Algorithmic Erosion: AI, Culture, and Social Relations in Contemporary Capitalism

Abstract: Artificial intelligence and algorithmic governance reorganize not only work, but also language, trust, friendship, identity, access to information, and the capacity for collective action. AI, however, is a tool rather than a destiny. Its social direction is determined by ownership, structures of power, economic objectives, and decision-making institutions.

Technology Within Social Relations

Two simplified stories dominate public debate. One presents AI as an automatic solution to social and economic crises; the other treats it as an independent and inevitable force that will destroy culture and human relations. Both hide the roles of people, ownership, and power. AI systems are the result of collective labor by researchers, programmers, data workers, content producers, users, and infrastructures often built through public investment and knowledge.

Society participates in producing data and knowledge, while control over data, computing power, chips, cloud infrastructure, models, and distribution channels remains concentrated in a small number of companies. This gap between the social production of knowledge and the private concentration of ownership and decision-making power is the point of departure for the analysis of algorithmic erosion.

Mechanisms of Algorithmic Erosion

Surveillance and the Reduction of People to Profiles

Data-driven systems translate human behavior into features, probabilities, and scores. These profiles can influence visible content, prices, employment opportunities, access to credit, and the intensity of surveillance. Harm grows when surveillance is invisible, continuous, and difficult to contest – when a company knows a great deal about an individual while the individual knows little about the logic of classification or the purpose of data collection.

Algorithmic Management and the Individualization of Risk

Algorithms in platform work, retail, warehousing, transportation, and professional occupations can assign tasks, measure speed, schedule shifts, score performance, and restrict access to work. Variable income, constant connectivity, and automated suspension transfer enterprise risk to the individual and turn a common problem into personal competition.

Reputation and the Engineering of Attention

Ratings can facilitate cooperation among strangers, but also reduce trust to a calculable and manipulable number. Recommendation systems decide which content is emphasized, hidden, or repeated. When success is measured by engagement, emotional and divisive material may gain an advantage. The research evidence is not uniform, however, and credible analysis must distinguish correlation, causation, and structural interpretation.

Economic, Social, and Cultural Consequences

Platform control through data, network effects, and centralized rules can move markets toward monopoly. Higher productivity does not automatically produce universal welfare. If gains remain privately owned while life security remains tied to full-time employment, the same technology that produces more wealth with fewer workers can intensify unemployment, weaken bargaining power, and increase insecurity.

The commodification of social infrastructure changes the incentives of interaction. Followers, likes, and reposts can displace relationship quality and slower dialogue. At the same time, platforms have enabled new forms of support and organization for migrants, minorities, dispersed groups, and social movements. Critical inquiry does not deny these opportunities; it examines how opportunity, risk, profit, and decision-making power are distributed.

Strategies for Reducing Harm

• Limit disproportionate surveillance and prohibit unrelated uses of sensitive data.
• Disclose evaluation criteria and guarantee understandable explanations and human review.
• Include workers and users in the assessment and design of high-impact systems.
• Give unions, workplace councils, and civil organizations access to relevant information and independent audits.
• Develop public infrastructure, platform cooperatives, data trusts, and a social share of digital wealth.
• Convert productivity gains into shorter working hours, public services, social security, education, and culture.

Connection to A New Social Order in the Age of Consciousness

The book treats AI as a leading expression of the socialization of knowledge production. Millions of people produce data, language, images, experience, and feedback, yet their right to participate in decisions about this collective product is not proportional to their contribution. The desired transition is from participation in producing knowledge to participation in power.

Such participation is impossible without freedom of expression, media, research, association, unionization, strike, and assembly. Data-driven planning without these freedoms can make domination more precise. AI should assist analysis, transparency, and decision-making; it should not become a source of political legitimacy or a substitute for democratic judgment.

The decisive question is not what AI will do to humanity, but how society – through its structures of ownership and institutions of decision-making – will direct this capacity and for whose benefit.

Back to the Home Page