By Javier Surasky-
This blog
post stems from a conversation with my international law students and their
"fears" about how AI will impact their lives. As I listened to them,
the question that arose in my mind was: What will people living in natural
interaction with AI think about us in the future? Will they see us as concerned
human beings or as frightened children?
From
"2001: A Space Odyssey" to "Terminator," from
"Westworld" to "Black Mirror," AI has been a central actor
in constructing dystopias for decades, even when it was barely learning to play
checkers acceptably.
Of the many
possible futures that await us interacting with AI, those leading to the
destruction of humanity or its domination by machines are not the most
plausible, but they are not entirely unrealizable either. So far, we know that
the final outcome will depend more on human capabilities to consciously and
safely manipulate a compelling technology than on the technology
itself in its autonomous projection.
It may be helpful,
however, to put things in context, at least to lower the levels of anxiety that
the arrival of AI in our lives can awaken. One possible strategy is to see how
humanity has reacted to different technological changes that have impacted
people's lives.
The case of
the Industrial Revolution and the struggle of the Luddites, even by violent
means, to destroy new textile machines to confront a technology that would lead
to the loss of countless job positions is well known. In an exciting
description of those times, Eric Hobsbawm explains in his book Labouring
Men: Studies in the History of Labour that this view is somewhat
simplistic: it was not just about the jobs that would be lost but about the
devaluation of the social value of expert knowledge of artisans in a framework
of exploitation of workers by factory owners. While we were then facing a
technological leap, and today we are facing an enormous but continuous advance,
the similarities between the fears and concerns of then and now in the face of
new machines are easy to identify.
Something
similar happens with telephones. Today, we worry about the amount of private
data they take from our daily activities (who we talk to, at what time, where
we go, what places we visit, who we are with, etc.), but when the first
telephones were put into use after their patenting by Alexander Graham Bell in
1876 (a historical case of "theft" of creation rights since the
inventor of the telephone was Antonio Meucci, in 1954, who did not patent it
because he lacked the economic means to do so), the fears it aroused around the
invasion of privacy through third-party eavesdropping accompanied by stories of
people who became seriously ill after having spoken on the phone generated no
few rejections, a story very well narrated by Tom Standage in his book The
Victorian Internet.
The current
autonomous car forces us to use the same prefix twice with a powerful
meaning and generates more than justified fears. We can cite as an
antecedent the rejection produced by the introduction of the automobile,
arousing fears of accidents and revolutionizing life in cities, where
pedestrians were accustomed to circulating on sidewalks and streets. However,
there is another antecedent that better applies to the case of cars without
human drivers: elevators without human operators.
It may seem
ridiculous today, but until the mid-1940s, elevators were people-operated, and
the transition to automatic elevators (not intelligent, which is different) was
not free from rejections by users who refused to use them. In the words of
David Mindell in his book Between
Human and Machine: Feedback, Control, and Computing before Cybernetics, automatic
elevators were considered by the general public as "too complex to
generate trust in the absence of a human operator."
The
"disappearance" of elevator operators did not occur without union
rejections and strikes, and there were even processes of what we would now call
labor reconversion to relocate many of them when their unions managed to impose
them, especially in the case of elevators in business buildings.
Other
technologies that suffered social rejections in their time were the train,
especially by rural populations who feared that its arrival would mean the end
of their traditional ways of life, and electric energy, which aroused
rejections in its integration into cities and homes due to the dangers
generated by the use of light bulbs.
The arrival
of computers was no more peaceful: unions demanded that job automation be
regulated and relocation programs be created for workers displaced by machines,
generating in the technologically more advanced Western countries a strong
political debate around industrial automation, public good, and control of
technological change.
In all
these cases, it was necessary to create trust in new technologies through
social and educational campaigns and generate response mechanisms to the social
changes that were generated, especially in the world of work, before new
technologies became entirely accepted.
There are
important lessons here for making the conscious introduction of AI into the
lives of individuals and societies analyzable in a realistic way, without
hiding its potential to introduce profound changes in our lives, which each
person and society will have to define as "good" or "bad,"
"desirable" or "undesirable." This can only be done by
democratizing the information and training necessary to process the most potent
technological change humanity has experienced.
