by Michael Roberts
The recent opening by Amazon of a new retail store in the basement of its headquarters in Seattle has provoked more talk that human labour is soon to be wiped out by the expansion of robots and AI.
At the store, which is clearly a ‘pilot’, customers walk in, scan
their phones, pick what they want off the shelves and walk out
again. There are no checkouts or cashiers. Instead, customers first
download an app onto their smartphones and then machines in the shop
sense which customer is which and what they are picking off the shelves.
Within a minute or two of the shopper leaving the store, a receipt pops
up on their phone for items they have bought. This development in
‘automatic’ retailing mirrors other automation: in offices, driverless
cars, social care and in decision-making.
So does this mean that humans will soon be totally replaced by intelligent learning machines and algorithms? In previous posts,
I have outlined the forecasts on the number of jobs that will be lost
to robots over the next decade or more. It appears to be huge: and not
just in manual work in factories but also in so-called white-collar work
like journalism, banking and even economists!
The techno-futurists think robots will soon replace humans. But I
think they are running before they can walk – or to be more exact, so
far, robots can hardly run and catch compared to humans. This is
‘Moravec’s paradox, namely that “it is comparatively easy to make
computers exhibit adult-level performance on intelligence tests or
playing games, and difficult or impossible to give them the skills of a
one-year-old when it comes to perception and mobility” (Moravec).
So algorithms can vote on whether to invest or not for hedge funds or
banks, but a robot cannot even hit a tennis ball, let alone beat a club
player. Indeed, robot development is heading more towards ‘cobots’,
which act as an extension of the worker, in factories with the heavy
work and in hospitals and social care for diagnosis. This does not
directly replace the worker.
The mainstream economic debate is whether ‘technology’ will create
more jobs than it destroys. After all, the argument goes, new
technology may get rid of certain jobs (hand loom weavers in the early
19th century) but provide new ones (textile factories).
One thought experiment is that provided by Paul Krugman. In Krugman’s celebrated example,
imagine there are two goods, sausages and bread rolls, which are then
combined one for one to make hot dogs. 120 million workers are divided
equally between the two industries: 60 million producing sausages, the
other 60 million producing rolls, and both taking two days to produce
one unit of output.
Now suppose new technology doubles productivity in bakeries. Fewer
workers are required to make rolls, but this increased productivity will
mean that consumers get 33% more hot dogs. Eventually the economy has
40 million workers making rolls and 80 million making sausages. In the
interim, the transition might lead to unemployment, particularly if
skills are very specific to the baking industry. But in the long run, a
change in relative productivity reallocates rather than destroys
The story of bank tellers vs the cash machine (ATM) is another
example of a technological innovation entirely replacing human labour
for a particular task. Did this led to a massive fall in the number of
bank tellers? Between the 1970s (when American’s first ATM was
installed) and 2010 the number of bank tellers doubled. Reducing
the number of tellers per branch made it cheaper to run a branch, so
banks expanded their branch networks. And the role gradually evolved
away from cash handling and more towards relationship banking.
That’s the optimistic view. But even then, as Marx pointed out with
the rise of machines in the 19th century, the loss of jobs in one sector
and their recreation in another is no seamless process of change.
Marx put it: “The real facts, which are travestied by the
optimism of the economists, are these: the workers, when driven out of
the workshop by the machinery, are thrown onto the labour-market. Their
presence in the labour-market increases the number of labour-powers
which are at the disposal of capitalist exploitation…the effect of
machinery, which has been represented as a compensation for the working
class, is, on the contrary, a most frightful scourge. For the present I
will only say this: workers who have been thrown out of work in a given
branch of industry can no doubt look for employment in another
branch…even if they do find employment, what a miserable prospect they
face! Crippled as they are by the division of labour, these poor devils
are worth so little outside their old trade that they cannot find
admission into any industries except a few inferior and therefore
over-supplied and under-paid branches. Furthermore, every branch of
industry attracts each year a new stream of men, who furnish a
contingent from which to fill up vacancies, and to draw a supply for
expansion. As soon as machinery has set free a part of the workers
employed in a given branch of industry, the reserve men are also
diverted into new channels of employment, and become absorbed in other
branches; meanwhile the original victims, during the period transition,
for the most part starve and perish.” Grundrisse.
And then there is the profitability of technology. Robots will not
be widely applied unless they can deliver more profit for owners and
investors in robotic applications. But more robots and relatively less
human labour will mean relative less value created per unit of capital
invested, because from Marx’s law of value, we know that value (as
incorporated in the sale of production for profit) is only created by
human labour power. And if that declines relatively to means of
production employed, then there is a tendency for profitability to fall.
So the expansion of robots and AI increases the likelihood and magnitude
of profitability crises. So it is very likely that slumps in
capitalist production will intensify as machines increasingly replace
labour. This is the great contradiction of capitalism: increasing the
productivity of labour through more machines reduces the profitability
Mainstream economics either denies the law of value or ignores it. Back in 1898, neo-Ricardian economist Vladimir Dmitriev,
in order to refute Marx’s value theory, presented a hypothetical
economy where machines (robots) did all and there was no human labour.
He argued that as there was still a huge surplus produced without
labour, so Marx’s value theory was wrong.
But Dmitriev’s thought experiment is irrelevant because he and other
mainstream economists do not understand value in the capitalist mode of
production. Value in a commodity for sale is double-sided: there is
physical ‘use value’ in the good or service sold, but there is also
‘exchange-value’ in money and profit that must be realised in the sale.
Without the latter, capitalist production does not take place. And
only labour power creates such value. Machines create no value (profit)
without humans turning machines on. Indeed, Dmitriev’s super abundant
robot only economy would no longer be capitalist because there would be
no profit for individual capitalists.
And here is the great contradiction of capitalism. As machines
replace human labour power, under capitalism, profitability falls even
if the productivity of labour rises (more things and services are
produced). And falling profitability will periodically disrupt
production of individual capitalists because they only employ labour and
machines to make profits. So crises are intensified well before we get
to Dmitriev’s hypothetical robot world.
But what to do, as jobs are lost to robots? Some liberal economists
talk of a ‘robot tax’. But all this would do is slow down automation –
hardly a progressive move in reducing toil. The idea of universal basic
income (UBI) continues to gain traction among economists, both leftist
and mainstream. I have discussed the merits and demerits of UBI before. UBI
is advocated by many neoliberal economic strategists as a way of
replacing the ‘welfare state’ of free health, education and decent
pensions with a basic income. And it is being proposed to keep wages down for those in work.
Any decent level of basic income would be just too costly for
capitalism to afford. And even if UBI were won by workers in struggle,
it would still not solve the issue of who owns the robots and the means
of production in general.
A more exciting alternative, in my view, is the idea of Universal Basic Services
i.e. what are called public goods and services, free at the point of
use. A super-abundant society is by definition one where our needs are
met without toil and exploitation ie a socialist society. But the
transition to such a society can start with devoting socially necessary
labour to the production of basic social needs like education, health,
housing, transport and basic foodstuffs and equipment.
Why use resources to give everybody a basic income to buy these
social needs; why not make them free at the point of use? Instead of
cutting people who are not working off from those that are working with
income handouts, we need to build unity at work through reducing the
hours of labour and expanding (free at use) public services and goods
Of course, this would require the many owning and controlling the
means of production and planning the application of those resources for
social need, not the profit of the few. Robots and AI would then become
part of the technological advance that would make a super-abundant
- AFSCME Local 444 negotiations assesment 1997
- The Internal lives of Revolutionary Organizations
- Socialist Alternative members: Questions and Answers
- Sanders: Our Alternative
- An Invitation to Our Readers
- Facts For Working People Weekly Phone Conferences and Discussions
- Help open The AFL-CIO AIFLD Archives