Two Books to Understand the Effects of the Technological Invasion ----
In a short book published at the end of 2011, Frédéric Metz invites thereader to join him in seeking answers to a simply formulated question:
what will happen "when Google, facing the world, knows how to see and
name"[1]? "Google" here isn't just the G of GAFAM. This is also why the
author removes its initial capital letter and transforms it into a
common noun. Google is the machine, in the generic sense of the term,
this large machine with indiscernible contours, both a network and an
aggregate of machines, of which only the "search engine" function
usually appears to us, but which we know well relies on a myriad of
systems and interdependencies.
At the time Metz wrote, Google still doesn't know how to see. It
manipulates only words, and recognizes images only by their captions,
the ones humans have added to them. It compares, discriminates, and
calculates only on the basis of words. It does not "see": it cannot
distinguish one face from another, nor a photo of a flower from a
drawing of one. It is already tremendously powerful; but the author
invites us to project ourselves into the next stage of its progression:
the one from which it could, through an infinite network of eyes and
sensors, see the world for itself and interpret it without apparent
human mediation[2]. The point is to consider what this ability, and the
use humans will make of it, will do to knowledge itself.
"We are not talking here about police use-nor about the destruction of
all possible clandestinity. Nor are we talking here about the fear that
machines would then take power, nor about fear for our "freedom," nor
even about the fear of police forces becoming omniscient and
all-powerful. We are talking, much more simply, about a
revolution-colossal and perfectly simple-in everyday life and in
everyday life[...].
We are talking about the definitive disappearance of things that were
initially unknown.
Of any person, before they have said anything, before you have looked at
anything of them, seen anything, your computer-no bigger than a phone,
being your phone-will give you their age, history, genealogy, CV, and a
few hundred photographs taken here and there[...]. And this doesn't just
apply to recognizing people. There is also the recognition of the
unknown bird in the hedge, the unknown painting, and the recognition of
the melody. The machine becomes capable of giving me the name of
everything that appears to me. It recognizes and names everything I see.
It is knowledge given without mediation-without preceding experience,
and without life[...]. The machine, everywhere, precedes my knowledge
and my desire to know. Where will I find the desire to learn, to become
acquainted, when my machine tells me everything, of its own accord, in
advance, always already everything?"
This is the gist of the argument, the dizzying observation to which
Frédéric Metz seeks to draw our attention, summoning the works of the
greatest philosophers (Aristotle, Kant, and Merleau-Ponty) on the nature
of knowledge, experience, and perception, and expanding on Walter
Benjamin's ancient warning that, with the advent of the modern era, "the
course of experience has fallen."
The Google that is coming, Metz tells us, is quite simply the end of
experience. It is a world in which knowledge precedes experience,
answers precede questions. The knowledge accumulated at a given time "t"
is returned to humans in a standardized form (validated by unknowable
biases) and thus spares them any adventure. The past, a certain past,
preempts any possible future.
"The world with Google will be a satiated perception; always already
finished, before having begun; always already satisfied; without desire;
always already done; without violence.[...]Google extinguishes the
violence of perception." And further: "A thing, henceforth, will never
again appear alone. In giving itself, it will always bear its name,
given by someone other than me, harnessed to it, and in addition to its
name, it will have the burden, this name harnessing it similarly,
doubly, quadruply, of all the knowledge accumulated about it-contained
in Google. Google therefore does not pose the problem of a virtual
world, but that of the real world become-reduced into-immediate
knowledge. Google will cover the things of the world with their names.
And the world will appear nothing but decked out, burdened, harnessed,
enclosed, adorned, and covered."
* * *
Whoever controls language controls thought, said Orwell. Yet by
preempting all experience, all knowledge about the things and beings of
the world, Metz spoke to us in 2011 of a formidable threat to thought
itself. And what's truly terrible, writing this note in 2025, is to
realize that this Google has since come into being; and that while "AI"
is undoubtedly the most recent name this machine has taken, it didn't
wait for this latest evolution to reach the stage Metz feared. AI would
rather be a new metamorphosis, which takes us even further than the
speculations of this first author.
Since 2011, we have indeed experienced not only the global distribution
of smartphones (which are as much machines for accessing Google's
knowledge as they are eyes and instruments to feed it), but also that of
cameras and satellites, but also automatic language recognition, the
permanent collection and cross-referencing of data, and finally, with
the advent of machine learning, the mechanical control of an increasing
part of the management of human affairs.
So here we are. Already. Not only has Google learned to see and name-but
now it speaks, advises, and determines certain aspects of human history.
We are subjects on its real-time global map-subjects, agents, and users.
This is what James Bridle calls, in a book published in 2022, the "New
Dark Ages"[3]. He adds from the outset that the phrase is not intended
to provoke anxiety, but is first and foremost an observation and an
invitation: technology has made the world dark for us. We are now moving
forward amidst unknowable causalities, groping in the fog, and we must
learn to live with it. The main cause of darkness, according to Bridle,
is what he calls "computational thinking," which he presents as a
radicalization of the concept of "technological solutionism."
Computational thinking is the digital ordering of the world, in all its
variations. It is the abyss within which every question must find its
answer with the help of measuring and calculating instruments, even when
it involves questioning measurement and calculation themselves. A finite
thought, which no longer knows an outside, which subsumes everything.
Bridle multiplies and expands on examples in all directions - from the
management of the economy via "high-frequency trading" to the monitoring
of air traffic and the weather using autonomous technologies, from the
automatic creation of YouTube videos and absurd and terrifying
commercial objects to the saturation of intelligence agencies with their
own accumulated data, from the unknowable nature of "truths" found on
the internet to real or presumed conspiracies...
We will pause before concluding with one of the anecdotes he presents,
because it sadly resonates with an even more recent reality, like a
final symptom at this time of the mad acceleration in which we are
caught. Bridle cites a story (albeit "probably apocryphal") in which
neural networks were trained by the military to spot enemy tanks hidden
in the woods. The gamble of "machine learning" is that by providing
phenomenal amounts of information to self-educating machines, they will
be able to develop new forms of perception that, even if inconceivable
to humans, will allow them to answer the questions they ask. In this
case, the training was successful: the machines were able to determine
with almost 100% success the aerial photographs of forests in which
tanks were located, and discard those in which there were none. However,
the real-life test was a resounding failure. The military would later
understand that what the machine was detecting was not the presence or
absence of tanks, but the cloud cover in the photos-the series of photos
used for training had in fact been taken at two different times of day,
one with tanks in the woods when the weather was nice, the other with
them removed when the sky had become overcast.
This ridiculous story unfortunately appears to be a paradigm of a tragic
reality when one considers that less than ten years later, soldiers in
the Israeli army receive some of their orders directly from the
"artificial intelligence" with which its command has been equipped. This
intelligence, cynically named "Gospel," allows them to identify
suspected Hamas members based on cross-referencing information collected
by various surveillance systems. These suspected fighters are then
killed based on an algorithmic calculation whose workings are unknown.
It's not science fiction anymore...
Tonio
Who isn't familiar with "CAPTCHAs"?
These little applications are used by some websites, which ask visitors
to select certain boxes to demonstrate that they are human.
However, beyond the strangeness of having to demonstrate to a machine
that you are not one yourself; beyond the (less obvious) technical
significance of the programmers' use of these applications, which is to
prevent autonomous machines from disrupting the intended operation of
their website, which is a reminder, in passing, that such machines
exist; Beyond these considerations, very few people realize that the
corollary purpose of this exercise in human discrimination is to
contribute, with each click, to refining machine learning to better
determine what is a bus, what is a pedestrian crossing, what is a fire
hydrant, etc., and that this learning ultimately contributes to
improving the ability of combat drones to choose their targets.
We are caught, even in our most mundane daily lives, in a vicious web,
in a new dark age whose extent we still struggle to measure, yet which
deepens with each new "innovation."
We must relearn how to think, how to fight... and find the way out.
Frédéric Metz, Les Yeux d'Œdipe (inutiles, sauves). Quand le Google,
face au monde, sera savoir voir et nom (The Eyes of Oedipus (Useless,
Saved). When Google, Facing the World, Will Know How to See and Name),
Pontcerq, 2011. Downloadable PDF version
James Bridle, A New Dark Age. Technology or the End of the Future
Notes
[1]Frédéric Metz, Les Yeux d'Œdipe (inutiles, sauves). Quand le Google,
face au monde, saura voir et nom (The Eyes of Oedipus (Useless, Saved).
Quand le Google, face au monde, saura voir et nom (The Eyes of Oedipus
(Useless, Saved). When Google, Faced with the World, Will Know How to
See and Name), Pontcerq, 2011. Downloadable PDF version at
http://i2d.toile-libre.org/PDF/2011...
[2]Of course, this mediation remains, but it is kept invisible: whether
through programming "biases" or the colossal amount of human
exploitation, AI, behavior-recognition cameras, and other connected
speakers can be constantly adjusted to meet their users' expectations.
[3]James Bridle, A New Dark Age. Technology or the End of the Future
(The Eyes of Oedipus (Useless, Saved). When Google, Faced with the
World, Saura Voir et Nom (The Eyes of Oedipus (Useless, Saved). When
Google, Faced with the World, Saura Voir et Nom (The Eyes of Oedipus
(Useless, Saved).)
http://oclibertaire.lautre.net/spip.php?article4424
_________________________________________
A - I N F O S N E W S S E R V I C E
By, For, and About Anarchists
Send news reports to A-infos-en mailing list
A-infos-en@ainfos.ca
Geen opmerkingen:
Een reactie posten