Talking about AI within a Broader Historical Context of General Purpose Technologies

On July 7, 2023, we had the pleasure of speaking with Simon Johnson, the Ronald A. Kurtz (1954) Professor of Entrepreneurship at the MIT Sloan School of Management, where he is head of the Global Economics and Management group. In 2007–08 he was chief economist at the International Monetary Fund, and he currently co-chairs the CFA Institute Systemic Risk Council. In February 2021, Simon joined the board of directors of Fannie Mae.

The topic: on May 16, 2023, Simon, along with his co-author Daron Acemoglu, published Power and Progress: Our 1000 Year Struggle over Technology and Prosperity. We can attest that it is a fascinating book from two angles:

1.  Many technological developments have occurred before and can provide us with their own historical analogues, even if imperfect. Looking at the rollout of the steam engine, development of electricity or the execution of the Suez Canal are useful. We can see some of the positives and the negatives they brought along with them, as well as how society interacted with them and how long they took to make their impact.

2.  There is also ample attention paid to the present day, where we have technologies like social media and automation and even artificial intelligence. It is worth thinking about the types of decisions that can be made with these technologies that can influence whether they end up benefitting small numbers of people, like ‘tech founders,’ or if they have a maximally beneficial impact across all of society, lifting the economic tide of all boats.

A refrain of both the book and our discussion is that nothing about technology is ever set in stone or pre-determined. Technology need not increase inequality across society, given appropriate decisions being made at the policy level.

Can Technology Hurt Society?

We were curious to look back through history, searching for an example where a new technology was introduced and did not lead to the average worker being or feeling ‘better off.’ Simon’s take was that the cotton gin was possibly the most illustrative example. At that time in history, in the parts of the U.S. where cotton was being grown, the laborers were slaves. The cotton gin increased the productive capacities of cotton farming, with the result that the market soon demanded more slaves. In addition to being a stain on America’s history, from an economic perspective, the picture is clear—more cotton was being produced more efficiently and none of those economic gains were touching the lives of the workers.

AI—the Path Is Not Yet Certain

The cotton gin itself is merely a machine—it doesn’t force slavery on society. Likewise, AI is merely a tool, software that can take in massive amounts of data and find patterns. The primary question then is about its precise use and the societal background in which it is being used. Simon mentioned surveillance as an example of a particularly negative AI use case, which is what China’s government uses in certain provinces. This is not an example of technology use that is expected to improve standards of living or productivity, but rather to create an environment of compliance or even hopelessness about the ability to ever express dissenting views to government.

As a global society, it is on us to decide if the primary use case for AI is merely surveillance technology, or if it can be applied to a much richer tapestry of different activities.

More Competition in Certain Areas of Technology Could Be Beneficial

We had the conversation with Simon during the week when Facebook’s ‘Threads’ platform was released as a competitor to Twitter. The evolution of social media is interesting to consider, in that a narrow set of large platforms has dominated. The classic, often-cited issue is that Facebook purchased Instagram and WhatsApp, two platforms that were large and growing fast in their own right. Simon’s view as that competition is good and he wished there were more companies in the space—also that the Twitter competitor wasn’t coming from Facebook, already a large player in its own right. We’ll see if recent government actions regarding M&A activities, where the FTC and different European authorities have expressed that they too favor encouraging competition, leads to a different look and feel for the tech landscape.

Work Will Continue to Evolve

At the core of the book, as well as our discussion with Simon, was the relationship between human beings and work. Simon noted the evidence suggests that people want to work, and that they are getting something beyond mere economic value in the bank account at the end of a given period. This is an argument against a future state that is driven largely by a measure of ‘Universal Basic Income,’ with many people just taking leisure time for the majority of their lives. We, as a society, do have an issue—growing inequality.

It is true that this issue has been growing since the 1980s, and that labor’s share in certain economic gains has been slow to materialize. Still, we have many options. Germany was cited as an interesting case, where for the introduction of new technologies to different industrial processes, there are many voices at the table setting policy and getting the effort off the ground. Workers being represented there gives them a better chance to share in the gains. Can something like Germany’s approach be used more broadly? Every country is different, but a core takeaway was that workers need to have more of a voice and an influence to share in gains, and that the best new technologies from a productivity perspective don’t merely ‘automate,’ but also bring with them new tasks, creating the need for new skills and even new jobs to go along with them.

If you are interested in this topic, we certainly note the pleasure we got from reading Power and Progress: Our 1000 Year Struggle over Technology and Prosperity.

Listen to the full discussion below: