Erik Brynjolfsson and Andrew McAfee -- The second Machine age
===============================================================

The first parts of "The second machine age" are a celebration of all
the cool and spiffy things that computers and digital devices can
do.  There is emphasis on how that progress has accelerated in just
the last several years and on how tasks that we thought, just 2 or 3
years ago, computers would never be able to do, they are now doing
or are in the early and promising stages of becoming able to do.

That's followed by descriptions of how wonderful all these blessings
will be, what Brynjolfsson and McAfee call the "bounty" of this
technological progress.  But, some of the later sections of the book
become less sanguine, as Brynjolfsson and McAfee consider both (1)
how the benefits (and costs) of the adoption of this digital
technology is going to be distributed unevenly and (2) the likely
effects and consequences of the elimination of so much work through
robotics and automation.

And, then an awareness of reality appears as Brynjolfsson and McAfee
describe some of the negative consequences of all that technology
and automation, notably the loss of employment opportunities and the
increasingly skewed distribution of wealth and income with a larger
are larger share of profits going to a smaller and smaller
proportion of the population.

With respect to the technology itself, it's not just that the
devices (desktop computers, laptop computers, smart phones, embedded
devices, etc.) are becoming more powerful and capable; it's also
that those devices have more (machine readable) data to work with
*and* that so many of those devices are connected.  This suggests
that much of the data that is being created is either accessible to
or being created and updated by these connected devices.  It also
suggests that these devices are interacting with each other, for
example that these devices send information and requests to each
other.  Eventually, our devices (and there are likely to be more and
more of them) will do increasing amounts of their processing in
cooperation with the devices around us (or, more appropriately,
around *them*, since you and I will likely become less important
assistants for their work) or  in cooperation with computers and
devices accessible across the Internet.  In the future, although we
may have an identifiable device in our hands or a computer on the
desk in front of us, that "computer" will effectively be composed of
it and the devices it is connected to (and the devices they are
connected to recursively).  We are not likely to be aware of many of
these devices (just as we are not aware of many of the embedded
computers and controllers in our automobiles), nor are we likely to
be aware of the data they are collecting.  We may have a dim
awareness that they (the computing devices) are doing tasks that
they did not formerly seem to be able to do and that they make
intelligent choices that they did not formerly make (e.g., suggest
restaurants that we prefer and purchases that entice us).

If innovation and technological advances lead to increased
productivity and increased productivity leads to improved standard
of living, then it is highly unlikely that we as a society or our
political institutions are going to chose to slow it down.  But,
that means that we are in the midst of a perma-growth economic
system.  If that is so, you have to ask what we could possibly hope
to do to slow down global warming and the consumption of resources.
Aren't our needs to deal with resource consumption and deletion in
direct conflict with the changes that Brynjolfsson and McAfee
describe are so important to improving our lives.  We can imagine 50
to 100 years from now any society that manages to survive will look
back at us and ask: What were they thinking?  Could they (we) have
possibly *not* been aware of the consequences of their actions and
policies?

It is a bit depressing that books like this one constrain the set of
"good" or beneficial policy choices to the set described by
perma-growth economics, to policies that promote more economic
growth and more production and more consumption.  Are there really
no other alternatives?  Possibly so, however, it does not seem to me
that we are always going to have the option of attempting to solve
our problems with "more": more economic growth, more
production/productivity, more consumption, etc.  At some time in the
not to distant future, we may have to start making do with less. We
may have to stop growing the total size of the pie so that everyone
(or most of us) can have a slightly larger slice.  We may need to
learn how to divide a smaller (total) pie, but share it more
equitably.

Believe it or not, Brynjolfsson and McAfee are worried about the
slowing down of innovation and increased productivity, and not just
about the possible negative consequences.  They worry that, with
respect to innovation, we have already picked most of the low
hanging fruit and that the most productive innovations are behind
us.  Honestly?  Do we need new ways to burn fossil fuels and produce
more plastic?  Do we really need to figure our how to do so more
quickly?  Shouldn't we be more concerned with understanding why the
benefits of innovation have not spread more evenly across the world?

Brynjolfsson and McAfee dismiss their own worry about innovation
slowing down based on these reasons: (1) many or most productive
innovations come about through recombination and adaptation of
existing ideas and innovations; (2) information and communication
technology spreads those ideas and makes them more readily
available; (3) many more people have access to the technology that
enables them to learn about and learn from existing ideas, which
will spread innovation possibilities more broadly across, say, more
minds capable of having more ideas and inspirations, which will
cause an increase in innovation activity.  As someone who has been a
computer programmer for many years, I can willingly agree with that.
In comparison with my experiences 20 years ago, I have quicker
access to more of the information I need about the computer
programming technologies I use (programming languages, libraries of
support code and functions, and programming tools such as text
editors and debuggers), answers to questions about programming
problems I run into, etc.  If you need evidence about that, look up
almost any programming language or technology at Wikipedia and
follow the links near the bottom of the page.  And, go to
StackOverflow (http://stackoverflow.com/) and look at the incredible
wealth of questions and answers there.  None of this readily
available information existed 20 years ago.  When it comes to
computer programming, we truly are living in information rich times.

Brynjolfsson and McAfee also attempt to explain why increases in
productivity do not immediately become visible when technological
innovations are adopted.  They say this is due to a time lag as
businesses adapt and improve the application of new technologies to
their needs.  And that leads to Brynjolfsson and McAfee's claims
that, during the last 10 to 15 years, worker productivity has
increased dramatically.

Brynjolfsson and McAfee do devote a good amount of space toward the
end of this book to the problems associated with the loss of jobs
and the problems that will result from an increasingly uneven
distribution of wealth and income that we are seeing, especially
currently in the U.S.  But, that discussion gives suggestions such
as the following: get more education, train for a job that requires
non-routine work, and so on.  These are suggestions and policies
that might help some individuals, but will not fix a society or an
economy.  A few of the points they make: (1) They explain how we (in
the U.S.) may be entering a time of "winner take all" era of
economic compensation where a disproportionate amount of earnings
goes to a small number of individuals and where income distribution
is best described by a power law, i.e., a "fat tail" distribution
curve.  (2) Advancing technology and innovation is a likely cause of
both a bounty in the way of improved medical care, better
communications, safer air travel, and much more, as well as negative
consequences such as the extremely skewed and inequitable
distribution of incomes and the loss of so many employment
opportunities.  (3) It's actually possible to argue that we are all
better off because of our access to improved technology and its
benefits, and Brynjolfsson and McAfee do a reasonably balanced job
of analyzing both the good and bad consequences of advancing
technology.  (4) The economic inequality in the U.S. is likely
producing increasing political inequality, which will, in turn,
enable a set of economic and political elites to protect their
favorable and unfair position and to deny all others access to
economic opportunities and the political process.  That will,
Brynjolfsson and McAfee think, lead both to even more inequality and
to a slowing of the innovation needed for the increased productivity
that, they think, will improve our lives.

Further reading -- (1) If you'd also like to read a book that takes
a similar view that innovations and advanced technology are the
prescription for improving our lives and raising our standard of
living, but written from a political point of view, you might want
to look at "Why Nations Fail: The Origins of Power, Prosperity, and
Poverty", by Daron Acemoglu and James Robinson, which by the way, is
referred to by Brynjolfsson and McAfee.  I'd also recommend that you
take a look at: (2) "The rise of the robots", by Martin Ford; and
(3) "The glass cage", by Nicholas Carr.

08/07/2015

.. vim:ft=rst:fo+=a: