AI is pervasive throughout society under chatbots, models, and agents that are decidedly reactionary and create communication and physical walls to protect the status quo.
Capitalism has chosen AI as its next tool for distributing and dismantling labor, building new power structures in the world, and suppressing social and political movements. If left unchecked, we will face collapse brought about by war and climate chaos.
Forget the Terminator stories about artificial general intelligence and artificial superintelligence. These are more science fiction than reality. There’s no need to speculate about things that don’t exist in the realm of AI. What we need to focus on is what already exists and is being deployed at scale.
The main purpose of AI is to automate historical automation itself. AI holds irresistible promise for capitalist elites. It will be able to automatically dictate most of the instructions that guide human activity, reducing the power of social classes other than those that own the algorithm. A complete economic and social plan for the wealthy. In particular, they want to reduce the power that the working class has wielded in the past, to move forward into the future and to obtain political, social and economic changes that reduce or eliminate inequality and injustice.
Today’s data centers are nightmare factories.
An important and complementary goal of AI is to create an overwhelming monopoly on knowledge codified through large-scale language models, computer vision, convolutional neural networks, and other machine “learning” models. This monopoly is designed to completely transform social relations and establish a reactionary hegemony that goes well beyond neoliberal capitalism and feeds a far-right dystopia.
A third important objective concerns the control of violence and political repression. To its effect, AI provides a variety of tools for use in states of declared and currently mostly undeclared wars. In the Gaza massacre, human targets were selected by AI, and the model was used to determine the maximum impact on a set of targets to achieve maximum infrastructure and human damage. Clearly, AI is being used to maximize the efficiency of all war logistics, including calculating payloads, schedules, and distribution of supplies. In Ukraine, much of the war is being fought with drones, many of which are autonomous and equipped with AI-powered self-selecting targeting capabilities. Automated killing machines with no command or goal are not only available, they are already deployed on a variety of fronts. Meanwhile, automated political repression and persecution on the streets and in protests is increasing, but is currently in the data collection and training stage. In the United States, Immigration and Customs Enforcement has deployed a variety of apps developed by companies like Palantir to maximize social disruption and capture the nation’s most vulnerable populations.
There is significant pressure to prevent meaningful regulation of AI, particularly AI used by police and the military. Facial (FRT) and body recognition surveillance is used outdoors to plan the movements and participants of protests and actions. Mapping connections and collaborations in movement can be done not only on the street, but also through pattern recognition online. Automatic protest suppression coupled with deliberate misinformation and disinformation can make regular protests completely unviable.
And of course, AI is being used for hacking by private companies and nation states. Given the hackable systems currently in place throughout society and the economy, such as banking systems, social security, electrical and transport systems, aviation and navigation systems, pension administration, surveillance equipment, medical systems, and of course all the Internet and data in public and private servers, large-scale disruptive events, large or small, are inevitable. Many political and social movements will be targeted. This could mean increased political repression of individuals by erasing their accounts, blocking their financial assets, and curtailing their ability to communicate. This can also occur on a larger scale, covering entire cities, countries, and regions.
For the most significant investments and political initiatives, AI is being deployed as a replacement tool for labor, a creator of cultural hegemony monopolies, and a military and surveillance weapon. Most of this is done through people actively involving and inviting models into their daily lives (now more than ever). While resistance to large-scale data center projects is important and encouraging, the overwhelming threat of AI goes far beyond emissions, water consumption, and land occupation (although there are plans, particularly in Europe, to double current numbers on many factors). Today’s data centers are nightmare factories.
So far, AI has not been successful in automating a key aspect: processes that could be replaced by effective algorithms to enable mass layoffs. This is clear. 95% of investments companies make in AI are not profitable, and capitalists are worried. However, we were unable to meaningfully stop its spread.
When we say AI, we mean machine learning, robotics, and expert systems. Currently, AI is primarily a process of recognition, classification, and computation with very high probability based on large amounts of data with a good human interface. The interface is the most important trick for the average person. The public debate surrounding this issue is so anti-historical and anti-materialist that it is almost entirely white noise.
AI does not copy or reproduce human intelligence. We are trying to encode human activity into repeatable steps that can create reproducible algorithms. It’s not trying to imitate our biological intelligence, it’s trying to imitate what we more or less “understand” about the aforementioned algorithms. That is, copying labor and social relations, their mechanisms and predictable outcomes. Like other abstractions that govern our lives, such as money, algorithms produce real results. AI offers irresistible promises for capitalist elites. It is about being able to automatically dictate most of the directions that guide human activity, reducing the power of social relations, especially the power of the working class to impose political, social and economic changes that reduce inequality and injustice.
AI’s neural networks do not imitate the human brain in any way; they automate the “perceptual work” of classifying, interpreting, and establishing associations between textual, numerical, and visual data. This creates a synthesis of knowledge obtained from social cooperation, a collective form of knowledge. As explained earlier, another of its aims is to establish a monopoly on the knowledge scraped and sourced from all websites, databases, and online encyclopedias. So it’s no wonder Elon Musk and the far right pursue Wikipedia.
These are some of the reasons why trying to hard-code ethical rules and constraints into these models doesn’t work. Because they do not change the data used for training or the underlying political and economic functions of the algorithms that are generated and built. Of course, we understand that language itself, all data, and of course the internet are algorithms. But with AI, we’re talking about a new level of control. The fundamental abstract objectives of AI as it currently exists are quantification, control, and enhanced exploitation. The labor theory of automation posits that AI is the result of a series of technological advances that abstract automation to a level where automation is possible. Since we now have the technical ability to create such machines, and capitalism has the economic incentive to deploy them on a large scale, it wants to use this to further reorganize the division of labor in its favor. It is the pinnacle of automation: the automation of automation.
In the face of these seemingly insurmountable challenges, social and disruptive movements have no choice but to ask what to do about AI. There are basically two options. Either we fall off the grid, or we gain the technological strength to resist the onslaught of collapsing algorithms.
