Power and Progress: How Technology Serves Power or the People

🇵🇱 Polski
Power and Progress: How Technology Serves Power or the People

📚 Based on

Power and progress
()
PublicAffairs

👤 About the Author

Daron Acemoglu

Massachusetts Institute of Technology (MIT)

Daron Acemoglu is a Turkish-American economist specializing in political economy, development, and labor economics. He is an Institute Professor at MIT, and was awarded the Nobel Prize in Economics in 2024. His works include 'Why Nations Fail' and 'Power and Progress'.

Simon Johnson

MIT Sloan School of Management

Simon Johnson is a British-American economist and the Ronald A. Kurtz Professor of Entrepreneurship at MIT Sloan. He was the Chief Economist at the IMF (2007-2008). Johnson co-authored 'Power and Progress' and 'Jump-Starting America'. He received the Nobel Prize in Economics in 2024.

1. The AI Illusion: The Primacy of Autonomy Over Machine Usefulness

This article analyzes the fundamental dispute over the direction of technological development: whether it should serve machine autonomy or the enhancement of human capabilities. The AI illusion is the drive to create systems that replace humans, where the individual becomes merely an "error in the process." The alternative is machine usefulness—designing machines as tools that increase human agency and competence rather than acting as their substitutes.

2. So-so Automation: Low Productivity Destroys Jobs

The currently dominant so-so automation displaces workers in low-value tasks without generating a leap in productivity. Such innovation destroys jobs but fails to create new, well-paying roles, leading to economic stagnation.

3. The Productivity Bandwagon: The Illusion of Universal Prosperity

We must debunk the myth of the productivity bandwagon, which assumes that increases in efficiency automatically raise wages. The locomotive of technology can pull away on its own if it is not coupled with institutions that ensure a fair distribution of profits and worker participation in success.

4. Marginal Productivity: The Real Measure of Worker Value

Marginal productivity is the real increase in value resulting from the work of an additional person. If technology makes humans a marginal addition to machines, the demand for labor falls and wages remain stagnant, even as overall corporate profits skyrocket.

5. The Vision of the Elites: A Monopoly on the Direction of Technological Innovation

The current direction of innovation is a vision of the elites, for whom "autonomy" is a slogan for building capital rents. This selection pattern ensures that funds flow into projects that replace people instead of strengthening craftsmanship and collaboration.

6. The Advertising Model: The Algorithmic Erosion of Democratic Debate

The digital advertising model monetizes human attention through surveillance. Algorithms prioritize polarizing and anger-inducing content, which destroys the deliberation necessary for democracy, replacing it with a stimulus-and-response loop.

7. Countervailing Forces: A Social Brake on Technocratic Power

Countervailing forces, such as labor unions and antitrust laws, are essential to impose constraints on technological elites. They act as the ABS system that prevents innovation from becoming a tool of expropriation.

8. Medieval Mills: A Lesson on Monopoly and Technical Coercion

The history of medieval mills teaches us that technology is not neutral. Although they increased efficiency, they served exclusively as a tool of monopoly for feudal lords, failing to improve the lives of the 90% of the population deprived of the rights to progress.

9. The Industrial Revolution: A Century of Exploitation Before Wage Growth

In the Industrial Revolution, productivity growth only brought widespread prosperity after decades of struggle for democratic reforms. Without countervailing forces, machines served only factory discipline and the degradation of workers' quality of life.

10. AI Scenarios: Between Market Monopoly and Regulation

The future of AI fluctuates between corporate monopoly and institutional breakthroughs. The key is whether organizations will redesign work around new tasks or limit themselves to cost-cutting, risking political destabilization.

11. Asymmetric Complementarity: Machine Dominance

Modern asymmetric complementarity is a trap: the human remains in the process, but only to be held responsible for the errors of a system over which they have no real power or right to audit algorithms.

12. Taxes and Regulations: Enforcing Machine Usefulness

Digital advertising taxes and strong competition laws are indispensable. These are instruments that can forcibly change the incentive system, compelling companies to invest in tools that enhance worker competencies.

13. The Oligarchy of Vision: Silicon Valley’s Dictate Over the Future

An oligarchy of vision controls both financial and narrative capital, imposing a belief in the inevitability of automation upon the world. The counterweight must be institutional pluralism, allowing for the testing of different developmental paths.

14. Machine Usefulness: Pillars of a New Technological Order

The machine usefulness program is based on four pillars: increasing human agency, creating new tasks, providing reliable information, and building open collaboration platforms instead of surveillance systems.

15. False Teleology: Technology is a Choice, Not Destiny

We must reject false teleology: technology has no goals of its own and does not inevitably lead us to a better world. In the age of algorithms, the crucial question becomes: is technology meant to empower us or control us? The answer will determine whether the future is a manifestation of our creativity or merely an echo of our fears. Progress is not destiny, but the result of institutional choices.

📄 Full analysis available in PDF

📖 Glossary

Machine usefulness
Koncepcja projektowania maszyn jako narzędzi zwiększających ludzkie możliwości i kompetencje, zamiast zastępowania człowieka w procesie pracy.
So-so automation
Automatyzacja „taka sobie”, która jest wystarczająco wydajna dla firm, by zastąpić pracowników, ale zbyt słaba, by wygenerować ogólnogospodarczy wzrost produktywności.
Produktywność krańcowa
Realny przyrost wartości wynikający z zaangażowania dodatkowej jednostki pracy, który determinuje popyt na pracowników i wysokość ich wynagrodzeń.
Siły przeciwważne
Instytucje i organizacje, takie jak związki zawodowe czy regulacje antymonopolowe, zdolne do równoważenia wpływów elit technologicznych.
Iluzja AI
Przekonanie, że technologia powinna dążyć do pełnej autonomii na wzór ludzki, co często prowadzi do marginalizacji roli człowieka i jego odpowiedzialności.
Instytucje ekstraktywne
Mechanizmy społeczne i prawne służące do przejmowania nadwyżek wypracowanych dzięki technologii przez wąskie grupy posiadające władzę lub kapitał.

Frequently Asked Questions

How is the AI illusion different from machine usefulness?
The AI illusion strives for technology autonomy at the expense of humans, treating them as a bug in the system, while machine utility enhances human abilities, making the machine a tool rather than a substitute.
Why does productivity growth not always lead to higher wages?
Wage growth only occurs when technology increases the marginal productivity of humans. If automation merely replaces human labor with capital, wages may stagnate despite increases in firm productivity.
What is the so-so automation phenomenon?
This is the type of innovation that displaces workers from their tasks but does not generate a sufficient leap in productivity to create new, well-paid jobs in other sectors of the economy.
What role do countervailing forces play in the development of technology?
Countervailing forces, such as trade unions and legal reforms, are necessary to ensure that technological progress benefits society as a whole by enforcing a fair distribution of profits and participating in defining the goals of innovation.
How does the advertising model influence the direction of AI development?
This model rewards algorithms that capture attention through polarization and emotional manipulation, which often runs counter to public interest and promotes innovations that serve data extraction rather than real utility.

Related Questions

🧠 Thematic Groups

Tags: Power and progress AI illusion machine usefulness so-so automation marginal productivity countervailing forces Daron Acemoglu human-in-the-loop corporate accelerationism extractive institutions productivity wagon revenue model automation demand for labor innovation