Despite the global Uber-hype about artificial intelligence (AI), there still is what only “we” have: natural intelligence (NI). NI allows us to deal with reality in terms of abstractions, predictions, data, information, and knowledge.
NI is a representation of cognition and reasoning, understanding and learning, problem-solving and decision-making. It allows us to do what makes us human: the ability to interact with other people. Thanks to NI, we can engage with our social, physical, cultural, and emotional environment.
By contrast, AI is a computer generated intelligence. AI still is a machine, a technology, and a tool. At its core is an algorithm that – more or less – calculates statistical probabilities. In short, the algorithm finds the most likely answer to “some” of the questions.
It also finds the most likely route to a specific location. Armed with such capabilities, we have been made to believe that it is “intelligence”.
Put simply, AI enables computers – calculating machines that use a code – to “simulate” human intelligence. Virtually all of this and despite some significant advancements is what furnishes AI with – still rather limited – problem-solving capabilities.
Basically, AI is a symbolic-logical system that statistically – with the aid of mathematical algorithms – suggests solutions to an (often rather simple) problem like “what soap to buy”. But perhaps not so much on “what is the meaning of life”.
Unlike NI, AI is not created by nature. There never was a natural evolutionary mechanism that resulted in AI. In contrast to AI, NI is the “biological intelligence” embodied in our brain.
As so often, AI is neither invented nor applied in a vacuum – nor is it neutral. Instead, AI – like most machines – is inextricably linked to the system in which it emerged: capitalism.
Yet, AI is anticipated to reshape the economy by transfiguring the human interaction with technology.
In that, the tasks assigned to AI cover machine learning, business reasoning (read: as directed towards profits), corporate problem-solving to assist management, understanding, and perhaps invent an ever more complex managerial language, etc.
In any case, AI systems, as used in companies, can analyse large commercial data sets, recognize sales patterns, and reach or support business decisions.
In relatively closed systems like chess, Go, and algorithmic share trading (algo-trading), AI has exceeded human capabilities in both accuracy and speed.
In the world of work, AI technologies can automate “some” managerial tasks, streamline operational processes, and optimize resource allocation.
This, in turn, will lead to increased “productivity” (read: more corporate profits) across various sectors of the economy. AI also fosters innovation by enabling the development of new products, services, and business models.
Meanwhile, AI also has the likelihood to improve human society through advancements in healthcare, transportation, education, trade union organizing, and creating environmental sustainability, leading to better overall welfare.
Notwithstanding AI’s undisputable range of applications, political-economics has – up until today – given limited attention to the economic effects, negativities, and even pathologies of AI.
Simultaneously, AI is rapidly becoming a hot, if not sexy, topic capable of attracting considerable attention. In the area of academia, for example, a threshold of fifty academic articles per year on AI was passed in 2017.
Academic writing in economics was outdone by studies in the areas of business and management. Even for business school professors, AI is, well: good business. By assisting corporations in enhancing profits, these business school publications focus on three key issues:
AI’s impact on resourcefulness and innovation that can drive corporate revenue.
AI’s role in sustainable business growth (read: not necessarily “environmental” sustainability), but productivity and its economic benefits (read: maximising corporate profits).
The money-generating integration of AI technologies into the economy and companies.Meanwhile, economists have also analysed the link between robotization and AI and its impact on two issues: economic growth and the income of workers.
This includes automation, the corporate application of industrial robots for routine tasks, and the displacement of low-skilled workers. For many neo-liberalism economists, for example, the unwavering maxim is:
“labour replacing technologies improve productivity”.
In short, one of the key lessons is – and this comes despite robots and AI – that low-skilled labour continues to be a significant factor in many production systems.
AI will hardly change work in areas such as, for example, agriculture, construction, bricklaying, beauticians, etc. In other words, there are plenty of jobs where the cost of a robot still exceeds that of a worker.
At the same time, in many industries, a worker’s job prospects will be influenced by how AI technologies interact with workers. In short, AI can replace workers or assist and support workers in their daily work.
On the financial side of AI innovations, venture capital is cranking up new entrepreneurs in the global AI sector that generates about $200bn a year and is predicted to have a strong growth in coming years.
Up until today, AI-based start-up businesses still raise “less” venture capital compared to non-AI startups. At the same time, there is a widespread funding gap in financing AI-based start-ups to nurture AI ventures and innovations.
For economics, the importance of a seemingly “omnipresence of AI” guides an ever-increasing focus on research into the “economics of AI”.
This includes the largely unexploited potentials of AI like its problem-solving and decision-making capabilities. In the area of labour economics, the three key issues are:
how does AI-based automation and computerization impact on the skills of workers,
the impact of AI on employment and unemployment, and
the influence of AI on a worker’s income – potentially leading to more income inequality.In overall economic terms, AI capitalism has a two-fold impact on employment and workers: AI can substitute or complement labour.
Predominantly, AI impacts on workers engaged in cognitive and manual tasks. This is particularly the case when work tasks are executed by following explicitly defined worker rules. Secondly, AI will also have an impact on workers engaged in non-routine problem-solving tasks.
Meanwhile, a somewhat gloomier outlook emphasizes AI’s role in reducing employment as well as the labour share in producing economic value. Indeed, AI has the potential to reduce labour demand as well as wages. This could lead to what economists know as technological unemployment.
Simultaneously, AI might also enhance the polarization of the labour market. This occurs when wage increases disproportionately benefit workers at the extreme ends of the income and skill distribution scale. It intensifies inequality.
Yet, there is also a chance that AI will have a balancing impact by cranking up corporate outputs in such a way that might result in a greater demand for labour.
The predominant view of economists seems to be that AI’s displacement effect is moderated through the emergence of a range of new professions and jobs.
This might well come at the advantage of workers particularly for those working in areas that demand specific problem-solving skills, a high degree of adaptability, and in an area that AI has, so far, comprehensively failed on: creativity.
However, others propose that the combination of corporate innovation and the enhancement of corporate productivity among low-skilled workers will ease the unfavourable impacts of AI technologies on labour working at the lower end of the skill and income pyramid. In other words, many economic growth theories propose “divergent results” in the short term. But they also tend to predict a more optimistic outlook for labour in the long run.
Some even suggest that productivity growth that result from general purpose technologies like AI is currently underestimated. They also suggest that others seem to undervalue the momentum of AI investments that is, in general, not accurately measured, and therefore, remains unrecognised.
Meanwhile, AI-based automation can contribute “positively” to economic growth – the ultimate measure of neoliberal economists.
Other economists even predict that a rather speedy AI growth beyond a certain threshold will accelerate economic growth through an increasing level of AI-based improvements in an economy. Indicators for this are the impacts of AI on marketing, design, and operations.
Consequently, this will crank up “algorithmic competition”. AI will be used to turbo-charge capitalism and it will also crank up the economic and political influence and power of corporations.
This power will be used to manipulate politics for pro-business support (state subsidies) and pro-business regulations. This, ultimately, will lead to enhanced profits.
As almost always in capitalism, there are already shifts towards concentration and collusion in the global IT industry. It is Big Tech – whether GAFAAMT or GAFAM. This occurs despite (read: because of) antitrust and other state regulations.
The raison d’etre of anti-trust rules is to camouflage the power of corporations through making people believe in free market competition. In other words, centralization and concentration in capitalism are camouflaged through an unwavering belief in free market competition.
Worse, the apostles of the “free market” (Hayek) are also those who support autocratic regimes. Free market advocates like Hayek despised any democratically elected majorities as Hayek’s love for Pinochet showed.
Beyond that, authoritarian regimes never shy away from even using corporate online platforms to fight and even destroy civil and democratic organizations.
In this context, corporate AI can wield enormous power of social control simply by manipulating data and by the dissemination of information about groups and individuals and their behaviour.
Moreover, AI also allows for “bundling”. This is the assembling of information on individuals depicting so-called “pro”-social (read: regime supportive) behaviour with other data (like consumer behaviour) and political attitudes.
Supercharged by AI, authoritarian regimes as well as undemocratic corporations can create something that is known as: “guilt by association”. Since their algorithms work by associating data, the “guilt by association” is an extremely short step.
AI’s algorithms are associated to a massive amount of data. This, too, can be a most effective apparatuses for the domesticating and regimenting of a society. In short, algorithms can use data produced by society and analysed by algorithms against society.
In other words, AI and autocratic regimes as well as authoritarian corporations mutually reinforce each other for the surveillance of society (states) and workplaces (companies).
Hence, Uber-surveillance AI as used in corporations, authoritarian regimes, and ultimately in capitalism can co-evolve inside a mutually useful bond. This “bond” is based on something that might best be called “interest symbiosis”. There are no more smoke-filled backrooms of evildoers and secret conspirators.
Meanwhile, AI also remains a useful tool for marketing, logistics, business, information management, and engineering.
On the other hand, AI will also impact employment, labour markets, productivity, growth, and competition. Under “free market” competition and powered by AI, petrol station, for example, can raise prices simultaneously even more efficiently.
Beyond all that, there are still the authoritarian potentials of AI that can endanger the freedom of individuals because of a concentration of data and knowledge in the hands of an authoritarian regime and an equally authoritarian company or corporation.
Yet, some Uber-optimistic views imagine a future AI society of abundance, free goods, high environmentalist standards, and almost zero marginal costs of reproduction.
Beyond capitalism may even be something called AI-Post-Capitalism, digital socialism and the even more famous fully automated luxury communism. A bright AI future awaits all.
Next to that, linking AI to futurism might be a bit of a wasted energy because futurism tends to suggest technological determinism. Futurism carries authoritarian connotations.
Seemingly, it resents the possibility of non-techno-determined future with undefined pathways that are open to the shaping of future via non-technological but human intervention. Instead, deterministic AI implies a doomed humanity – the AI-driven killer robot is coming.
Unlike futurism’s pre-determined dystopia, the future development of AI is mutually dependent on several mechanisms that shape AI within a specific economic order – that of capitalism. In addition, there are also nation-specific institutions that shape capitalism and AI in somewhat different ways.
Yet, even the most neoliberal version of capitalism has what economists like to call “coordinating agents”.
Next to imperative state institutions that issue money, provide a secure legal framework for commerce and trade, safeguard intellectual property, issue ISO standards, and so on, these “coordinating agents” are found in today’s managerial class. Overall, the managerial class can be divided into two sub-divisions:
Apostles: There are the apostles of neoliberal capitalism furnished with a quasi-religious belief into the ideology of neoliberal capitalism. These are employed in the outer ring of managerial capitalism. They, for example, are working in the one-trillion-dollar worth management consultancies ($1,000,000,000,000), the 15,000 global business schools, corporate think tanks, pro-business research institutions, the business press, etc. They work in the outer ring that supplies stabilising ideologies for managerial capitalism. And then there are the,
Apparatchiks: Corporate apparatchiks are working in the inner ring, e.g. companies and corporations. More often than not, these work as middle- and top-managers as well as CEOs of corporations rather than line-managers. While the apostles are more involved in ideology-creation, corporate apparatchiks are more engaged with using ideology to legitimize three things (in that order!): a) themselves, b) the corporate hierarchy and power enforced inside companies and corporations (e.g. Big Tech), and c) to camouflage the environmental, human, and other pathologies that corporations create. At the level of society, the apostles and at the level of corporations, the apparatchiks push the ideology of liberal market economies of laissez-faire capitalism. In this hallucination, the free market will create and push AI. In reality, AI isn’t so much shaped by the beloved free market of neo-liberalism.
Instead, AI is shaped by the “Big 5” of Big Tech of Apple, Meta (Facebook), Amazon, Alphabet (Google), and Microsoft (Windows, Github, Xbox, Skype, etc.):
AI might not be formed by the illusive free market at all but by a democratic (EU, USA, etc.) or an authoritarian state (Russia, China, etc.) regime that supports capitalism and corporate AI.
Even within authoritarian capitalism, the known varieties of capitalism might still nurture slightly different politically, regulatory, legally, and socially determined forms of AI.
Whether operating in LME or CME, the interdependence of corporate and research networks with formal as well as informal structures (read: the old-boys’ network, collusion, business sleaze and corruption) also impacts on the way AI is introduced and used.
Yet there are, almost inevitably, also capitalism-supportive regulations and laws that shape AI such as, for example, the EU’s AI regulation. Beyond everything said above, there are five overarching scenarios that link AI to capitalism:
Free Market Neo-Liberalism: AI will crank up an Uber-capitalist future with vibrant AI entrepreneurs existing in a neoliberal free market system. This, so the assumption goes, would incentivize innovation in the AI area.
Big Tech & Big Money: Alternatively, the monopolistic Big Tech concentration will be turbo-charged by AI resulting in even less than five large hierarchically structured corporations. These monopolies have ample financial cloud to push and finance AI innovations.
Networks: In this version, corporate networks will harness the organizational power of command, coordination and control that reside in corporate networks. The corporate networks consist of groups of business organizations that work in conjunction with state institutions and public universities to foster AI.
Using the State: AI might enforce some sort of legal, political, and economic planning that is conducted by a state as a single central entity. This entity takes most of the decisions on AI and its development (read: China). Ideologically, this is supported by a kind of national mission that shapes AI policies.
Avoiding the State: in this model, an unregulated or minimally regulated market allows AI organisations and Big Tech (read: corporate lobbying) to escape, bypass, or evade regulation.
In all of these cases, AI has serious consequences for the corporate exploitation of workers through the use of the technological and managerial opportunities AI offers, e.g. algorithmic management.
In the end, AI – as with any other technology, tool and machinery – exists in the context of capitalism. Because of this context, AI is likely to be used to enhance corporate profits rather than human wellbeing.
Often that comes at the expense of workers and the global environment as AI is extremely power hungry. AI’s energy consumption is mind-numbing.
Yet, there are, as mentioned above, five different scenarios in which companies and ultimately capitalism are motivated to use AI for their all too often non-human (read: profit-maximising), anti-human (read: AI supported corporate surveillance), and perhaps even dehumanising benefits. In the end, AI under capitalism might take slightly different forms and shapes.
What is certain is that AI will be used to serve the interest of companies, corporations, and capitalism.