nothing to suggest that it would provide for a life.
While UBI certainly has its merits, the proposal has the feel of an economic vaccine that one would happily prescribe to others, but not necessarily to themselves. Moreover, it constitutes an admission that in order to survive, many of us will be forced to rely on the goodwill and generosity of our government – and consequently that of our fellow citizens. But just as importantly, it involves an acceptance that many, if not most, of us will be forced into a substantially reduced standard of living from which there will be no escape.
If the “utopia” prophesised by the UBI’s champions ever came to fruition, the new economic reality would be unforgiving. Millions would exist in a state of total dependency on a government that would itself be firmly under the control of a handful of tech billionaires. In such a scenario, the eye-watering levels of financial inequality would only be reinforced by the citizenry’s near-total loss of political power.
But notwithstanding UBI’s more obvious limitations, the concept is really most deficient in its inherent pessimism. As an idea, it treats automation as an existential threat to humanity – a problem to be overcome, not an opportunity to be seized. Indeed, the economic hypothesis that underpins UBI is one that equates the displacement of human labour with the termination of free market capitalism itself. Ultimately, it concedes that there is no place for humans in a market economy beyond the performance of traditional labour-intensive occupations.
Thankfully, however, the pessimism of UBI’s supporters is merely a product of their lack of imagination. When we think beyond the parameters of the state, precious opportunities begin appear. Moe concretely, if we are to find a way through the automation problem that allows individuals to retain – and enhance – their economic independence, our public policy response needs to revolve not around the distribution of money, but around the distribution of assets. In plainer terms, our efforts should be directed toward ensuring that as many people as possible are able to own the technology that replaces them.
In order to achieve this we will need to establish a set of new institutions centred around ensuring that the victims of automation become the beneficiaries of automation. These institutions, that could be called Asset Redistribution Tribunals (ARTs), would be charged with managing the transfer of assets (whether they take the form of shares or robots) to employees who face redundancy due to the emergence of new technology. Put simply, ARTs would be responsible for ensuring that anyone who found themselves displaced by technology would have the opportunity to own that technology. And on a more philosophical level, they would be responsible for saving capitalism from itself.
Thus, the real purpose of an ART would be to calculate and administer redundancy programmes that reflect the realities of the New Economy. Each ART would be charged with deciding:
- whether a particular redundancy was the result of automation/computerisation, and
- the level of ownership an employee would need maintain a satisfactory standard of living (via dividend).
In practice, this would mean that if an ART was satisfied that a redundancy was indeed the product of automation, it would then insist on a share issue, or asset transfer, capable of providing the former employee with a level of company stock that could realistically be expected to produce a substantial annual dividend. We can call this arrangement Shareholder Income Replacement (SIR).
Although these institutions would be new and somewhat revolutionary, their objective would not be without precedent. Employee Stock Ownership Plans (ESOP), which allow employees to own a share of the company through a retirement trust, are already a growing feature of the economy. The work of ARTs would be geared toward ensuring that displaced workers were granted the same opportunity as those fortunate enough to benefit from ESOP plans. Ideally, of course, ESOPs would expand organically, and without the oversight of statutory tribunals. But the stakes are simply too high to hope that the market democratises itself without nudges from the state.
Critics of ARTs and SIR might point to the forced redistribution of assets and see an entirely unnecessary abuse of state power. Under normal circumstances, I would be sympathetic to such criticism. However, in this case it is important that we are not naive about the scale of the automation problem, and ignorant of what will happen in the absence of a concerted defence of free market economics. This plan might sound like a strange defence of the free market, and I can understand why. But I hope that if it is viewed in the proper context, it will become clear that the driving ambition is a liberal one. The ART model is about ensuring that displaced labourers are able to re-enter the market as private investors. It is this central ambition that differentiates ART/SIR from UBI. Whereas the latter constitutes an attempt to equalise society and institute an oligarchy, the ART approach seeks to empower individuals by offering economic independence and a new role in the economy.
Yes, the process of Making ART would constitute the single greatest act of economic intervention since the birth of the Soviet Union, but it would ultimately save capitalism from itself, thwart the emerging plutonomy, and grant individuals true economic independence.