Opinion 2

Thesis

The abstraction of AI is a political choice designed to create reverse-centaurs and make democratic self-government seem impossible.

Core Claim

AI is often presented as an autonomous, inevitable, abstract force beyond human control.

This framing is political: it discourages regulation and democratic contestation.

In reality, AI is materially built, socially organized, and intentionally deployed in ways that subordinate human judgment.

Schneier + Mumford

Technologies embody forms of power and authority.

Physical infrastructures can be contested because they are visible and concrete.

• during all stages of development: creation, adoption, adaptation

Presenting AI as invisible code or neutral intelligence hides the political choices embedded in its design.

If AI appears to “just happen,” elites can avoid accountability.

Crawford

Argues AI is not abstract at all.

It depends on mines, data centers, labor exploitation, environmental extraction, and global supply chains.

It also has material consequences: energy use, labor displacement, surveillance expansion, and workplace restructuring.

Calling AI immaterial conceals these costs.

Doctorow

Contrasts genuine human-machine cooperation with systems built to invert that relationship.

A true “centaur” model uses tools to enhance human judgment.

A reverse-centaur turns the human into a peripheral serving machine commands.

Example: the Amazon driver follows algorithmic instructions so completely that the person functions as an appendage to software.

Mumford

distinguishes tools, which remain subordinate to human purposes, from machines, which reorganize humans around machine logic.

Reverse-centaur AI represents the machine form: humans adapt themselves to thinking alongside the encroaching presence of AI

Democratic Consequences of AI

Democracies depend on citizens capable of judgment, deliberation, and self-rule.

If people are habituated to obey opaque systems, their civic capacities weaken.

AI systems can settle political questions in advance—about efficiency, risk, hiring, policing, productivity—without public debate.

Decisions appear technical rather than political.

Schneier

Warns that privacy is necessary for free thought and experimentation.

AI-driven data extraction and behavioral monitoring invade the private sphere where opinions are formed.

This creates a public-private surveillance partnership: corporations collect data, states may access or share it, and both gain power.

Citizens become more legible.

Tufekci

Notes a democratic paradox: technologies that empower mass participation can also enable mass surveillance.

AI can increase access and efficiency while simultaneously centralizing control.

The same tools that help people organize can also monitor dissent and automate inequality.

Why Abstraction Matters

If AI is seen as too complex or inevitable:

1. Individuals feel incompetent to challenge it.

2. Corporations monopolize expertise, power, and agenda-setting.

3. Governments defer to private actors rather than regulate them.

This produces unaccountable sovereignty inside nominal democracies.

Conclusion

AI is not an uncontrollable external force.

It is a material political system designed by institutions making contestable choices.

Citizens must reject the myth of abstraction, recognize AI’s material consequences, and reclaim democratic authority by refusing to become reverse-centaurs.