Artificial Intelligence: Technology of Capital?

Artificial intelligence could make the economy immensely more productive in the coming years. But the crucial question is not how big the cake will be. It is: Who gets it? – A comment by Mohannad Lamees.

Artificial intelligence (AI) is repeatedly sold to us as a technological revolution. In fact, it is a gigantic business. Behind the race for ever more powerful models are capitalist companies that promise themselves a simple advantage: more production with less labor, lower costs and higher profits.

Anthropic, the US company behind the Large Language Model (LLM) Claude, has now calculated where this development could lead. In onesimulationfor the US economy through 2030, the company is studying how AI could impact growth, employment and the distribution of income. Anthropic records three possible scenarios, which differ depending on the progress of AI technologies. However, depending on the scenario, the effects of AI differ mainly quantitatively; according to Anthropic, the fundamental developments are the same.

So Anthropic is sure that AI will ensure economic growth. In the extreme scenario, the gross domestic product (GDP) of the USA even increases by 32.4 percent compared to an economy without the additional AI effect. Even according to the less extreme predictions, AI still generates several trillion US dollars in additional GDP.

At the same time, the study says, the share of the value produced that goes to workers is falling. While workers today take home 60 cents per dollar produced in the form of compensation, in the extreme model it would only be 45.2 cents. According to Anthropic's calculations, the share that capitalists capture will increase from 40 percent today to almost 55 percent in the future - almost a complete reversal of the current ratio. The same applies here: the more moderate simulations only produce significantly smaller shifts, but the development is the same.

AI video surveillance at Kottbusser Tor: The next step towards a surveillance state

AI in capitalism

So the question is less whether AI can create wealth, but rather who this wealth belongs to. And this question cannot be answered without taking into account the fundamentals of the mode of production under capitalism. This means first of all recognizing that the AI ​​boom is not being driven forward by companies in order to improve the living conditions of workers and, for example, to create more free time.

It is driven forward because capitalist companies compete with each other. In capitalism, AI becomes a weaponcompetition of capital. Their development requires enormous amounts of computing capacity and thereby favors the concentration of technology in a few tech monopolies. These use AI to automate production, save labor and profitably exploit control over huge amounts of data.

What could increase productivity for society initially serves to reduce its own costs, drive out competition and acquire additional profits. That's why capitalists are investing billions in data centers, models and automation. Increased productivity is particularly interesting for them if it can be converted into profits.

AI thus exacerbates the fundamental contradiction between the interests of workers and capitalists. For workers, higher productivity potentially means less necessary working time. For capital it initially means the possibility of spending less on labor and getting more out of production. What could actually enable an improvement in working conditions becomes a threat to the job under capitalism. This is also shown by Anthropic's study, which predicts a tripling of unemployment in the USA in extreme cases.

The AI ​​question is therefore always a class question. Who owns the means of production decides how the new productivity is used and who benefits from it. The Anthropic scenarios clearly show that a society can become significantly richer through new technologies, but also how a greater share of this wealth is transferred to capital. This is nothing other than a reflection of the long-established capitalist distribution logic.

When old knowledge becomes new: How AI companies destroy books and hoard knowledge

Is AI threatening humanity?

While the economic consequences are already becoming concrete, a large part of the public debate is currently focused on a different scenario: the extinction of humanity by an out-of-control superintelligence.

The warnings about this are now becoming louder and louder within the AI ​​industry. The former Anthropic employeeJacob Coxonrecently warned, in the manner of a whistleblower, that the development of self-improving AI models could lead to the destruction of humanity.

Other AI researchers and developers also believe such an outcome is possible and are calling for the development of powerful AI to be more closely controlled and slowed down. Even the demand for stronger regulation has recently been joinedHeads of leading AI companiesin the USA.

Apocalypse as a distraction

Technology that is becoming increasingly autonomous and has capabilities that far exceed human intelligence can actually create real dangers. But the fixation on the end of humanity today has primarily a political effect: it shifts the focus from the damage that AI is already causing today to a hypothetical future monster.

The AI ​​critic also emphasizes thisTimnit Gebru. The idea of ​​an all-powerful, out-of-control AI distracts from concrete problems, namely increased exploitation, military use, expansion of surveillance, environmental damage and the interests of the companies that develop this technology.

Schwarz Group is planning an AI data center near Rostock – German economy in catch-up mode

The crucial question is actually not: Will AI kill us all? But rather, how we can succeed in putting the advances that are possible with AI at the service of the majority of people and not at the service of a few capitalists.

The answer is not determined by the algorithms. It is determined by the ownership and power relationships under which these algorithms are developed and used. A technology that generates profits for capitalists today can, under different social conditions, become a machine that improves the lives of billions of people.

The postArtificial Intelligence: Technology of Capital?appeared firstperspective.