De cerca, nadie es normal

On Social Justice and AI Automation & Augmentation

Posted: January 29th, 2023 | Author: | Filed under: Artificial Intelligence | Tags: , , , , | Comments Off on On Social Justice and AI Automation & Augmentation

The year has gotten off to a bad start for many families.

On January 5th, Amazon announced that it would lay off 18,000 employees. Days later Google stated it would lay off 12,000 employees; and the last to join the merry-go-round was Microsoft, which announced on January 18th that it would lay off 10,000 people. Twitter kicked things off when, in November last year, it announced the layoff of almost 4,000 employees.

What’s going on in the industry? I am not going to be the one to do an in-depth analysis -which has already been done- of the economic and financial reasons that have led these companies to make these decisions. What is clear is that, sad as it may seem, some positions made little or no sense at all from a business point of view (Chief Happiness Officer!), and the labor market in this sector was totally “overheated” concerning salaries with all the cash volume that both governments and central banks -directly or indirectly- had pumped into the economy.

However, let’s move on to a reflection that has gone somewhat unnoticed these days and which is the one that interests me: has or will the progressive implementation of AI in these companies have anything to do with these layoffs? Before pondering on it and answering…Blue pill or red pill? As always, red pill.

As happened in the first and second industrial revolutions with the steam engine, electricity, the telephone or the radio, we have before us a new and likely the most general of all general-purpose technologies: artificial intelligence. AI is not only an innovation itself, but also one that triggers cascades of complementary innovations, from new products to new production systems.

In both the first and the second industrial revolution, there were initial phases of adaptation that meant job losses for thousands of workers, since their jobs and skills no longer made any sense in the new economic scenario. And this is where we begin to go deeper into the analysis: automatization versus augmentation.

Let’s be positive, at least from the outset: both automation and augmentation can boost labor productivity. Nevertheless what happens with automation is that, as machines become better substitutes for human labor, workers lose economic and political bargaining power and become increasingly dependent on those who control the technology and on their financial business plans. 

How are we envisioning AI nowadays? Towards automation or augmentation? There are many who deem AI should be focused on augmenting humans rather than mimicking them.  Augmentation through AI creates new capabilities and new products and services, ultimately generating far more value than merely automating human tasks. In this approach humans and machines become complements. Complementarity implies that people remain indispensable for value creation and, when humans are indispensable, economic power and political decision-making tend to be more decentralized and democratized. 

Nonetheless, there are currently excess incentives for automation rather than augmentation among technologists, business executives, and policy-makers. When AI replicates and automates existing human capabilities, it tends to reduce the marginal value of workers’ contributions, and more of the gains go to the owners, entrepreneurs, inventors, and architects of the new systems. Entrepreneurs and executives who have access to those AI models can and often will replace humans in those tasks. 

There are some voices which defend a fully automated economy, such as one which could, in principle, be structured to redistribute the benefits from production widely, even to those people who are no longer strictly necessary for value creation. However, the beneficiaries’ incomes would depend on the decisions of those in control of the technology. This opens the door to increased concentration of wealth and power. 

What is the solution regarding this dilemma? Clearly it is not slowing down technology, but from my standpoint rather eliminating the excess incentives for automation over augmentation. Think for instance on the US tax legislation, it encourages capital investment over investment in labor through effective tax rates that are much higher on labor than on plants and equipment. The US tax code treats labor income more harshly than capital income.

As a conclusion, the more technology is used to replace rather than augment labor, the worse the disparity may become. At the same time, automating a whole job is often extremely complex. Every job involves multiple different tasks, including some really challenging to automate. Think on industries such as health, legal, domestic security…

As mentioned once in a workshop, human beings and AI models should be -using the image of the Greek mythology- centaurs: a perfectly coordinated and unbeatable mix of wisdom and power.

Let’s see if, for once, we can think on the general benefit.


Comments are closed.