My Times column is on technology and jobs:
Bill Gates voiced a thought in a speech last week
that is increasingly troubling America’s technical elite — that
technology is about to make many, many people redundant. Advances
in software, he said, will reduce demand for jobs, substituting
robots for drivers, waiters or nurses.
The last time that I was in Silicon Valley I found the
tech-heads fretting about this in direct proportion to their
optimism about technology. That is to say, the more excited they
are that the “singularity” is near — the moment when computers
become so clever at making themselves even cleverer that the
process accelerates to infinity — the more worried they are that
there will be mass unemployment as a result.
The McKinsey Global Institute argued last year that perhaps 40
per cent of jobs in clerical and professional services could be
automated by 2025. All sorts of professions, including accountants
and even actors, should begin to fret. With Google’s driverless car
having had just one (human-error) accident in 200,000 miles, there
is every reason to suspect that taxi drivers are heading for the
same fate as wick trimmers and ice harvesters.
For large stretches of every flight, pilots are already mere
spectators; might an incident such as the Malaysian Airlines
mystery make drone cockpits acceptable? Hal could come to seem more
trustworthy than Dave. (No article about the future is complete
without Arthur C. Clarke allusions — Dave being the astronaut who
has to disconnect Hal, the sentient computer, in 2001: A
Space Odyssey.)
Will there be any jobs left for our children? A new book much
talked about in techie circles, The Second Machine Age: Work, Progress, and
Prosperity in a Time of Brilliant Technologies, by Erik
Brynjolffson and Andrew McAfee, hedges its bets. The two authors
accept that previous scares about technology leading to
unemployment were overdone, but they are worried that this may not
happen for ever. They have seen “one bastion of human uniqueness
after another fall before the inexorable onslaught of innovation”
and think that there may be no human activity immune to
automation.
In the 1700s four in every five workers were employed on a farm.
Thanks to tractors and combine harvesters, only one in fifty still
works in farming, yet more people are at work than ever before. By
1850 the majority of jobs were in manufacturing. Today fewer than
one in seven is. Yet Britain manufactures twice as much stuff by
value as it did 60 years ago. In 1900 vast numbers of women worked
in domestic service and were about to see their mangles and dusters
mechanised. Yet more women have jobs than ever before.
Again and again technology has disrupted old work patterns and
produced more, not less, work — usually at higher wages in more
pleasant surroundings.
The followers of figures such as Ned Ludd, who smashed weaving
looms, and Captain Swing, who smashed threshing machines (and, for
that matter, Arthur Scargill) suffered unemployment and hardship in
the short term but looked back later, or their children did, with
horror at the sort of drudgery from which technology had delivered
them.
Why should this next wave of technology be different? It’s
partly that it is closer to home for the intelligentsia. Unkind
jibe — there’s a sort of frisson running through the chatterati now
that people they actually know might lose their jobs to machines,
rather than the working class. Indeed, the jobs that look safest
from robots are probably at the bottom of the educational heap:
cooks, gardeners, maids. After many years’ work, Berkeley
researchers have built a robot that can fold a towel — it takes 24
minutes.
None the less, it is hard to see where people go next. If we are
reaching the point where robots could do almost anything, what is
there left for people to do? To this I suggest two answers. The
first is that we will think of something. Half the new professions
that are thriving today are so bizarre that nobody could have
predicted their emergence — reflexologist, pet groomer, ethical
hacker, golfball diver. In a world where androids run supermarkets,
you can bet that there’s a niche for a pricey little shop with
friendly salespeople. The more bulk services are automated, the
more we will be willing to pay for the human touch as well.
Automation has made us so much richer than our ancestors, by
cutting the cost (in hours worked) of most of the services that we
desire, that we have been able to afford to employ more and more
people to amuse or pamper us. Most people can afford to eat out,
for example — an unimaginable luxury only a century ago.
If the worst comes to the worst, and the androids take over
absolutely every kind of work, providing all our daily needs so
cheaply and efficiently that we just don’t need people at all, not
even as politicians — why, then what’s the blooming problem? The
point of work is so we can consume, not vice versa. Do not forget
that the poor benefit more than most from automation — as consumers
of ever cheaper goods and services.
Keynes predicted that we would eventually have more stuff than
we needed and would start to ration work down to 15 hours a week.
When you consider that we work far fewer days a year and hours a
week than in his day, and make allowance for the fact that we spend
much longer in education and retirement, we are already there in a
sense. As Arthur C. Clarke put it, “the goal of the future is full
unemployment so we can play”.
In 1700 nearly all of us had to dig the soil from dawn to dusk
or everybody starved (and some did anyway). Technology liberated us
from that precarious and awful world. If it does so again, so that
our grandchildren never have to think in terms of “jobs” at all,
but merely in terms of how they can fill their days fulfilling
their wishes and helping others, mixing bits of work with bits of
leisure, while drawing on the output of Stakhanovite machines for
income, will they envy us our daily commutes and our office
politics? I don’t think so.
I might be wrong, but I think that of all the bad things that
might happen in the world, beginning in Crimea, hyperproductive new
robots are the least of our worries.