Politics aside, and perhaps not then, the big news of the past few years has been the rapid maturing of artificial intelligence. This has long been inevitable, and it is now reshaping our lives. I think it a glorious leap into the future. Others, of course, see it as a calamity. Here is a story from last week’s Guardian: the supermarket Ocado is cutting five hundred jobs as AI systems take over roles once held by human workers. This is only one example of a broader trend, and I can understand individual fears. If you are an accountant, or a conveyancing solicitor, or a copywriter, or a graphic designer; or if you analyse the movement of share prices or if you interpret medical test results – if, that is, you have earned or are looking forward to earning a nice middle class income, you are being replaced.

I can understand these fears. I do, even so, take pleasure in these fears at the individual level. The occupations now endangered are filled with people who cared nothing when working class jobs were mechanised or exported to the Far East, who swallowed every ruling class lie about how diversity makes us strong, and who voted obediently for politicians who are further destroying a country already in the final lap of its history. Artificial intelligence machines do not have the braying, self-congratulatory voices of the English middle class. They do not vote Labour. It helps, I will admit, if I contemplate the future from an occupation that is unlikely in the short term to be made redundant by artificial intelligence. That does greatly assist my finding of comfort in the suffering of others:

suave, mari magno turbantibus aequora ventis,
e terra magnum alterius spectare laborem

But enough of this, and enough of individual fears. When an established occupation is swept away by technological progress, those who have the intelligence and enterprise to succeed somewhere else will go forward to succeed. The rest can push trolleys round a Tesco car park, and I shall walk past them. What, though, if we extrapolate from that story in The Guardian. What if AI should continue to replace workers? Arguments about how mechanised looms only increased overall employment in the textile industry are brushed aside. This time, we are told, it is different: we live at the start of an age either of mass-impoverishment, or where people will live as drones under the thumb of some omnipotent and omnipresent Big Sister.

I do not believe that either outcome is likely – rather, I do worry about the second; but I see no economic forces that will bring about the first. I will begin with what I hope is a clear explanation of the common fear.

Let us imagine a closed economy, Plutonia, with only one employer: Megacorp plc. There is no saving or investment, only production and consumption. Every week, Megacorp pays out £100 in wages, and collects the same amount in sales of a single product called “consumer goods”. Let us further assume that Megacorp is magically given an AI automated system and replaces all human labour with robots. In the first post-automation week, Megacorp still makes £100 in sales, but no longer pays wages. By week two, consumers have no money to spend. Because of this, Megacorp ceases production. The result is that the economy collapses, and everyone is at risk of starvation.

Though not a realistic model, this illustrates the fears that surround the quickening use of AI automation in our lives. I will not deny that automation will come with problems. But the most basic fears it has raised lie on a foundation of bad economic analysis.

To see this, let us introduce reality to the model. Let us now assume that many, though not all, firms in a country completely automate, so that they no longer need ordinary labour. Let us further assume that demand for their output is inelastic. The result of this will be a large fall in production costs, followed by the usual fall of prices of output in the direction of average cost. This being so, the cheapening of prices for some goods and services will open demand for other goods and services, and therefore of labour. Many of these new good and services will be unexpected. It is a natural failure of our imagination that we do not know what these will be. But, looking to the history of the past quarter millennium, this has happened again and again. There was a time in England when more than ninety per cent of the population had to work on the land just to keep a population of a few million from starving. Today, a greatly enlarged population can be fed very well, even without imports, by an agricultural workforce of less than five per cent. Later, as already noted, the mechanisation of weaving reduced the demand for certain kinds of labour in the textile industry. This did not cause mass-unemployment, but instead shifted the demand for labour to other areas within the textile industry. It has been the same in living memory with automation of factories and the introduction of digital technology into the service sectors. We do not know what new goods and services will emerge. That is because we cannot know how people will want to spend the extra purchasing power they gain from lower prices of existing goods and services. Even so, we should know by now that the future is always a surprise. Where technology is concerned, it has always so far been a pleasant surprise.

But let us continue by assuming that AI automation will progress to the point where it has replaced all but a few essential human jobs. Now, at this point, the prices of goods will reflect only the rents and interest and raw material element of their production costs – and there is no reason why these should not themselves fall, since their present administration and extraction costs have a large element of wages paid to human workers. In the same way, the prices of most services will approach zero. At this point, a displaced industrial or professional worker pushing trolleys in that Tesco car park might be able to afford a better real standard of living than before the AI revolution.

One objection here is that automation will replace human labour even in places like supermarket car parks. What if there is no demand at all for human labour? What if there is no means whatever of working for a living? That may happen – but even this does not mean universal poverty. The fact that any production of goods and services may continue will be sure evidence that the producers themselves want something in return. If they cannot have money, they will need to settle for the applause of flattery of their customers. This may seem an odd state of affairs, but we already see something like it in the business model of Google. This provides a whole range of useful and even essential services: the only payment required from us is our attention. Imagine a world where everything was free at the point of use, but carried paid advertising from religious or ethical organisations. Would this be a terrible place? It would take us into an age of near-universal abundance, where the only real “work” left for humans was intellectual and social interaction. It would allow us to live in a world approximating to the Greek and Roman ideal of dignified leisure. Instead of miserable slaves to do the work, there would be various kinds of machine.

The further objection here is that the majority of people are not morally and intellectually fitted to live like gentlemen. Without jobs to give them purpose, they might vegetate in front of their big televisions, or turn to more actively vicious habits. Our experience of state welfare has not been encouraging. When large numbers of the unintelligent receive something for nothing, the result is often a collapse of discipline and ambition. A world where no one needs a job as presently understood – where, indeed, no one needs to think about most of the things that now take up the thoughts of moderately intelligent people – might be a world of human vegetation. These may at present be valid concerns. But our science and technology are moving forward at wonderful speed in all directions. Give a few generations of parent-driven genetic engineering, and everyone should be up to exercising and chatting all morning in the gymnasium, hunting in the afternoon, and string quartets in the evening – that or whatever else rich and intelligent people fancy.

Now, so far, I have assumed a continuation of our present corporatised economy, which has a clear distinction between organisations that produce goods and services and individuals who sell their labour to these organisations. Even if there is no more sale of labour, the automation of work will have clear benefits. But this is not the only model of economic organisation. I do not think it a particularly desirable model. There are alternatives. We can imagine a world of independent households, where people use AI automation to satisfy virtually all their wants. If there is exchange between households, it will be because some choose to specialise in certain kinds of production and have a surplus. Just because production costs may head towards zero, some opportunity cost will remain, and differences here will encourage continued exchange.

I have seen two objections here. The first is that AI technology and general automation technology is very expensive, and will be too expensive for ordinary people. However, I see no reason why this should be so for much longer.  On the one hand, the price of capital goods is determined by the price of what they help to produce. When the price of final goods is heading towards zero, capital goods are unlikely to be that expensive. On the other hand, modern capital goods are like personal computers. They evolve rapidly, and lose most of their value long before they become useless. You can buy a mobile telephone from 2019 for pennies in the pound of its original price. Fit a new battery, and it works as well now as five years ago, when it was considered a technical marvel. I have no doubt, I could ask on Facebook, and someone within half a mile of where I live would offer me a free notebook computer from about 2015. That also would work just as well as new, when the reviewers fell over each other to praise its new features. Or I will say that, a few years ago, I paid £20 for a 40 inch television from 2007. It cost over £1000 new, and worked very well until last year, when I replaced it with something new and bigger that Samsung had offered me at a steep discount. We can easily imagine a world of independent households, mostly running almost free but still useful electrojunk from the recent past.

The second objection, I suggest, is worthless as soon as stated. This is that AI automation, used as I suspect it will be, will bring about a shrivelling of markets, and that this is somehow bad for freedom. But this really is a worthless argument. Markets, as they have developed since the end of the last Ice Age, have been the best available answer to the problem of scarcity. Take away scarcity, and the need for extended market activity will at least diminish. Markets, though, are just one specialised example of interaction between free people. Interaction in itself will continue for as long as free people wish to mingle with each other.

How we get from where we are to a future of universal abundance depends on political and social structures. We may decorporatise, so that independent households quickly emerge that can take advantage of quickening technological change. Or we may go through a corporatist-social democratic phase, in which the unemployed are given a universal basic income. Or we may proceed to the kind of total state provision described in Huxley’s Brave New World. I have my own preferred model for the future. But I see no reason why universal automation should result in mass-starvation.

A final worry is that we are heading towards “the Singularity” – where AI becomes so advanced that it will take on its own consciousness. It may then decide that we are a danger to it, and will help us to destroy ourselves. We may find ourselves not in a world of universal abundance, but in the world of the first Terminator film. This is a possibility. It is a possibility even if we view consciousness as a spiritual thing that cannot be reproduced by adding more microscopic etchings on a piece of silicon. We only need to programme a computer to respond as if it were conscious, and our destruction may follow from that. Again, though, I am optimistic. These fears of a Terminator future rest on the assumption of a continued separation of ourselves from the technology that we use – that we shall continue making machines of increasing sophistication, while remaining unchanged ourselves. I think it more realistic to imagine that we shall soon start enhancing ourselves. We are already looking towards the direct manufacture of replacement of organs and limbs. Why not look forward to brain implants that expand our learning and processing abilities? At the moment, even with my constant attentions, it can take several years to learn Greek well enough to read Homer with confidence. It will be a glorious thing to upload a programme to our brains that cuts out all the preliminary work of memorising and internalising. More practical may be the ability to read and comprehend and make use of a thousand words a second. AI will not replace us if it becomes part of us.

Speaking personally, I have now reached an age where death is not far in the future, and where, before that, the good health I have always enjoyed will no longer be something to be taken for granted. These facts have always so far been inevitable – rather as putting up with bad eyes or an irregular heart rhythm were once inevitable. We now live at the start of an age where was has been inevitable will not remain inevitable. I feel sorry for those people who have lost their jobs at Ocado. But I do suggest that these are local costs to a set of changes that will bring universal benefits.