Search for a report, a publication, an expert...
Institut Montaigne features a platform of Expressions dedicated to debate and current affairs. The platform provides a space for decryption and dialogue to encourage discussion and the emergence of new voices.

Power and Progress

Power and Progress
 Soli Özel
Senior Fellow - International Relations and Turkey

How can we reconcile social advancements with technological progress? In his review of Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity, by Daron Acemoğlu and Simon Johnson, Soli Özel warns against a “vision trap” that leads us to blindly believe AI narratives. When can the third wave of AI not just save money but also benefit society and boost productivity? A brief recall of History shows that it should be possible. The authors advocate for changing the political agenda and establishing the right institutional framework to ensure that technology supports shared welfare for multitudes rather than working against them. Democracy is at stake.

From the early 19th century novel Dr. Frankenstein to today's debates on Artificial Intelligence (AI) and ChatGPT, humanity has almost always had a tense relationship with technology, which is fueled by the belief that technological evolution cannot be interfered with, because technological development on its own leads first to productivity gains and then to progress. 
This approach did not even raise the possibility that the direction of technological development would in fact be determined by political preferences, power relations and the distribution of power in society. More importantly, it ignored the relationship between technology and general social welfare. The negative consequences of certain choices were not even open for discussion.
Daron Acemoğlu and Simon Johnson, professors at the Massachusetts Institute of Technology (MIT), one of the world's leading universities, challenge this mechanistic approach to technological development in their co-authored book Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. They also caution the reader against a danger they call the "vision trap". According to them, "once a vision becomes dominant, its shackles are difficult to throw off because people tend to believe its teachings". This ever-existing danger today results in a failure to question the narrative of the technology giants and to consider alternative paths for technological development and its applications.  

These are the reasons why the authors argue that it is neither possible to understand "Progress", which includes technological development, nor the issue of distribution, without understanding this dimension of power. Their thesis, which they support with historical examples, is that the contribution of technology to general social welfare was only made possible because of social struggles. 

Acemoğlu and Johnson demonstrate that the contribution of technology to general social welfare was only made possible because of social struggles.

The relationship between technology and general and shared welfare is not spontaneous. Considering these, the authors object to the isolation of the debates over AI from their social context. They argue that monopolistic technology companies should be prevented from singlehandedly determining the direction of this technology. Such is also the outlook that informs legislative debates in the EU and even the United States. 

Technology, progress, and distribution from past to present

Acemoğlu and Johnson analyze some of today's important issues that gravitate around the issue of technology and its use. These include the future of globalization, income inequalities, the vast surveillance and controlling powers that AI provides to states, the ability of big tech companies to abuse the power they hold by collecting free data and enriching a limited number of individuals. This use of technology has the effect of weakening democracies, as income inequalities and employment opportunities shrink due to technological developments. 

Arguing against the adherents of the "productivity bandwagon" thinking that link technology automatically to productivity and to development, the authors present a comprehensive technological and economic history of the past 1000 years. For them the key issue is to understand whether technology contributes to general and shared social welfare. They conclude that great leaps in technology in any given period do not necessarily contribute to an increase in general welfare.

In Europe, for example, the standard of living of the peasantry remained almost constant until the 17th century, despite the technological and productivity-enhancing advances of the Middle Ages. Even during the industrial revolution, the most decisive factor of the transition to the modern era, the middle classes that led this revolution did not initially share the gains of increased productivity with the working classes. Such a development only became possible in the second half of the 19th century, when countervailing forces took shape in society.

The authors present a comprehensive technological and economic history of the past 1000 years: great leaps in technology do not necessarily increase general welfare.

Although this dimension is underemphasized, the international conjuncture too plays a significant role in the increase in shared prosperity. The presence of the Soviet Union and the context of the Cold War were factors in the period of about 30 years after the Second World War, when the distribution of income was the most equitable in world history. 

Indeed, the period between 1945 and 1975, which the French refer to as "les Trente Glorieuses", had led to significant productivity gains, as well as an increase in employment opportunities. At the same time, the working class was able to increase its share up to 70 percent of the value produced. The existence of an alternative economic system and the balances established thanks to the organization of working people in developed countries have led to such a result in terms of distribution. According to the authors, two developments, the foundations of which were laid in the early 20th century, defined the post-war structure: the first was "a direction of travel for new technology, that generated not just cost savings through automation but also plenty of new tasks, products and opportunities; and second an institutional structure that bolstered countervailing powers from workers and government regulation.

The paradigm of this period of "shared prosperity", lost its appeal in the 1970s after the quadrupling of oil prices due to the 1973 Arab-Israeli war. The alternative model was not favorable to the egalitarian distribution approach at all. 

New vision, Artificial Intelligence and its consequences

By the late 1970s, the thesis that the sole responsibility of corporations was to maximize profits and protect shareholders' wealth became widely accepted. In this context, since maximizing profits required lowering labor costs, digital technologies evolved in a way that facilitated automation and the elimination of unskilled labor. The different development paths that had begun to germinate in the early 1970s by the original band of techno-hackers, were abandoned. For Acemoğlu and Johnson, it is not automation itself that should be considered problematic, but the technology portfolio that emphasizes automation and does not create new lines of work for workers. This choice combined with globalization and China’s ascent contributed to income inequality in the following years, as well as to social disintegration due to the flight of some industries abroad. 

By the late 70s, digital technologies evolved in a way that facilitated automation and the elimination of unskilled labor to maximize profits.

The authors argue that the social problems that this development system has generated gave way to a deep crisis, especially in the US, and that this is where one ought to look for the roots of Trumpism. Western European countries went through a similar process, albeit in a less destructive way. Later, when technological visionaries found a toy called "Artificial Intelligence", things took a turn for the worse economically, socially and politically.   

The authors take a very negative view of what they describe as "the mother of all inconvenient technologies". Their criticism is harsh: "(AI) is a vision rooted in ideas but it receives a further boost because it enriches and empowers elites corralling technology toward automation and surveillance...". 

This technology is being used more and more widely because it can perform routine human functions, and thereby it facilitates the suppression of labor. It does not increase productivity and is not being used to create new jobs in the same way as in previous eras. According to the authors, this socially problematic situation hides a fallacy, since "only a small fraction of human tasks is truly routine. […] We deal with new situations or challenges by coming up with solutions that draw analogies on the basis of past experience and knowledge. We employ flexibility when the relevant environment changes constantly". 

AI facilitates the suppression of labor, does not increase productivity and is not being used to create new jobs in the same way as in previous eras. 

AI is not able to do this. However, the AI myth promotes the adoption of a rigid automation approach, disregarding or minimizing the flexibility and creativity that humans can exhibit, especially in service industries and even in industrial production lines. Alberto Romero, who has worked in this field for many years, is quoted in the book by the authors as saying: "The marketing power of AI is such that many companies use it without knowing why. Everyone wanted to get on the AI bandwagon".

In addition to a blind love of automation, AI also helps establish a regime of constant surveillance. This trend, which initially started in the field of production of blue-collar workers, is gradually spreading to white-collar workers. 

Human-Complementary Technologies

Arguing that this path was not pre-destined, the authors write that "The current approach that dominates the third wave of AI, based on massive data harvesting and ceaseless automation is a choice. It is in fact a costly choice, not just because it is following the bias of elites toward automation and surveillance and damaging the economic livelihood of workers. It is also diverting energy and research away from other, socially more beneficial directions for general-purpose digital technologies."

Machines and algorithms can and should improve the productivity of workers in their current jobs, and create new jobs for workers.

According to their approach, which they call "Machine Utility", machines and algorithms can and should first and foremost improve the productivity of workers in their current jobs. [...] Second, technologies need to create new jobs for workers. In sectors such as education and health, the application of new technologies can enable the system to be designed differently. Undifferentiated education, for masses with unequal abilities and aptitudes, can now be designed for small groups, or even for individual students, thereby enabling everyone to improve themselves according to their capacities.

Thirdly, new technologies and the machines that use them can be designed to be more useful in making accurate information more accessible, to improve decision-making. Fourth, if new technologies can be used to facilitate the coming together of people with different skills and know-how, they will also increase economic dynamism. However, recognizing that companies do not want to spend money on such a change of course, the authors close on a pessimistic note: "human-complementary machines are not very attractive for organizations when they are intent on cost cutting". 

To date, the use of digital technologies and AI has generally driven out workers or pressured down wages in low-education, low-skill, blue-collar sectors. One important reason for the recent flurry about AI is the fear that white-collar jobs and industries for the highly educated will be next. This is the first time that the jobs of the educated - lawyers, journalists, teachers/academics and the like - who have so far benefited from the applications of digital technologies are under such a serious threat. Their greater ability to speak out, and their capacity to organize a social opposition, may prevent technology from moving in the direction of continued and greater automation.

But Acemoğlu and Johnson are not very optimistic about this. They believe that the currently dominant narrative reflects a formula that suits the interests of big tech companies although "AI based automation fails to increase productivity by that much." In such a case, "if everyone is convinced that AI technologies are necessary, companies will invest in AI even if there are alternative and more useful ways to organize production". If the process unfolds in this way, the relative impoverishment of the middle classes, the pillar of democracies, will accelerate. And historically such times are also times of rising authoritarianism. 

If the process unfolds in this way, the impoverishment of the middle classes will accelerate. Historically, such times are times of rising authoritarianism.

How can democracy be saved? 

One of the most pessimistic chapters of the book is titled "Democracy Breaks". After recalling that the internet and later social media were once the tools that strengthened democratic movements and facilitated societies' resistance to corrupt regimes, the authors summarize the current situation as follows: "[...] Once [digital]  tools started being used primarily for massive data collection and processing, they became potent tools in the hands of both governments and companies interested in surveillance and manipulation. As people became more disempowered, top-down control intensified in both autocratic and democratic countries, and new business models based on monetizing and maximizing user engagement and outrage flourished."

"Once [digital] tools started being used [...] for data collection and processing, they became potent tools [for] governments [...] interested in surveillance and manipulation."

The People's Republic of China has taken the lead in using these technologies to control its own society in the most extreme way possible. Iran and Russia have not lagged China in this regard, especially in terms of censorship. However, the authors note that "abuse of digital tools directed against opposition groups is not confined to dictatorships."Programs such as Pegasus, developed by the Israeli company NSO, can also be used by democratic countries. As exposed by Edward Snowden's revelations to the world, the US National Security Agency, together with giant companies in the digital world, monitored citizens' communications and tapped their phones. It has done the same to the leaders of various countries, including those of allies. 

Digital technology is neither democratic nor anti-democratic. Decisions and choices about how to use this technology define its function. The direction in which tech companies have devoted all their power to the collection of personal data and the revenue they generate from it has fostered authoritarianism. "(E)arly hopes associated with digital democratization have been dashed because the tech world put its effort where the power and money lie - with government censorship."

In addition to facilitating authoritarianism and state surveillance of citizens, the monopolistic power of social media giants have accelerated the erosion of trust in democratic societies. The irony of this situation is that the owners and managers of these companies see themselves on the left of the political spectrum. But what they do is violate people's privacy, exploit their data for free, and easily cooperate with the most repressive regimes when their commercial interests require it. However, looking at the Wikipedia experience, which is also used as an example in the book, it is possible that the available technology and data production could be utilized in much different ways that encourage participation. 

Digital technology is neither democratic nor anti-democratic. Decisions and choices about how to use it define it.

After all this criticism, the authors conclude with some suggestions that they believe will slow or even reverse this trend in the coming period. Their optimism about the possibility of a new path and a new vision for AI technology is inspired by the valiant political struggles and successes of the early 20th Century Progressive movement when their efforts in a place like the United States were deemed almost delusional. The authors quote attorney (later Justice) Louis Brandeis approvingly at the beginning of their final chapter, Redirecting Technology: "Most of the things worth doing in the world had been declared impossible before they were done."

For the authors, what is needed is "the right institutional framework and incentives shaped by government policies, bolstered by a constructive narrative".

Some of their ideas may seem utopian in today's conditions, while others are well grounded and could be the starting points of a new agenda. They summarize their approach succinctly as follows: what is needed is "the right institutional framework and incentives shaped by government policies, bolstered by a constructive narrative, to induce the private sector to move away from excessive automation and surveillance and toward more worker-friendly technologies".

A democratic political environment and a space for debate that is not poisoned by self-serving high-tech giants and the social media under their control are the prerequisites for moving in such a direction. 


Copyright Image: WANG Zhao / AFP

Receive Institut Montaigne’s monthly newsletter in English