I’m updating my SVA course async material, and I’m adding some ‘AI Effects’ reading gems that are probably gonna concern a few folks.
I’m pretty sure that the survey report isn’t gonna to sit too well with my students… but it’s the reality everyone needs to understand. When I read things like this I still struggle to answer my students who ask, “What’s the good news?”
And… here’s another Brookings Institute paper from 2022, “Preparing for the (non-existent?) future of work” that discusses the… <cough>… “non-existent” future of work.
https://www.brookings.edu/articles/preparing-for-the-non-existent-future-of-work/
This paper outlines 3 “concerns” about the future of work where AI and automation can substitute for human labor: “Labor saving progress”, “Perfect substitutability of Labor”, and “Economic Redundancy of Labor”. All three of these map to the Replacement Effects as GenAI takes over tasks previously performed by humans.
However, there’s some interesting 2nd order nuance to be found in here. “Labor-saving” technologies will augment human work, helping workers to become more productive… at least in the short term. But remember, if your tasks are augmented today, they will be replaced tomorrow.
I also see some hints of Transformation Effects in the “Perfect substitutability of Labor” and “Economic Redundancy of Labor”. If history can shed any possible light on where we are… I’m optimistically leaving the door open for new roles and industries that we cannot even imagine yet. However, I’m still thinking we’re gonna need some sort of Universal Basic Income to prevent chaos in the streets while society transforms into a post-labor economy… but that’s for another post.
Take a look at the left side of this Figure 2 from the paper1:
This shows an individual worker’s “Non-labor income” (T) with respect to their “Labor Productivity” (w). This helps us visualize and understand the transition people are facing due to the advancements in Generative AI. As the marginal cost of cognition and creativity approaches zero, people will experience a decline in the economic value of their labor. In other words, they can no longer exchange their time, labor, or knowledge, for a similar value of money. If these people have significant non-labor income, this will push them from the "Work" region to the "Do Not Work" zone. However, if they are below the C0 threshold (the “subsistence consumption” line, a minimum level of income necessary for an individual's basic survival needs), this will push them into the Perish zone.
The “Do not work” zone indicates a positive outcome from GenAI for the “Capital Class” of society. Remember, the non-labor income in on the X-axis. The population of people who benefit from maximizing shareholder value… the folks who have ample non-labor income and assets will not have to worry. However, if you have a job which you need to maintain to support yourself or your family, then you need to pay very close attention to the changes that are happening. My advice, do everything you can to get some non-labor income. The more the better.
As the labor productivity value of GenAI increases, the relative labor productivity of humans decreases. AI will produce the same or better outputs and at a lower cost to humans, this introduces “economic redundancy effects” in the market.
And what happens to workers when you’re seen as redundant? More often than not, you get fired. Unless, the leaders of your organization are smart, and understand that they can now start to address their “wish lists”… aka the Tushman “Performance Gaps” and “Opportunity Gaps”.
The stark truth is that AI's march forward is gonna undeniably displace many traditional roles. Tasks that were once the domain of human ingenuity and effort are increasingly being performed by machines with unparalleled efficiency and at a fraction of the cost. This is all about the effects of technology on overall labor demand. There’s a possible future where “the marginal costs of cognition and creativity dropping to $0”2 parallels the “marginal product of labor being less than the cost of human subsistence”3.
For those whose labor is augmented today, the unsettling prospect is that it will likely be replaced tomorrow. This unsettling transition, is an unavoidable consequence of technological advancement.
History shows that new stuff is rapidly coming… but the challenge is preparing for the unknown. I teach my students that adaptability, continuous learning, and an entrepreneurial mindset are key.
And if you wanna learn more… please signup for my course at SVA! Class starts June 4th! Everything’s on Zoom. I teach using the HBS Case Method, and it’s the best way to learn!
Korinek, Anton, and Joseph E. Stiglitz. "Artificial Intelligence and Its Implications for Income Distribution and Unemployment." Brookings, August 2022. https://www.brookings.edu/wp-content/uploads/2022/08/2022.08.10_Korinek-Juelfs-Final.pdf. Figure 2.
The “marginal cost of cognition and creativity is dropping to $0” is a quote from Karim Lakhani @ HBS.
The “…marginal product of labor would be less than the cost of human subsistence”. Korinek, Anton, and Joseph E. Stiglitz. "Artificial Intelligence and Its Implications for Income Distribution and Unemployment." Brookings, August 2022. P10. https://www.brookings.edu/wp-content/uploads/2022/08/2022.08.10_Korinek-Juelfs-Final.pdf.