(These are excerpts from my book "Intelligence is not Artificial")
The media are full of stories about how machines will automate most of today's
jobs. Many conclude that machines will cause massive unemployment.
I suspect that the truth is closer to the exact opposite conclusion:
that technology will lead to hyper-employment (no unemployment and more
than one job per person), and hyper-employment will cause a
the global catastrophe similar to and possibly including
hyperinflation, with consequences that we can't imagine.
Many of the jobs that existed in 1900 had been automated by 1960.
Today most of us have jobs that didn't exist in 1900.
Therefore the fact that tasks will be automated by machines is nothing new.
It would be a scary sign of technological decline if it didn't happen.
Despite all the jobs that were automated between 1900 and my birth,
i wasn't born into a world of massive unemployment. On the contrary,
i was born in a world that offer much better and better paid jobs than the
jobs that had been available to my grandfather.
So the real question is whether in the near future the percentage of jobs that
will be automated is rought the same, less or more the percentage of jobs
that were automated in the past; and whether the new jobs created by this
future automation will match the new jobs created by automation in the past.
For example, the world will need a lot more engineers to build, program and maintain millions of software and hardware robots.
A 2016 study by Robert Atkinson and John Wu "False Alarmism: Technological Disruption and the U.S Labor Market, 1850-2015" doesn't show any major deviation
from the past... so far.
I certainly wish that jobs like plumber, electrician, etc were automated.
Alas, i suspect that the world will need a lot more of them, not fewer,
in the next 25 years.
Personally, i think that, as it is often the case, the media tend to look
in the wrong direction.
Machines that replace human jobs are nothing new. Machines to which you
can outsource part of a job are, instead, something really new.
These machines allow you to outsource
not just one job but many, as many as you can, and desire to, manage.
I suspect that in the near future many people will have two or three or 25 or 2,000 jobs, thanks to tools that will allow us to massively multitask.
In that case, unemployment will de facto become negative: more than one job per person,
even for the very old and the very young. A 90-year-old man will be able
to carry out the activities of a 20-year-old and in fact the activities
of many young men. Energy and health will not be obstacles anymore.
A 5-year-old will be able to generate money by using bots to make and
My prediction is therefore exactly the opposite of the most popular
"doom and gloom" predictions:
the global threat to social stability of the 21st century will be hyper-employment.
We may already be living in that age. Most of us work way more than 8 hours/day.
We see studies about employment all the time, but those count only traditional
work, the kind of work that will disappear.
I'd like to see a study about the number of hours worked per capita.
I suspect that number is going up. Automation is creating so much work
that each person can work a lot more than in the past, potentially 16 hours
a day, potentially 24 hours a day: you'll have a swarm of bots working for
you even when you sleep.
We forgot that, in medieval times, people worked only when they wanted,
and some worked only six months a year.
In the last century it was common in Italy
to have a lengthy breakfast, a 3-hour lunch break and play bocce or cards in the
evening before dinner.
Juliet Schor in "The Overworked American" (1992), using a variety of historical
studies, estimated that in the
13th century an adult male peasant in Britain worked 1620 hours compared with
1980 hours worked between 1400 and 1600 by an adult male farmer or miner
and compared with 1949 hours worked in 1987 by the average worker in the USA.
The Bureau of Labor Statistics report of 2000 showed that more than 25 million Americans (20.5% of the total workforce) worked at least 49 hours a week.
OECD maintains statistics about the
average annual hours actually worked per worker that show a 2% decline between
2000 and 2015 in the USA (from 1836 hours to 1790),
but these statistics
made in the age of the "net economy" can be misleading. They don't count
all the independent additional income made by many of us on the Internet,
and certainly don't count all the work made for free by the majority of
Internet users when they supply content to social media.
Ian Bogost in the Atlantic magazine
argued that we're already hyperemployed, except
that we mostly work for free, providing Internet companies the content that
they monetize (for themselves).
Until now we were working multiple jobs without even knowing that we were
doing so, but we will soon become more and more aware that our
leisure activities are sold as commercial goods by some corporations.
In 1930 the economist John Maynard Keynes wrote the essay
"Economic Possibilities for our Grandchildren" in which he
argued that future generations would soon be able to replace work with leisure thanks to widespread wealth and surplus.
"I predict that both of the two opposed errors of pessimism which now make so
much noise in the world will be proved wrong in our own time-the pessimism
of the revolutionaries who think that things are so bad that nothing can save us
but violent change, and the pessimism of the reactionaries who consider the
balance of our economic and social life so precarious that we must risk no
He was wrong on human psychology though: people's preferred form of
entertainment is not poetry, nor painting, nor music, but making money.
We, turned into hyper-employed workers by swarms of bots, will use our
"leisure time" to make more and more money, selling our
products to people who are making more and more money selling their things
to people who are making more and more money selling their...
Back to the Table of Contents
Purchase "Intelligence is not Artificial"