artificial intelligence is capturing the
hopes and fears of people from all over
the country from our politicians to tech
leaders to celebrities or once again
faced with the question about how
technological disruption will impact the
livelihoods of the workers it makes
redundant or once again faced with the
machine question and the proposition
that this time is different today
popular media is filled with think
pieces on the topic and our public
discourse has become particularly
polarized on the web side people like
Elon Musk offer apocalyptic warnings
about the perils of coming autonomous
machines and on the other savvy
executives and economists salivate at
the bounty and increases in productivity
that AI seems to promise they’re not
disturbed by the resurgence of fears
about technological disruption who is
right about artificial intelligence well
AI spur extraordinary improvements in
health care and security oh wow boosting
standards of living and increasing
productivity or will a I Cod catalyze
massive disruption and widespread
unemployment will it undermine the
socio-economic fabric of American
society these are questions that I’ve
been researching for the past year and
I’ve come to believe that in order to
answer them we need to first turn to
history and consider how today’s fears
about technological disruption compared
to past historical ones doing so reveals
that the greatest risk we face today is
not joblessness or sentience but rather
income inequality and though there are
great risks in embracing new labor
saving technology we have a tremendous
ability to meet these challenges here’s
an apt summary of technological fear
every day sees the creation of some new
mechanical miracle that enables one
worker to do better and more efficiently
what many workers used to do nothing
like it has ever been seen on earth but
what did these machines do for us they
dispensed with workers that’s a quote
from US Secretary of Labor James Davis
in his 1927 address his speech titled
the problem of the worker displaced by
machinery was concerned with the impact
of new industrial designs and labor
saving devices on workers what would
happen to the worker employed in
glassware if he saw his job automated
what would happen to the livelihoods of
the family these are questions that the
Luddites also famously asked now the
Luddites were a collection of British
agricultural workers who feared the
innovations in labor saving devices that
would catalyze the Industrial Revolution
they were so concerned about the
permanent redundancy of their labor that
they took to destroying machine designs
destroying machines themselves and even
murdering machine creators who will
undertake the arduous task of learning a
new trade
they asked let’s start with some facts
about where we are today America has
never been wealthier and more prosperous
than it is now
instead of the end of work we’ve created
more jobs than ever before and the US
has seen an increase in work in nearly
every decade this past century Davis and
the Luddites were wrong or at the very
least they were wrong about the main
point what happened in Britain in the
18th and 19th century and America in the
early 20th century was the temporary
displacement of some workers as the
economy left forward and during that
time productivity skyrocketed and the
world was on a trajectory to experience
almost unfathomable improvements in
standard of living and public health in
literacy and in wealth and yet Davis and
the Luddites were right in one profound
sense though long-run prosperity was
bolstered and jobs ultimately came back
there were serious costs to that initial
period of disruption during both periods
of displacement income inequality
increased and the reassurance that
long-run prosperity would be bolstered
and that jobs would return that would
have been of little comfort to a British
textile worker who’d seen his or her job
automated by the spinning jenny we
should take lessons from history when
speculating about the future impact of
artificial intelligence and on the
surface the nature of today’s disruption
seems similar about 3% of all Americans
are employed in transportation and the
advent of the self-driving car threatens
to make millions of Americans largely or
partially redundant and we can see other
stories in many other sectors such as
manufacturing we should take lessons
from history when speculating about this
issue but I think it’s worthwhile to not
fall into the deep pessimism that some
of the more the most alarmist voices on
the topic have put forth though a
University of Oxford study found 47% of
all jobs could be automated and a more
recent McKinsey Global Institute study
found roughly 60% of jobs could be 30%
automated I think we should avoid
falling into alarmism for three reasons
first each study speculates about the
percentage of jobs that might be
automated not the percentage of jobs
that will and in reality there are
significant barriers to adopting
artificial intelligence in the workplace
and most AI is used to a company but not
replace human labour
and second economics tells us that as AI
makes the production process cheaper and
more efficient Goods become less
expensive for consumers and that’s money
that can be spent in other industries
which creates new jobs and third there
are already a shortage of workers in
industries that we know about take elder
care and child care services for example
with more women working than ever before
and the baby boomers retiring there’s a
serious shortage of workers to look
after our children and elders and that’s
work that can’t be automated but despite
that income inequality remains a crucial
issue just as the Luddite saw a period
of widespread inequality during the
industrial revolution AI and the
automation it enables may couple with
outsourcing and globalization to spur
greater inequality today and we can
already see the beginnings of this from
1980 to 2014 the share of national
wealth for the bottom 50% of American
households fell from 20 percent to 12
percent and during that same period the
share of national wealth going to the
top 1% of American households increased
by the same amount from 12% to 20%
so today the top 1% of American
households owns about as much wealth as
the bottom 20% and the story is even
more dramatic for the top 1% of the top
1% which is seeing their share of
national wealth growth by 6 times what
it was in 1970 we should all be
concerned about how AI may exacerbate
these recent trends in income inequality
and though there are obvious moral
reasons why we might object to high
levels of inequality there are also some
very practical
reasons to on the economic side high
levels of inequality make a country less
competitive as the resources needed to
innovate get disproportionately
allocated and it correlates heavily with
a country’s homicide rate it can help
spur the rise of anti-democratic
populist and since the time of Plato has
been cited as a chief cause of
democratic erosion and a recent book by
Walter Shadle finds that throughout
history periods of inequality only ever
get remedied or leveled by one of four
things plagued mass military
mobilization civil war or government
collapse this is why when we consider
the risks posed by artificial
intelligence we should view it not as a
problem about joblessness or about
sentience but rather one about income
inequality and we should view it as an
issue that strikes at the foundation of
our democracy so whether the solution is
better retraining or a universal basic
income we need to seriously reevaluate
the way we approach social security and
education for a modern era and by doing
so we can begin to navigate these
transformative times so from our health
care systems to our education to our
militaries to our economies and
ultimately to our democracies the risks
and rewards of artificial intelligence
are great it’s up to all of us to decide
which is greater thank you very much [Applause]
hopes and fears of people from all over
the country from our politicians to tech
leaders to celebrities or once again
faced with the question about how
technological disruption will impact the
livelihoods of the workers it makes
redundant or once again faced with the
machine question and the proposition
that this time is different today
popular media is filled with think
pieces on the topic and our public
discourse has become particularly
polarized on the web side people like
Elon Musk offer apocalyptic warnings
about the perils of coming autonomous
machines and on the other savvy
executives and economists salivate at
the bounty and increases in productivity
that AI seems to promise they’re not
disturbed by the resurgence of fears
about technological disruption who is
right about artificial intelligence well
AI spur extraordinary improvements in
health care and security oh wow boosting
standards of living and increasing
productivity or will a I Cod catalyze
massive disruption and widespread
unemployment will it undermine the
socio-economic fabric of American
society these are questions that I’ve
been researching for the past year and
I’ve come to believe that in order to
answer them we need to first turn to
history and consider how today’s fears
about technological disruption compared
to past historical ones doing so reveals
that the greatest risk we face today is
not joblessness or sentience but rather
income inequality and though there are
great risks in embracing new labor
saving technology we have a tremendous
ability to meet these challenges here’s
an apt summary of technological fear
every day sees the creation of some new
mechanical miracle that enables one
worker to do better and more efficiently
what many workers used to do nothing
like it has ever been seen on earth but
what did these machines do for us they
dispensed with workers that’s a quote
from US Secretary of Labor James Davis
in his 1927 address his speech titled
the problem of the worker displaced by
machinery was concerned with the impact
of new industrial designs and labor
saving devices on workers what would
happen to the worker employed in
glassware if he saw his job automated
what would happen to the livelihoods of
the family these are questions that the
Luddites also famously asked now the
Luddites were a collection of British
agricultural workers who feared the
innovations in labor saving devices that
would catalyze the Industrial Revolution
they were so concerned about the
permanent redundancy of their labor that
they took to destroying machine designs
destroying machines themselves and even
murdering machine creators who will
undertake the arduous task of learning a
new trade
they asked let’s start with some facts
about where we are today America has
never been wealthier and more prosperous
than it is now
instead of the end of work we’ve created
more jobs than ever before and the US
has seen an increase in work in nearly
every decade this past century Davis and
the Luddites were wrong or at the very
least they were wrong about the main
point what happened in Britain in the
18th and 19th century and America in the
early 20th century was the temporary
displacement of some workers as the
economy left forward and during that
time productivity skyrocketed and the
world was on a trajectory to experience
almost unfathomable improvements in
standard of living and public health in
literacy and in wealth and yet Davis and
the Luddites were right in one profound
sense though long-run prosperity was
bolstered and jobs ultimately came back
there were serious costs to that initial
period of disruption during both periods
of displacement income inequality
increased and the reassurance that
long-run prosperity would be bolstered
and that jobs would return that would
have been of little comfort to a British
textile worker who’d seen his or her job
automated by the spinning jenny we
should take lessons from history when
speculating about the future impact of
artificial intelligence and on the
surface the nature of today’s disruption
seems similar about 3% of all Americans
are employed in transportation and the
advent of the self-driving car threatens
to make millions of Americans largely or
partially redundant and we can see other
stories in many other sectors such as
manufacturing we should take lessons
from history when speculating about this
issue but I think it’s worthwhile to not
fall into the deep pessimism that some
of the more the most alarmist voices on
the topic have put forth though a
University of Oxford study found 47% of
all jobs could be automated and a more
recent McKinsey Global Institute study
found roughly 60% of jobs could be 30%
automated I think we should avoid
falling into alarmism for three reasons
first each study speculates about the
percentage of jobs that might be
automated not the percentage of jobs
that will and in reality there are
significant barriers to adopting
artificial intelligence in the workplace
and most AI is used to a company but not
replace human labour
and second economics tells us that as AI
makes the production process cheaper and
more efficient Goods become less
expensive for consumers and that’s money
that can be spent in other industries
which creates new jobs and third there
are already a shortage of workers in
industries that we know about take elder
care and child care services for example
with more women working than ever before
and the baby boomers retiring there’s a
serious shortage of workers to look
after our children and elders and that’s
work that can’t be automated but despite
that income inequality remains a crucial
issue just as the Luddite saw a period
of widespread inequality during the
industrial revolution AI and the
automation it enables may couple with
outsourcing and globalization to spur
greater inequality today and we can
already see the beginnings of this from
1980 to 2014 the share of national
wealth for the bottom 50% of American
households fell from 20 percent to 12
percent and during that same period the
share of national wealth going to the
top 1% of American households increased
by the same amount from 12% to 20%
so today the top 1% of American
households owns about as much wealth as
the bottom 20% and the story is even
more dramatic for the top 1% of the top
1% which is seeing their share of
national wealth growth by 6 times what
it was in 1970 we should all be
concerned about how AI may exacerbate
these recent trends in income inequality
and though there are obvious moral
reasons why we might object to high
levels of inequality there are also some
very practical
reasons to on the economic side high
levels of inequality make a country less
competitive as the resources needed to
innovate get disproportionately
allocated and it correlates heavily with
a country’s homicide rate it can help
spur the rise of anti-democratic
populist and since the time of Plato has
been cited as a chief cause of
democratic erosion and a recent book by
Walter Shadle finds that throughout
history periods of inequality only ever
get remedied or leveled by one of four
things plagued mass military
mobilization civil war or government
collapse this is why when we consider
the risks posed by artificial
intelligence we should view it not as a
problem about joblessness or about
sentience but rather one about income
inequality and we should view it as an
issue that strikes at the foundation of
our democracy so whether the solution is
better retraining or a universal basic
income we need to seriously reevaluate
the way we approach social security and
education for a modern era and by doing
so we can begin to navigate these
transformative times so from our health
care systems to our education to our
militaries to our economies and
ultimately to our democracies the risks
and rewards of artificial intelligence
are great it’s up to all of us to decide
which is greater thank you very much [Applause]
Please follow and like us: