Monday, January 16, 2023

Book Review: The Autonomous Revolution by Davidow & Malone

More than one year back, Madhu, another longtime friend of mine in my "book exchange program" gave me "The Autonomous Revolution - Reclaiming the Future We've Sold to the Machines" by W.H.Davidow & M.S.Malone, saying it should keep me entertained on my long return flight. I got around to reading it only now, and realized his prediction was correct for couple of reasons. Content & narration wise, it is an easy read and so potentially one could finish reading it in few hours. Since it is less than 200 pages long, it is not a big tome either. Second reason is a bit more personal. Since I have been giving this lecture titled "Ethics & Emerging Technologies" for a while now, I had thought/read about most of what the book had to discuss and so the material was quite familiar to me. I was looking for new insights, ideas, concepts that I could use in my future lectures. I did find some but not that many. 


First chapter titled "Autonomous Revolution - A third social revolution" discusses how what is going on now, constitutes the third revolution after agricultural and industrial revolutions of the past. After using the word revolution twice in the first chapter title, second chapter, a bit confusingly, talks about transportation revolution (when cars became common in the 20th century), telecom revolution and says some are calling the current autonomous revolution as the "Fourth Industrial Revolution" and tries to explain why it is not appropriate. While the authors points of how "this time it is different" is well taken, the overuse of the term revolution by themselves and then saying why this is not fourth industrial revolution didn't come across as well articulated clear cogent writing to me. 

The substitutional equivalence that leads to societal phase change is a good insight presented in the following chapter. What this means is when a new method introduced to do the same old task more efficiently, it provides a substitutional equivalence. This is the case when spreadsheets replace handwritten ledgers in a bank, since the work performed is still the same. But when the new technology enables the industry to do its business in a totally different form, it leads to societal phase changes. Thus, when banking has been moved online, eliminating the need for buildings, tellers and so on, that is a phase/societal transformation. While the first equivalent change leads to productivity gains but may not affect the number of people working in the industry much, phase changes eliminate large numbers of jobs permanently. Travel agents, agricultural labor are all examples of this kind. 

Book also discusses non-monetizable productivity increase that adversely affects certain industries, including ours (i.e., semi-conductors). Authors give the example of Intel specifically. As everyone knows, our industry is legendary when it comes to efficiency improvements in cramming more transistors per square millimeter and reducing the power consumption, leading to orders of magnitude better microchips. But since the competition is fierce worldwide (e.g., there is no license or regulatory protection that prevents one country/company from selling its products anywhere in the world), price keeps coming down, preventing us from monetizing this phenomenal productivity increase. But the efficiency improvement is still felt/enjoyed by the whole world in the form of cheaper electronics over the past few decades. Luckily since the volume of sales increased thanks to improved affordability, the pie size became bigger. So, we managed to make a good living. But in cases where the volume of sales cannot increase that much due to market saturation, industry might be doomed. For example, if AI improves newspaper article writing efficiency, the reduced cost is not going to keep increasing newspaper sales, helping newspaper industry to thrive. If self-driving cars reduce the costs involved in providing transportation, the usage may increase a little bit. But most of the car sales will go away since there will be no need to own/maintain your own transportation anymore. Since these changes are coming, society has to rethink how the fruits of these efficiency improvements are going to be shared across the population more equitably. 

Authors do touch upon "algorithmic prisons" we are starting to live in (since social media algorithms limit what you get to consume), loss of privacy and other issues I keep talking about in my lectures and posts. Everyone know they have become big topics of discussion though no one seems to be seriously trying to get out of these constraints, as we are used to the freemium models and don't know how to get out of them. It is not easy to come up with a neat silver bullet solution and so authors talk about well-known solutions like basic income guarantees, end users owning their data and being able to decide who gets to see how much (and possibly getting paid for it), etc. that I also often talk about. For those who may not be thinking about these issues regularly, book may be a good, easy read. 

No comments:

Post a Comment