Robotisation is probably going to be a temporary phenomenon: planetary limits will (within a generation or at most two) severely limit the supplies of raw materials and energy needed to enable large-scale robotisation, and pollution-crises – part-speeded-up by huge investments in automation/robotisation – will have the same effect. The question is whether we can rein in robotisation soon enough to ensure that ‘Peak Robot’ occurs under our control, and not as a result of a crash forced on us by collapsing ecosystems.

The delusion of endless growth

In the heady days of early industrialization, when infinite economic growth was the unchallenged dogma that promised to dispel the twin evils of inequality and poverty, a dangerous fantasy began to take place in the minds of our most celebrated economists. The idea was that as production refined its efficiency through technology, we would inevitably reach a state of ever more widespread industrial automation. To this day, ‘growthist’ capitalism and socialism still tend to argue that the refinement of technology promises us diminished working hours and ever increasing consumption, leading (allegedly) to ever higher standards of material wellbeing.

It is perhaps no wonder that the ‘left’ has been as keen to embrace this catastrophically deluded daydream as the ‘right’ – for both are simply versions of economism (the delusion that the economy is all that matters in society) and productivism (the delusion that what society should be aiming to do is to maximise economic ‘production’). Considering that economic growth has so far failed to deliver on fewer-working days as promised, and that the fantasy that it will do so has largely been an excuse for withholding adequate remuneration and redistribution, now is an important moment to question whether robotisation itself can deliver on the promises so often made by its proponents.

Fully-automated luxury communism?

As my title suggests, I am as yet thoroughly unconvinced that robotisation is or indeed ever will become as automated as the fantasy of the complete leisure society would demand. Undoubtedly technology has allowed us to automate many jobs, and increase efficiency in others, which could free up more time for leisure if its fruits were fairly dispersed. However, there is a vital and unaddressed constraint that may well mean that we will be at ‘Peak Robot’ sooner than one thinks. This constraint is the bread and butter of my think tank, Green House (Green House Think Tank): it is the set of planetary boundaries that are popularly known as ‘the limits to growth’. As we in Green House maintain and have detailed, not only is it the case that economic growth is no longer necessary (for there is enough stuff to go around already, if only we shared it out better), it is also no longer desirable (for it isn’t making us happier; and it is destroying our collective future by gradually eroding the shared basis of our planetary home). Moreover, it is now only even possible at heavy cost: i.e. it is possible only if one is prepared to create on balance a less desirable future. And before too long, it won’t be possible at all.

What this means for the daydream of full automation is potentially profound. For the ‘march of the robots’ idea relies tacitly on the assumption that the limits to growth are negotiable, or indeed non-existent. It buys into the idea that there can be a complete – or at least near complete – decoupling of production from carbon emissions; despite the fact that, as Tim Jackson and others have shown, there is no evidence that any such decoupling is likely to happen or is indeed even possible. It should not surprise that this is the case; after all, what would ever more automation even look like if it did not include ever more consumption and environmental degradation?

Robotisation depends upon a ready supply of metals and plastics; on a ready supply of concentrated energy, permanently. And on much more. Robots will also without any shadow of a doubt produce far more GHG emissions than their blue-collar counterparts, to operate and to remain operable; for robots are not alive: they need repair, maintenance, and monitoring, and constantly risk being broken, superseded, etc. And when the robot ‘dies’, it will need to be fully recyclable –  which is a very big ask indeed, and again will be highly energy-consumptive if it is to be anything-like achieved. (Recycling always requires energy-loss. This is a basic result of the profound limit imposed by entropy.)

In a world which takes seriously the limits to growth that are constituted by the steadily advancing threat of global warming, then there will be ever-rising constraints on the rise of the robots. (‘Incidentally’, if we do not soon become such a world, then the whole discussion will become academic, as civilisation collapses). And in a world which takes seriously those limits to growth that are constituted by limits of resource depletion (fuel supplies, rare earths, etc.) then again there will be yet further ever-rising constraints on the rise of robots. While we are not yet in such a world quite as much as the original Club of Rome projected, the moment of truth here has only been slightly delayed, not overcome (See on this the important research recently conducted by the new All-Party-Parliamentary-Group on Limits to Growth: Limits To Growth). Given these twin constraints, whether one looks at the front or back end of the pipe, the limits to growth will place severe limits on the march of the robots.

Don’t overdose on sci-fi

If one reads too many sci-fi books, one imagines that it will be centuries or millennia (or never) until ‘Peak Robot’: One then fantasises, as Dr. Who and Star Trek and many many more do, routinely, a cornucopian tech-based future. Or, at worst, a Battlestar Galactica style robot-apocalypse… But such scenarios, which have for instance exercised Saint Stephen Hawking recently, are of course premised on assuming, utterly rashly, that the limits to growth can be gainsaid. The clever people spending their time worrying about the robot-apocalypse would probably be better employed worrying about more ‘down-to-Earth’ threats to our common survival – such as human-triggered runaway climate-change, reckless geo-engineering, reckless genetic engineering, and so forth.

If I were forced to bet, I would bet that, on an optimistic scenario for our future, Peak Robot will come within 20 years or so from now at the most. For, as a species, we are already living as if we have more than one planet. How are we going to rein that in in time, and survive (let alone flourish), if we pour more and more resources into producing fragile metal people and energy-hungry super-computers to replace low-skilled workers? The Club of Rome scenarios suggest that, unless we rein in impacts within a generation or so, then it will be too late to avert a crash. A better, Greener route will of course be to reduce our impact and footprint pro-actively, to live more locally etc., to share more – and then we won’t need very many of either shelf-stackers or of robots, and all without causing poverty.

On a pessimistic scenario for our future, I would guess at Peak Robot occurring within 50 years or so. Because by then, if we haven’t massively changed our way of life, climate chaos and others of our ongoing or incipient exceedings of planetary boundaries will have become so gross that the complex systems that robotisation relies on will be collapsing. In the long run, as the doyen of ecological economics Herman Daly, has shown us in his vital books, human labour is going to be in greater supply than ‘raw materials’, and also usable (low-entropy) energy. The extraordinary thing about people is, of course, that we are a truly renewable ‘resource’.

Conclusion: The ultimate resource

The real challenge for the future, then, is to use our human ingenuity – what some have called ‘the ultimate resource’ – to navigate boldly and precautiously the difficulties presented by this time, a time when something worth calling ‘a leisure society’ is possible, if we are willing to take our power back from the 1%, and ensure that the benefits of what automation is ecologically viable are fairly shared. Though of course it follows from what I have argued here that in any viable society there is always going to be an essential and substantial role for human work – especially once we take seriously that robots are not going to replace us for long without hastening our (and their) collective demise. The leisure society is going to be one in which we aren’t working excessively to make more and more money for big corporates. But it isn’t going to be a society in which we are laying around being served by robots either. It is going to involve us doing a great deal of what we want to be doing: including such things as growing food.

The leisure society is not going to be achieved through endless ramping up of our desires/ satisfaction of those desires by robots. It is going to be achieved, if at all, by our reducing our desires. Limiting our selfish wants. Cleaving rather to our needs, and to each other, and to our fragile home. ‘The leisure society’ in its true sense will be one in which we still engage in lots of activity, but less of it in return for money, and more of it just because we actually want to do it. (A society in which growing much of one’s own food, for example, is not considered a chore. There are already many of us – would-be smallholders, for instance – who are in this mental space.)

My Green House colleague Molly Scott Cato and I have a favourite quote, from John Ruskin: “There is no wealth but life“. In the long run, the amazing capacity of human and non-human life to organise ecological webs of activity with relatively low entropy and – potentially – with high wellbeing, is going largely to trump the novel fad that is robotisation. Unless we fall fully afoul of the limits to growth, in which case, we’ll all be trumped.

I would like to propose then that we disavow ourselves of our misplaced trust in robotisation, and instead focus our (renewable) energies, as I’ve argued previously here at the Ecologist, (Lets Build A Post-growth Economy That Works) on creating a truly fair, sustainable and post-growth economy – an economy for people.

A final thought: when I put these contrarian ideas to robotisation-fans, they often reply by claiming to me something like the following: “But robots are going to be more energy- and resource-efficient, overall; think of robot-workers who don’t need holidays or sleep.” I reply by pointing out what everyone seems to miss: that, if this scenario is actually going to generate a net-efficiency, then it can only be at the cost of eliminating a significant chunk of the (human) population altogether. Because, if the human population remains the same, and all those humans are still eating and holidaying and so on, then adding a population of robots on top of them certainly does not lead to an overall efficiency: i.e. it obviously leads to more of the really scarce resources (e.g. non-renewable materials and energy) being used up. So, bear this carefully in mind: when the claim is made that robots will be better overall, ecologically speaking, what the speaker has in mind, either consciously or unconsciously, is a society where robots are not merely replacing human labour; they are replacing some humans altogether.

Perhaps the fans of robotisation will be kind enough to tell us, before they sketch their next utopia, just which humans they propose to cut out.

This Author

Rupert Read is philosopher working and writing at UEA, the Chair of Green House think tank, and a former parliamentary candidate for the Green Party of England & Wales. He Tweets at: @GreenRupertRead & @rupertread

[Thanks also to Atus Russell and Dave Burnham for help with preparing this article, which had its origins in a much shorter piece published in Green World.]