AI, Automation, Globalisation, and the Inevitability of Socialism

There is a long and honourable literary tradition of dystopian visions based on and around the idea of mechanical and/or electronic usurpation of the role of human labour and the resultant impact on social stratification:

These are merely the best-known, perhaps most influential, instances of futuristic dystopianism. My own particular favourite, however, is Kurt Vonnegut’s 1952 novel Player Piano. Vonnegut homes in on the central problem of all-pervasive mechanised automation, that of the fate of the great majority of the population whose labour is no longer needed. Player Piano suggests that this human surplus finds employment either in the military (including its civilian equivalent, the police) or a massive, scarcely productive, labour force known as the ‘Reeks and Wrecks’, formally the ‘Reconstruction and Reclamation Corps’. This ‘solution’ is compelling because it describes what is likely to happen as a consequence of a slow, organic progress towards the ascendancy of AI, Automation, Globalisation (AIAG) in a capitalist or mixed economy system. How does such an evolution occur?

• At least one distinguished economist, Robert Skidelsky, envisages the evolutionary process becoming stuck at the point of AI’s ascendancy: ‘Science fiction has raced ahead of economic analysis to imagine a future in which a tiny minority of rich rentiers enjoy the almost unlimited services of a minimally-paid majority.’
• There is an inherent contradiction between the Surplus Labour Group’s (SLG) role as wage-earning consumers and its unemployment status. The SLG needs to have enough earning power to buy the outputs of the AIAG process. As AIAG accelerates so this problem becomes more acute.
• For a while, globalisation disguises the nature of the problem. Moving production to countries where wage costs are lower has a number of effects: competitively reducing wage rates in wealthier countries; creating a process of low wage ‘virtual employment’ in those countries; focusing attention on job losses to lower income economies; accelerating the drive towards automation in order to remove the transport costs involved in overseas manufacturing (especially where this has, in turn, been automated). In effect, globalisation creates a major distraction from the economic effects of automation and AI in wealthier economies
• A gradual process of replacing productive work with unproductive work. David Graeber’s polemic against ‘Bullshit Jobs’, painfully reminiscent of Vonnegut’s Reeks and Wrecks, describes this perfectly: “a form of paid employment that is so completely pointless, unnecessary, or pernicious that even the employee cannot justify its existence even though, as part of the conditions of employment, the employee feels obliged to pretend that this is not the case.” Chillingly, however, an examination of Graeber’s classification of bullshit jobs reveals a list most of which are acutely vulnerable to the advent of the smart systems that AI promises. They may be waged now in their administrative black holes, but these jobs are all candidates for extinction , every bit as much as the assembly line tasks taken over by robots or the typing pools rendered redundant by word processing.

As worthwhile, productive, adequately waged employment becomes less and less available, the system searches for a quick fix. Cometh the hour, cometh the concept: the Universal Basic Income is now firmly on the horizon. The idea is being investigated and discussed on a global basis, and even forms part of the proposed programme of an American presidential candidate. There is little doubt that, benignly implemented, UBI has potential benefits, but in practice, what is to stop it becoming the subsistence wage of a growing army of Reeks and Wrecks? Some commentators, exercised by ‘the future of work’ considerations, have embraced the UBI idea as facilitating a utopian vision of productive and rewarding leisure for the SLG as the fruit of AIAG, but there are some compelling reasons why this utopian vision is unlikely to materialise:

• We can’t go from where we are to a comparatively work-free utopia in a single step, and the intervening stages will throw up some seriously destabilising situations. Those still in work, the ‘lucky’ ones with non-bullshit fulfilling jobs, will see themselves as higher income earners paying tax to subsidise the lifestyles of a growing army of indolent ‘layabouts’. The potential for social conflict is enormous. This is especially so when there are attempts to raise UBI above the subsistence level of its initial implementation.
• It will be difficult to overcome social conditioning whereby self-esteem and social status is defined by the work we undertake. The question “What do you do?” – a commonplace conversation-opener between strangers in any social interaction – is indicative of how deeply embedded this conditioning has become.
• Our entire education system is based on the assumption of preparing people for work. Without work for the majority the required adjustments in the system will be difficult to identify and even more difficult to implement. Around the crossover point between work and leisure for the majority, there will arise the issue of motivation. How will schools and parents persuade pupils to engage with education that is decoupled from material security?

And then there is the issue of surveillance. AIAG isn’t just about automation of more and more modes of work, reaching from routine industrial tasks into the professional domains. Digital technology now enables the collection of vast quantities of data, while AI algorithms become ever more adept at manipulating that data to effect systems of social control. John Harris has characterized this process as a ‘battle for control of the Internet’, contrasting the Chinese system of generating ‘social credits’ (a form of compliant digital citizenship) with the western trend towards a new version of capitalism wherein data and the ability to gather it becomes the most valuable commodity, while the AI manipulation of that data through monetization becomes the greatest source of wealth. Are there any escape routes from surveillance? Outside China the best hope would seem to be through regulation, and the GDPR indicates a first step in putting up barriers to data collection and exploitation. In the Chinese model there is no hope for regulation because the state is the culprit and the state makes the rules. Here the only way out is through anonymisation – the capability of the individual to operate outside the identifying protocols of the networks. In either case there is a race between the inexorable progress of the the technology and the ability of the individual to evade it or be protected from its abuse. At present it feels like a runaway win for the technology, whether through the omnipotent ubiquity of 5G or the simple loss of social credits simply for having one’s mobile phone switched off. Regulators tend to be disastrously slow at catching up with the technology, and, where the state controls the networks, rapid technological progress works against the interests of the individual who has to run ever faster in order to stand still in the quest for anonymity.

AIAG and Socialism

All of the dystopian visions I have thus far mentioned tend to avoid, or only very marginally address, the issue of ownership. However, the question of who actually owns the tools and processes of AIAG is of critical importance. It is in addressing this question that I believe the most promising solutions may be found. If AIAG is a potential means of controlling the populace, then control of AIAG is the only means that the populace has of redressing the balance, and I take it as axiomatic that control can only be achieved through ownership. But the history and mechanics of ownership is in itself highly problematic. The best illustration of the problem is obtained through a close examination of the history of the Labour Party’s controversial Clause IV:

• In 1918 Clause IV became part of the constitution of the UK Labour Party, calling for ‘common ownership of the means of production, distribution and exchange’. This was seen as a defining commitment to the principles of socialism. Historically, this commitment ran in parallel with the Russian revolution and its subsequent implementation of rigid state ownership and control of the means of production, distribution and exchange. Critics of Labour’s commitment to socialism were able to identify it with all the faults of the Communist system, not least its processes of political and social repression.
• This identification was facilitated by Labour’s inability to imagine any other means of achieving Clause IV’s objectives than through nationalisation – full and complete ownership and control by the state. This failure of imagination was instantiated by the post WW2 Labour government’s programme of nationalisation.
• Gradually, Clause IV’s ‘ownership of the means of production’ phrase came to be seen as a back-door route to Sovietisation: by right wing critics who used it against Labour on a McCarthyite ‘reds under the bed’ basis; and by entryist left wing groups who saw the Labour Party as a suitable ‘Trojan Horse’ on the road to power.
• Labour leader Hugh Gaitskell tried and failed to revise Clause IV in 1959, and it was not until 1995 that Tony Blair succeeded in implementing a revision that made no mention of ownership of the means of production. Blair saw this as an essential step in making ‘New Labour’ electable.
• In 2018, marking the centenary of the original Clause IV, a campaign was launched to reinstate its commitment to common ownership of the means of production, distribution and exchange. Long on (rather antique) rhetoric and short on analysis, the campaign saw nationalisation as the only available response to a perceived ‘crisis of capitalism’.

So it would appear that in a hundred years the left has come full circle on the question of ownership of the means of production. All of this without ever properly considering the use of the capitalist system itself, its markets in shares, to achieve ownership on a gradual, cumulative basis. Historically, this would have been an alternative strategy, a matter of choice if had been considered. But now, with AIAG upon us, the issue takes on an entirely different urgency. The rate of advance of AIAG is so rapid that its ascendancy is irreversible – relative competitiveness lends it an air of inevitability in a globalised economic system. At the same time it is equally inevitable that the resultant social stresses will be politically seismic, and this is where I believe that socialism becomes inevitable – not the socialism of nationalisation or of ‘organically’ in step with AIAG’s advance. I have already outlined the institutions and mechanisms required to achieve this. They are neither new nor original, and are of proven efficacy. All that is required is the political vision and will to bring about their implementation. To be clear, this is not a form of socialism that threatens to supplant or replace existing capitalist processes, it is a subtle, gradual adjustment that complements them and evolves in parallel with the changing reality of capitalism’s implementation. I see much of its inevitability stemming from the fact that it promises to finance UBI without having to deal with the resentment arising from taxation of those privileged enough to remain in productive, more rewarding employment. If there is a key role for politicians and media in preparing electorates for this new reality, then it rests in challenging the traditional thinking of ‘right’ and ‘left’ in the running of modern states.

At first sight it might not seem obvious where globalisation fits into this. The answer is that technology will make the location of physical production an irrelevance. 3d printing promises to make it possible to create any product in any location. When this happens, transport becomes a much more important factor in the cost profile and localised production offers distinct advantages. 3D printing will therefore be highly disruptive of previous production patterns. The ‘means of production’ becomes an entirely different thing. Attention focuses, probably for historical reasons of habit, on the physical plant, the printers. But value is more likely to reside in the software (templates, designs, code) that determines the output of the plant. Once this is properly understood, the way is opened to an entirely different economic pattern, one that combines public and private ownership:

• Plant, that is to say the 3d printing equipment itself, can be widely disseminated on a publicly owned basis. Think of public libraries as an operating model. Access to the physical production process becomes a public good, freely available.
• Software, the digital templates that determine productive output, is privately owned, developed and licensed on an entirely entrepreneurial basis.
• Investment in the supporting infrastructure (raw materials, distribution etc.) required to make 3D printing work can be supported by public funding wherever and whenever the private sector proves inadequate.

None of all this will happen instantly or overnight. It is the gradual nature of AIAG’s ascendancy that makes growing public ownership inevitable. At every stage there are key political choices: either assert a degree of public ownership and control (if only to pay for ameliorating policies like UBI), or move closer to a dystopian future where increasing democratic majorities are left out of the loop of wealth creation and distribution. The form of socialism that results will be unlike any envisaged to date. Its defining feature will be a progressive harnessing of the processes and mechanisms of capitalism rather than a rejection of them on ideological grounds.