Book review by Emphour

Power and Progress

by Daron Acemoglu and Simon Johnson

PublicAffairs, New York, NY, 2023.

“What if AI fundamentally disrupts the labor market where most of us earn our livelihoods, expanding inequalities of pay and work? What if its main impact will not be to increase productivity but to redistribute power and prosperity away from ordinary people toward those controlling data and making key corporate decisions? What if along this path, AI also impoverishes billions in the developing world? What if it reinforces existing biases—for example, based on skin color? What if it destroys democratic institutions? The evidence is mounting that all these concerns are valid.”[1]

 

Acemoglu and Johnson paint a bleak picture of the present and near future in their 2023 book Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. Neither are the suggested solutions effective nor are their goals desirable.

The book is meant to serve as a wake-up call for anyone who believes that the current trajectory of technology, especially automation and AI, will bring benefits to everyone without requiring any effort on their part. The authors don’t address the purposes and end goals of all this “progress,” and the inevitable negative consequences it brings. Acemoglu and Johnson argue (similar to how Acemoglu already did in his 2012 book Why Nations Fail) that strong democratic institutions are needed to ensure that the broad public gets their share of prosperity. As of right now these are insufficient and instead the visions of powerful tech entrepreneurs are shaping our world. Since around the 1980s, things have been constantly getting worse: US median real wages crumbled, while worker productivity still rose. Automation and offshoring raised productivity, but did not bring shared prosperity. The now two-tiered society is on course to worsen in the next decade. The book tries to offer solutions for these (relatively narrow, in the wider scope) problems.

The authors cover the last millennia of human history and pinpoint examples where new technologies brought better productivity but only at the expense of mass human suffering and only to the benefit of a small elite. (Although the general population tended to recoup these “benefits” later.) For example, during the Middle Ages food production got much more efficient, but the situation for the peasants got worse. The aristocratic and religious orders had the necessary verbal and manipulative powers to enforce their regimes which funneled the increasing productivity disproportionately in their favor. During the Soviet Union’s period of rapid industrialization, the mechanization of agriculture resulted in the state requiring ever more grain transfers from the farmers, constantly squeezing them dry.

“The soviet collectivization episode makes it clear that the specific way in which technology was applied was not just biased but also a choice.”[2]

After going through multiple examples in human history where it always applies, the authors lecture the reader that, in fact, the process of elite-profits-first, average-joe-later is not inevitable, but rather a choice. This is quite confusing. After seeing this shown as a common thread throughout history one can reasonably hypothesize, if not directly infer, that this is a fundamental social aspect that is extremely difficult if not impossible to change, and something to be expected in the future too.

As mentioned before, the authors are solely concerned with enabling everyone to “profit” from ever increasing productivity. They suggest a three-step plan for bringing in change:

“The first is altering the narrative and changing norms. The Progressives enabled individual Americans to have an informed view about troubles in the economy and society—rather than just accepting the line coming from lawmakers, business tycoons, and the yellow journalists allied with them. ...

The second is cultivating countervailing powers. Building on the change in the narrative and social norms, Progressives helped organize people into a broad movement that could oppose robber barons and push politicians to reform, including via labor unions.

The third prong is policy solutions, which Progressives articulated based on the new narrative, research, and expertise.”[3]

This may have worked for the progressives, whose aims coincided with the most efficient way a society works. The progressives did work against powerful individuals when fighting monopolies, corruption and the robber barons, but were only accelerating a pre-existing historical trend, not pushing against one. The current trend is heading in a direction that does not seem to be to the benefit of the general population at all. Also, decision-makers in human societies cannot simply do what they want. Their options for action are very limited. This is because the technology itself dictates what the society that introduced it must look like. Different political systems with the same technology are not as different from each other as a technologically advanced society is from a less technological one.[4]

One can easily see how difficult this plan would be to implement on a world-wide basis. To take an example: China, as of right now, seems impossible to reform by the suggested scheme. The authors explain the findings of an experiment where Chinese university students were given free VPN access and were encouraged to use Western news outlets.

“Nonetheless, without the extra encouragements, the vast majority of the students had no interest in visiting foreign websites and did not even want free VPN access. They were so convinced by the propaganda in schools and in the Chinese media that there was no relevant or reliable information about China in Western sources that they did not really need to be actively censored. They had already internalized the censorship.” [5]

Power and Progress emphasizes the need of new technologies to create new jobs instead of only replacing existing ones. As if that can be done intentionally. Technology is not a rational solution-finding process for human problems, but an autonomous force that develops and progresses of its own accord.  Even if new job-creating tech could be consciously implemented at every turn, this entails the need for constant training and retraining of workers. The accelerating speed of technological growth puts an unbearable amount of stress and pressure on workers who can (already) only laugh about the idea of working only one job for the rest of their life.[6] Today, many people can’t keep up with the accelerating change and are cast aside when they're no longer productive like disposable slaves.[7] The suggested solution implies that even more people will get pushed around at increasing speed. If you are already out of the loop, you are in for a rude awakening and a wild ride. And these are only the predictable consequences of the coming change.

Every major technical advance is also a social experiment. New technologies always have social long-term consequences that cannot be predicted in advance. This is because human societies are complex systems that behave in unpredictable ways, like the financial market which is only a tiny part of our current society. When Newcomen or Watt made their respective tweaks for steam engines, they surely did not intend all the social changes that would result. Even if the current trajectory can be brought in the right direction to make sure everyone profits from the coming AI technologies, the social consequences in the long-term cannot be foreseen. What can be virtually assured is that it will be used for “bad” things eventually as well. Acemoglu and Johnson seem to know that, too.

“In Mexico, spyware was originally acquired as a weapon against drug cartels and deployed in the operation that led to the capture of the head of the Sinaloa cartel, El Chapo. But it was subsequently turned against journalists, lawyers investigating the massacre of forty-three students, and opposition parties, including one of the opposition’s leaders, Andrés Manuel López Obrador, who later became the country’s president.”[8]

And once the genie is out of the bottle, it won’t go back in (within the context of a civilization). Outside of a breakdown in the organizational order that supports the technology (e.g., the civilization),the technology will be used, and people will have to live with the consequences the technology inevitably brings.[9]

“Pegasus spyware, snooping by NSA, and Clearview’s facial-recognition technology illustrate a deeper problem. Once out there, digital tools for extensive data collection will be adopted by many, if not most, governments to suppress opposition and better monitor their citizens. They will strengthen nondemocratic regimes and enable them to withstand opposition much more effectively. They could even create a slippery slope for democratic governments to become more authoritarian over time.”[10]

Eventually a historical trend[11] might establish itself:

“Once AI technologies strengthen authoritarian impulses, they create a vicious circle. As governments become more authoritarian, their demand for AI to track and control their population increases, and this pushes AI further in the direction of becoming a fully fledged monitoring technology.”[12]

Another completely false claim the authors make is that “we” choose if a technology is worth it. Who is this “we” they speak of? After thousands of years of human history there still is no consensus on ethics or morality on basic matters, let alone the subtle and complex aspects of advanced technology. Furthermore, the bad effects of any given technology are often not discovered until it’s too late, and the technology has integrated itself into the society such that it is impossible to remove it. Here is just one example of how a “banned” item of technology resurfaces: In the 1960s, pharma company Boehringer Ingelheim developed Methylenedioxypyrovalerone (MDPV), a drug for treating chronic fatigue. Its addictive properties were quickly noticed and development halted. The drug resurfaced as a street drug in the 2000s under the name “flex” and is now seen among heroin, cocaine, and meth. This writer sees no practical and feasible way a world-wide, lasting consensus can be arranged about the use of technologies that offer some kind of benefit. Our society can’t even manage to eliminate all our nuclear weapons—things which offer no benefit, only death and destruction. Virtually everyone agrees that the world would be a far better place without them…yet they proliferate. 

What's the end goal Acemoglu and Johnson push for? An obvious first answer would be the prevention of the worsening of the two-tiered society to ensure material prosperity for everyone. First, material “prosperity” does not appear to strongly correlate with overall happiness. On the contrary, the kind of material “prosperity” that comes with advanced technology appears to bring with it widespread psychological suffering. The average person gets to receive more material goods and opportunities for escapism, but apparently to little end: modern life has proven itself empty and meaningless and no reform to fix this seems to be on the horizon, or even possible. It will be necessary to push around the average worker at accelerating speed. Humans (while they’re still needed) will inevitably be forced to continually adapt to the changing work environment brought about by rapid technological change. Not only that, they will also be forced to behave in accord with the system’s needs in all aspects of their lives. Even if solutions to these particular social problems were possible, many consequences and problems fundamental to technological progress are still inevitable. The ongoing existence of the industrial system threatens the entire Earth through the climate problem. Then there is the plastic problem, the mass species extinction, the overcrowding, the AI threat, and so on.  Future technological development knows no final stop, so no matter how efficiently and “sustainably” the system operates, it will continue gobbling up more and more resources and producing more and more negative environmental effects. Genetic engineering is on the brink of destroying what was formerly known as being a human and will inevitably do so. Ongoing technological development will keep politicians and citizens busy trying to catch up to the changing situation to mitigate negative consequences of the newest items of technical progress. The authors offer nothing to the average man. Power and Progress is the defense of the status quo, for taking the edge off, to keep the populace calm and prevent any meaningful change.  

___________

NOTES:

[1] Acemoglu, Daron; Johnson, Simon, Power and progress: our thousand-year struggle over technology and prosperity, New York, PublicAffairs, 2023, Ch. 1 Control over Technology, Fire, This Time

[2] Ch. 4 Cultivating Misery, A Technological Harvest of Sorrow

[3] Ch. 11 Redirecting Technology

[4] See Kaczynski, Theodore John, Anti-Tech Revolution: Why and How, Ch. 1, 2

[5] Ch. 10 Democracy Breaks, A Braver New World

[6] https://time.com/5876604/machines-jobs-coronavirus/

[7] The World Economic Forum, Future of Jobs Report 2023, p. 28 (“... [T]his report estimates a mean structural labour-market churn of 23% for surveyed companies across sectors and countries over the next five years. Labour-market churn refers to the pace of reallocation of workers and jobs.”)

[8] Ch. 10 Democracy Breaks, From Prometheus to Pegasus

[9] Unless, of course, the entire social system that supports the technology (the civilization) breaks down...

[10] Ch. 10 Democracy Breaks, From Prometheus to Pegasus

[11] For an elaboration on historical trends see Kaczynski, Theodore John, Industrial Society and Its Future, para. 99ff

[12] Ch. 10 Democracy Breaks, Surveillance and the Direction of Technology

Copyright 2024 by Emphour All rights reserved. This is published with the permission of the copyright owner.