|
A - I n f o s
|
|
a multi-lingual news service by, for, and about anarchists
**
News in all languages
Last 30 posts (Homepage)
Last two
weeks' posts
Our
archives of old posts
The last 100 posts, according
to language
Greek_
中文 Chinese_
Castellano_
Catalan_
Deutsch_
Nederlands_
English_
Francais_
Italiano_
Polski_
Português_
Russkyi_
Suomi_
Svenska_
Türkurkish_
The.Supplement
The First Few Lines of The Last 10 posts in:
Castellano_
Deutsch_
Nederlands_
English_
Français_
Italiano_
Polski_
Português_
Russkyi_
Suomi_
Svenska_
Türkçe_
First few lines of all posts of last 24 hours
Links to indexes of first few lines of all posts
of past 30 days |
of 2002 |
of 2003 |
of 2004 |
of 2005 |
of 2006 |
of 2007 |
of 2008 |
of 2009 |
of 2010 |
of 2011 |
of 2012 |
of 2013 |
of 2014 |
of 2015 |
of 2016 |
of 2017 |
of 2018 |
of 2019 |
of 2020 |
of 2021 |
of 2022 |
of 2023 |
of 2024 |
of 2025
Syndication Of A-Infos - including
RDF - How to Syndicate A-Infos
Subscribe to the a-infos newsgroups
(en) France, OCL CA #351 - The Devastating Effects of AI Today (ca, de, fr, it, pt, tr)[machine translation]
Date
Sun, 13 Jul 2025 07:45:42 +0300
Currently, two seemingly opposing narratives dominate regarding
Artificial Intelligence: 1/ We are led to believe that AI will be
capable of solving all current problems (ecological, climatic, economic,
etc.) tomorrow; 2/ We are warned of the apocalypse through the takeover
of AI and the end of humanity. ---- These two dominant narratives are
paradoxically produced by the same people, the leaders of major AI
companies. Their goal: to focus us on a future, sometimes idyllic,
sometimes frightening, in order to prevent us from thinking critically
about the present and the current damage generated by AI[1].
That said, in this article we will discuss not the future but the
present of AI and the human and geopolitical consequences that they
cause today: the destruction of the environment, the frenzied
exploitation of humans, imperialist competition and a political
evolution of contemporary capitalism that may be major. Each area
discussed would merit lengthy developments; we instead propose a
synthesis to attempt a more panoramic view of the real issues behind AI[2].
This article is part of the continuation of the file on the
computerization of the world, published in issue 349 of April 2025 of
Courant Alternatif , and in particular the first article on AI .
Introduction
AI is dominated in the Western world by 7 trusts, called "the
magnificent 7" by the world of finance because they have an incredibly
higher capitalization than any other company (all sectors combined)...
in short, the 7 most "rotten": Apple, Microsoft, Amazon, Meta, Google,
Tesla, Nvidia. In China too, there are dominant players like Alibaba,
even if they are often forgotten in the French press. Capitalism is
banking on the development of AI to revive its profit economy and with
very clear goals of political domination. AI is an industry of
capitalist profitability: reducing costs, developing new markets,
generating new monopolies... with therefore gigantic investments that
could be nothing more than a speculative bubble that would soon burst.
Associated with this development is a whole "ethical" discourse to, it
seems, frame AI. This ethic is neoliberal and in no way hinders the
development of AI as a tool of exploitation and oppression. For example,
the ethical framework for AI in Europe, touted by our media, was openly
written largely by these companies via "exceptions" to the constraints
imposed... in fact authorizing just about anything.
Capitalism has already stripped us of our work by proletarianizing us,
technoscientific development seeks to strip us of our thinking (our
cognition) under the pretext of simplifying our lives via digital
technology and social networks (among others). To this end, tech
companies[3]use all the data produced digitally (emails, SMS, internet
queries, ... but also literary and academic databases, ...) and this
most often in a completely illegal manner. As AI is dominated by 7
trusts with economic power over States, these companies impose
themselves politically and allow themselves everything without constraint.
Beyond the exploitation of labor, environmental impacts include those
related to the extraction of raw materials, manufacturing, distribution
and end-of-life of equipment. This leads to overconsumption of water,
electricity, depletion of materials, soil pollution, etc. In addition to
this immense waste, strategic access to these resources fuels deadly
conflicts, such as in the Democratic Republic of Congo[4]. This
development of AI is in symbiosis with a political evolution of
capitalism of which Trump is the most exemplary figure, generating
growing imperialist tensions.
Human exploitation
AI boasts of automated work, freeing us from thankless tasks. The
reality is quite different. There are permanently between 150 million
and 450 million people, or between 4.4% and 12.5% of the global
workforce, working every day to teach AIs not to produce morally
unacceptable answers or to correct response errors. This work is paid by
the task, often carried out in countries of the "South", and therefore
very poorly paid. For example, social networks censor morally
unacceptable images (rape, murder, pedophilia, etc.). To build
algorithms capable of recognizing such images, we must go through
humans. Poor countries then serve as a garbage dump for AI, asking them
to perform tasks that would not be acceptable in the Western world. The
most characteristic situation is these employees in Kenya exposed 8
hours a day to such images or unbearable texts for 2 dollars a day[5].
Others are tasked with using their mouse to indicate where pedestrians,
vehicles, and traffic lights are located on thousands of images to
"teach" the algorithms of future autonomous cars to drive without human
assistance... for 0.01 euros per image processed in Madagascar, for
example. In Finland, prisoners do this job.
AI also involves the exploitation of strategic minerals such as lithium,
copper, zinc, gold, and cobalt in countries such as Zimbabwe, the
Democratic Republic of Congo, Zambia, South Africa, Bolivia, Chile, and
Argentina. Working conditions are among the worst: exhausting and highly
accident-prone work performed by adults and children, with very
significant health risks for the environment. In the DRC, at least
40,000 children (some as young as 4) work 14 hours a day in perilous
conditions for a paltry income.
Artificial intelligence is therefore, above all, human labor, and the
fantasy of automating everything quickly falls flat: Amazon tried to
replace cashiers with cameras; for fewer than 200 stores, a thousand
Indian workers were needed to ensure reliability. Self-driving cars
require remote human supervision, estimated at 1.5 humans per autonomous
vehicle.
In short, AI is the exploitation of hundreds of millions of people,
including 1.5 million in Amazon's logistics warehouses. To exploit, all
dissent must be prohibited. Within AI companies, as in many other
sectors, there is repression against any unionization. Amazon, for
example, decided to close all its warehouses in Quebec (January 2025)
because there was the threat of unionization. Other tech companies are
not far behind, as unionization is virtually prohibited there. Behind
AI, there are very capitalist companies with all the brutality that goes
with it.
Resource exploitation
Data center servers are replaced on average every two or three years. AI
therefore requires a huge amount of raw materials upstream, particularly
rare earths... essential factors in the wars in Africa for the control
of these resources because many of them are still unexploited on this
continent - unlike other regions of the world. Greenland has suddenly
appeared as also highly coveted, the melting of the ice makes
unexploited mineral resources accessible, particularly rare earths.
The extraction of metals necessary for AI is particularly
energy-intensive and polluting, with the use of acid and water which,
through runoff of these highly pathogenic metals, leads to dramatic
pollution of groundwater and rivers for local populations and from which
results a large number of pathologies.
Furthermore, electronic components contain toxic products such as
arsenic, mercury, etc., which are released into the environment during
the processing of AI waste, polluting the surrounding inhabited areas.
Specifically, the population living around the largest waste center in
North America (in Quebec) has, according to an epidemiological study, a
life expectancy five years lower, significant growth delays in newborns,
and a 45% higher incidence of lung diseases, including cancer.
Beyond raw materials, AI also involves electricity. Already, the energy
consumption of mines uses a lot of electricity (grinding machines,
pumping, etc.), not to mention the transportation of their ores. The
energy consumption of AI data centers is unknown; we only have
estimates. Facebook indicates, for example, that its models behind its
search engine are retrained every hour; roughly the annual carbon
footprint of 500,000 French people. The International Energy Agency
estimates that, globally, by 2026, the AI industry alone will consume as
much electricity as Japan-the world's fourth-largest economy. In Texas
and California, this type of installation already accounts for more than
10% of global consumption. In Ireland, it's 20%. In France, 10% of
electricity is already consumed by digital technology today. Based on
the promises of investment in gigantic data centers recently announced
in the USA, Europe, or China, a virtually inexhaustible supply of
electrical energy would be needed. A
lot of electricity needs to be produced... and so, in the eyes of tech
executives, nuclear power appears to be a way out of the energy impasse
toward which artificial intelligence is leading us. Microsoft recently
announced the restart of Unit 1 of the Three Mile Island power plant -
the site of a nuclear accident in 1979 that halted the development of
this technology in the United States. Today, many AI players say they
are studying the installation of mini-reactors ("SMRs," a technology
that still doesn't work) next to their main research centers and are
investing in nuclear fusion (still at the promising stage for decades).
AI also means water because data centers need to be cooled. One of the
largest data centers in operation is that of the US intelligence agency,
the NSA, which covers more than 100,000 square meters in Utah, a region
particularly prone to drought. It alone consumes 60% of the state's
water consumption. It needs 90 million liters of water each month, 35
times the size of the Paris 2024 Olympic swimming pool. In 2027, it is
estimated that between 4.2 and 6.6 billion m³ of water will be needed to
run AI and cool servers, almost the water needs of Denmark, or half
those of the United Kingdom. Such consumption will worsen shortages and
exacerbate problems of access to water.
Geopolitical impact
The United States has been waging a technological war with China for
several years, and the "AI battle" is only just beginning. Biden had
prevented China from gaining access to microchips manufactured by Nvidia
(the most powerful chips that power artificial intelligence systems) and
ultraviolet lithography machines-two essential technologies for
semiconductor manufacturing. However, US sanctions have not prevented
China from advancing in the technological war against the United States.
Recently, the Chinese company DeepSeek updated a highly sophisticated
artificial intelligence model using less powerful semiconductors... and
this at a cost announced to be significantly lower than that of ChatGPT.
Other Chinese AIs are emerging to compete with American AIs and reveal
the current frenzied competition. Thus, Macron, through his "Summit for
Action on Artificial Intelligence," sought to make an impression with
hundreds of billions of euros in investment promises in France.
Above all, China has continued to gain momentum in the technological
race that pits it against the United States today. China is the world
leader in scientific publications, particularly in artificial
intelligence, and in patent filings. But China does not currently have
the technological expertise to produce semiconductors as efficient as
those produced in Taiwan and from which US corporations benefit, hence
in part China's designs on Taiwan. Trump recently imposed a more
significant blockade on the export of certain chips to China and should
succeed in ensuring that part of the production of these chips is
carried out in the USA and no longer solely in Taiwan. On the other
hand, China currently has a virtual monopoly on 30 of the 50 critical
metals necessary for AI: lithium, copper, essential rare metals, etc.
China, in response to the economic war launched by Trump, has blocked
the export of a certain number of these rare materials. This explains
the recent imperialist aims of the United States for Canada,
Greenland... or Ukraine[6].
Trump's current policy, and the associated trade war, is therefore part
of a confrontation with China for, among other things, technological
control, that of minerals and rare earths as well as means of
communication (such as the Panama Canal). The desire for new US
conquests forces European states to find new strategies, with increasing
tensions, and therefore a policy of frenzied rearmament. Imperialism has
never ceased, but it takes a more brutal form through the declared
desire to conquer new territories by force, and therefore becomes more
conflictual. This could become widespread and in the line of sight chaos
and war. Consequently, AI has recently been experiencing accelerated use
for military purposes. This is done by the rapprochement of tech giants
and states for the emergence of autonomous weapons systems, capable of
functioning without the need for humans: "killer robots", drones,
automated ships, ... (here, however, no moral filters for these AIs).
Libertarian capitalism
The US government is clearly the government of billionaires. While the
bourgeoisie has always dominated the state apparatus, the difference is
that billionaires hold key positions in the US government today-14
currently, of which E. Musk is the emblematic figure. If Trump was
supported by all the major tech companies, it is no accident. Trump's
libertarian and authoritarian vision is in line with the economic and
political projects of these trusts. Masculinist, racist, and a promoter
of unbridled capitalism, this movement offers a way out of the crisis in
which capitalism has been mired for several decades. The far-right
libertarian movement is gaining ground all over the world, popularized
by all the media owned by these trusts (the majority of newspapers and
digital applications). The goal is to be able to act and prosper without
hindrance, not even from a State that initially guarantees the long-term
interests of the bourgeoisie: remove all environmental legislation,
abandon the minimum wage, twist the neck of social protection, etc.
Tech companies have a predatory vision of our lives to impose themselves
politically and economically. They use the data produced by digital
users to format us, dictate our behaviors in order to optimize their
profits and channel us politically. This is the capitalism of permanent
surveillance. There is a trend of state supervision by this sector
because, beyond the control of the State and its data, this offers a
significant market to AI companies via administrative management by AI
to replace civil servants (as in the USA currently where thousands of
people are being laid off).
This political technocracy dons the garb of "science" to create an
alliance of AI and the State that thus generates permanent control over
the population, creating for some a new form of totalitarianism. Already
today, we can see what "AI + State" means: a Trumpist decree provides
for the use of artificial intelligence to exclude research projects,
those containing the words (now banned): antiracism, racism, altruism,
prejudice, DEI (diversity, equity and inclusion), diversity, diverse,
confirmation bias, equity, egalitarianism, feminism, gender, gender
identity, inclusion, inclusive, inclusiveness, injustice,
intersectionality, privilege, racial identity, sexuality, stereotypes,
pronouns, transgender, equality, ALT (alternative).
That said, if tech trusts are calling the shots, it's not without
contradiction. Logically, capitalist sectors don't want to let these new
billionaires preempt all positions, all resources, all power for their
own benefit, especially since tech represents only a small part of the
economy in terms of production, employment, and activity. The
agricultural, mining, industrial, and service sectors have become
accustomed to working with the state, diverting its resources to their
own benefit. Challenging entire sections of the state, disrupting the
system, is to risk unbalancing everything.
Conclusion
AI-led capitalism seeks to renew the capitalist profit rate by reducing
a fraction of the workforce (so-called intellectual labor), by
streamlining/optimizing decision-making, and finally, at the consumption
level, by opening up new markets. This is somewhat similar to what
Taylorism did with the "scientific organization of work," which was
first imposed on workers, before being adapted to employees and managers
in the form of "management." In both cases, the aim is to closely
control the intensity of work. AI is a new, extreme form of capitalist
exploitation, based on an equally extreme exploitation of natural resources.
Like every advance of this type (technological revolution), it is
accompanied by a new dispossession of knowledge and know-how (here in an
early stage, since AI is already massively used by pupils and students
to avoid learning and especially thinking for themselves), an even more
advanced degradation of the social bond, and propaganda serving to
destroy culture and common sense, under the pretext that the same
technology can more effectively detect the beginnings of cancer or
signals of an upcoming earthquake.
To do this, bourgeois "democracy" is superfluous for the ideal of a
world where everything is automatically regulated by algorithms: we must
be nothing more than passive masses, reacting to digital stimuli to work
and consume, monitored and without possible critical thinking. To do
this, we must manufacture a generalized consent based both on the
impossibility of escaping AI and the paralyzing stupor that AI produces
to make us believe that the future is already over.
Current investment announcements look like a headlong rush: $500 billion
in the USA, EUR150 billion in France, and the same in China, essentially
for gigantic data centers needed for AI; not to mention, for example,
the project for the world's largest cable (50,000 km) launched by
Meta... And this, without knowing whether these investments will turn
into profits and therefore whether they will not lead to a classic
crisis of capitalism, or simply to a financial crisis stemming from the
current speculative bubble in the shares of these AI companies. Not to
mention warlike crises arising from imperialist tensions (with the
significant growth in arms budgets). It is therefore authoritarian
capitalism but also turbulence that AI promises us.
We are not helpless and can be the vector of this turbulence. We can of
course start by boycotting everything related to AI as much as possible,
but it's a bit like peeing in the shower to combat global warming: AI
should be attacked structurally. We must first claim to have critical
autonomy, popularize the catastrophe generated by AI (human,
environmental, political, warlike, etc.), and participate in its
protest. Above all, AI is, above all, human work, and therefore
potential associated revolts. Because people do not spontaneously accept
their new absurd working conditions (human employees in the service
sector, responsible for "smoothing out" texts produced by the machine),
or downright ignominies (machines responsible for the care and education
of children, the administration of solidarity, etc.). It is we, the
proletarians, who suffer but above all who make AI work, not the tech
leaders, and therefore it is up to us to collectively destroy it through
sabotage, strikes, and above all the expropriation of all these parasites.
RV
Notes
[1]Thibault Prévost's recent book "The Prophets of AI - Why Silicon
Valley Is Selling Us the Apocalypse", published by LUX, effectively
discusses this.
[2]Sorry for the lack of references in the article, but there are far
too many to include.
[3]We will use this term to generically designate companies that develop
their profits based on new digital technologies, including AI.
[4]See CA 345: " Digital Barbarism. The Criminal Exploitation of
Technological Metals in Congo (DRC) "
[5]The documentary "The Sacrificed of AI", freely accessible on France
TV, reports on the working conditions and psychological damage to these
invisible AI workers.
[6]The United States demanded control of 50% of Ukrainian minerals
without guaranteeing military protection, hence Zelensky's refusal in
February 2025. To force Ukraine to accept, Musk reportedly threatened at
that time to deprive Ukraine of its Starlink satellite network, which is
absolutely necessary for the Ukrainian army today. At the time of
writing, however, an agreement seems to be emerging on the management of
Ukrainian rare materials for the benefit of the USA.
http://oclibertaire.lautre.net/spip.php?article4456
_________________________________________
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
Subscribe/Unsubscribe https://ainfos.ca/mailman/listinfo/a-infos-en
Archive: http://ainfos.ca/en
- Prev by Date:
(en) France, UCL Leaflet: The Plane Destroying the Forest (ca, de, fr, it, pt, tr)[machine translation]
- Next by Date:
(en) Italy, Sicilia Libertaria #460: Ahmed: The Mafia as a Metaphor of Power (ca, de, it, pt, tr)[machine translation]
A-Infos Information Center