This interview with Azhar has been edited and condensed.
You claim we’re living in a “wholly new era,” the Exponential Age. How can technological improvement be exponential?
We can see exponential change in the mathematics of compound interest. If your bank provides the same amount of interest each year, your capital will grow more quickly over time.
An exponential technology gets 10 percent better every year for roughly the same price. This means in seven years, the technology will double its performance. After 14 years, two doublings will occur, making the technology four times better. The rate of improvement for an exponential technology continues for an extended period, normally decades.
What are examples of technologies that are improving exponentially?
Many of the key technologies needed for renewable power are on exponential price declines. Solar power and wind turbines have had 23 to 25 percent price improvements every year for decades now. Lithium-ion batteries, which are important for storing energy, have been on an exponential price decline on the order of 17 to 19 percent per annum for a couple of decades.
Computer chips improve on a price-performance basis by roughly 50 to 60 percent every year. That has been compounding like the gift that seemingly never stops giving since the 1950s. When I was born in the early 1970s, there were fewer than a million computers in the world. Today, there are more than five billion, including smartphones that are more powerful than any computer that existed back then. As computers become cheaper, more services use them, and their numbers grow rapidly.
You and I can probably remember a time when very few people were on a social network. And then, within seven years of Facebook launching, everybody’s on one. Then along comes TikTok, and it grows faster than Facebook, going from nothing to suddenly being everywhere in a matter of seemingly months. And it seems that the same dynamic we see in the underlying technologies — compounding acceleration — appears in the products built with these technologies and the businesses that run these products.
What’s driving these huge reductions in the costs of technologies?
The critical thing seems to be how well we learn how to make them. Many things can affect the rate of learning. For example, in a global economy that is good at sharing ideas, someone prototypes a new app or shows off a product and it spreads quickly across the Internet. This allows for collaboration and outcomes like rapidly emerging copycats and clones. Through global supply chains, these products quickly get into the hands of many customers who make new demands. Apple launched its first smartphone in 2009 in one store in San Francisco. But when it launched the most recent one, it was available in dozens of countries. Now that giddy rate of change isn’t contained in a single city.
You say an exponential gap has arisen because technology changes faster than societal norms. But haven’t societies long faced technological challenges?
We’ve often contended with the pace of technology going faster than institutions can respond. Certainly, at the point of industrialization, Victorian England had real issues where business applications of technology ran ahead of societal needs. As I note in my book, Charles Dickens describes the poor living conditions of industrial towns in the north of England, where soot and ash permeate the air outside the factory and workers get paid next to nothing and yet work constantly.
What I think is distinct about the exponential gap is the tempo of these changes. Let’s go back to TikTok. It went from not being on anyone’s agenda to being a national security threat [because the Chinese app collects data on users] within a few weeks.
The question, therefore, isn’t “Can’t we just figure things out as we’ve done before?” but “Can we quickly mitigate harms and provide the right benefits?” There’s an argument for moving policy more rapidly and concertedly during moments of great change. William Ogburn was a US sociologist who studied employer liability laws in New York state between the 1890s and the 1930s. He observed that decades of foot-dragging resulted in thousands of casualties that could have been avoided by passing laws earlier. The same was true in Victorian England. By the time laws were passed to protect worker health and safety, decades had passed when large numbers of people were miserable.
How should we think about power when the Exponential Age is creating superstar companies and a winner-take-all economy?
Suppose companies become so big and powerful that they start looking like infrastructure, like roads and water systems. In that case, we should begin by saying, “You’re an essential part of society providing the oxygen we need — not an innovator taking enormous amounts of risk.” I think we should ask these companies to make their services more competitively available and adhere to higher levels of obligation in terms of who gets access to them and on what basis.
For example, Facebook’s hold on social networking is reminiscent of the type of control AT&T had over the phone networks until the mid-1980s. Go to another service and you’ll lose access to your connections, messages, and friends. Interoperability — allowing a user on one network to access another — is a common principle for opening markets and giving consumers more choice. It reduces the strength of network effects and introduces competition that keeps service providers honest.
Evan Selinger is a professor of philosophy at the Rochester Institute of Technology and an affiliate scholar at Northeastern University’s Center for Law, Innovation, and Creativity. Follow him on Twitter @evanselinger.