What if we have been thinking wrong about artificial intelligence?
After all, ai is often talked about as something that could replicate human intelligence and replace human labor. But there is an alternative future: one in which ai provides “machine utility” to human workers, increasing but not usurping jobs, while helping to generate productivity gains and spread prosperity.
It would be a quite flattering scenario. However, as MIT economist Daron Acemoglu emphasized in a public lecture on campus Tuesday night, society has begun to move in a different direction: one in which ai replaces jobs and increases social surveillance, and in the process it reinforces economic inequality while concentrating political power. in the hands of the ultra-rich.
“We have transformative and far-reaching options ahead of us,” warned Acemoglu, a professor at the MIT Institute who has spent years studying the impact of automation on jobs and society.
Great innovations, Acemoglu suggested, are almost always tied to issues of power and social control, especially those involving automation. technology generally helps society increase productivity; the question is how closely or widely those economic benefits are shared. When it comes to ai, he noted, these questions are very important “because there are many different directions in which these technologies can be developed. It is quite possible that they could provide broad-based benefits, or that they could actually enrich and empower a very small elite.”
But when innovations augment rather than replace workers’ tasks, he noted, they create conditions in which prosperity can spread to the workforce itself.
“The goal is not to make machines intelligent per se, but to make them increasingly useful to humans,” Acemoglu said, speaking to a near-capacity audience of nearly 300 people in Wong Auditorium.
The Productivity Wagon
The Starr Forum is a series of public events organized by MIT’s Center for International Studies (CIS) and focused on prominent topics of global interest. Tuesday’s event was presented by Evan Lieberman, director of the CIS and Full Professor of Political Science and Contemporary Africa.
Acemoglu’s talk was based on themes detailed in his book “Power and Progress: Our 1,000-Year Struggle for technology and Prosperity,” which was co-written with Simon Johnson and published in May by PublicAffairs. Johnson is the Ronald A. Kurtz Professor of Entrepreneurship at the MIT Sloan School of Management.
In Tuesday’s talk, as in his book, Acemoglu analyzed some famous historical examples to point out that the widespread benefits of new technology cannot be assumed, but are conditional on how the technology is implemented.
It took at least 100 years after the start of the Industrial Revolution in the 18th century, Acemoglu noted, for the productivity gains of industrialization to be widely shared. At first, real incomes did not increase, working hours increased by 20 percent, and working conditions worsened as textile factory workers lost much of the autonomy they had had as independent weavers.
Similarly, Acemoglu observed, Eli Whitney’s invention of the cotton gin further worsened conditions of slavery in the United States. That general dynamic, in which innovation can potentially enrich the few at the expense of the many, Acemoglu said, has not gone away.
“We’re not saying this time is different,” Acemoglu said. “This time it is very similar to what happened in the past. “There has always been this tension about who controls the technology and whether the benefits of the technology will be shared widely.”
To be sure, he noted, there are many, many ways that society has ultimately benefited from technologies. But it is not something we can take for granted.
“Yes, in fact, today we are immensely more prosperous, healthier and more comfortable than 300 years ago,” Acemoglu said. “But again, there was nothing automatic about it, and the path to that improvement was winding.”
Ultimately, what society should aspire to, Acemoglu said, is what he and Johnson call “The Productivity Bandwagon” in their book. That is the condition in which technological innovation is adapted to help workers, not replace them, spreading economic growth more broadly. In this way, productivity growth is accompanied by shared prosperity.
“The productivity bandwagon is not a force of nature that is automatically applied in all circumstances and with great force, but rather it is something that is conditioned to the nature of the technology and how production is organized and the benefits are shared. “Acemoglu said.
Fundamentally, he added, this “double process” of innovation involves one more thing: a significant amount of worker power, something that has eroded in recent decades in many places, including the United States.
That erosion of worker power, he acknowledged, has made it less likely that multifaceted technologies will be used in ways that help the workforce. Still, Acemoglu noted, there is a healthy tradition within the ranks of technologists, including innovators like Norbert Wiener and Douglas Engelbart, of “making machines more usable or more useful to humans, and ai could follow that path.” ”.
On the contrary, Acemoglu noted, “there is a danger that putting too much emphasis on automation will not generate many productivity gains either,” since some technologies may simply be cheaper than human workers, not more productive.
Icarus and us
The event included a commentary by Fotini Christia, Ford International Professor of Social Sciences and director of the MIT Sociotechnical Systems Research Center. Christia stated that “Power and Progress” was “a tremendous book about the forces of technology and how to channel them for the greater good.” He also noted “how prevalent these themes have been even since ancient times,” referring to Greek myths involving Daedalus, Icarus and Prometheus.
Additionally, Christia raised a number of pressing questions about the topics of Acemoglu’s talk, including whether the advent of ai represented a more troubling set of problems than previous episodes of technological advancement, many of which ultimately helped many people; which people in society have the greatest capacity and responsibility to help bring about change; and whether ai could have a different impact on developing countries in the Global South.
In an extensive audience question-and-answer session, Acemoglu answered more than a dozen questions, many of them about income distribution, global inequality and how workers could organize to have a say in implementing the ai.
Broadly speaking, Acemoglu suggested that how greater power can be gained from workers remains to be determined, noting that workers themselves should help suggest productive uses for ai. At multiple points, he noted that workers cannot simply protest circumstances, but must also seek policy changes, if possible.
“There is a certain degree of optimism in saying that we can really redirect technology and that it is a social choice,” Acemoglu acknowledged.
Acemoglu also suggested that countries in the global South were also vulnerable to the potential effects of ai, in several ways. For one thing, he noted, as the work of MIT economist Martin Beraja shows, China has been exporting ai surveillance technologies to governments in many developing countries. On the other hand, he noted, countries that have made overall economic progress by employing more citizens in low-wage industries could find that labor force participation is undermined by ai developments.
On the other hand, Acemoglu warned, if private companies or central governments anywhere in the world accumulate more and more information about people, this is likely to have negative consequences for the majority of the population.
“As long as that information can be used without restrictions, it will be undemocratic and will generate inequality,” he said. “There is a danger that ai, if it goes down the path of automation, will become a highly unequalizing technology around the world.”