Will AI take our jobs? And how will society cope in the aftermath?
Rapid advances in AI performing multiple tasks and being indistinguishable from people while discoursing - begs the question of how our society will look if and when AI becomes ubiquitous?
As a guide, let's look at what the impacts of technology in our society have been.
Historical Trends of Technology in Society
According to a McKinsey article, as technology grew, jobs have also grown, starting from 1960. The following example illustrates the economics of how this happens:
As the assembly line was introduced, the price of Ford Model T dropped more than half from $950 → $440 per car. This spurred demand for these cars, and so rather than declining, employment in the auto sector soared.
However, there has been a lack of corresponding rise in worker wages, along with the boost in productivity. The gains from automation are not all going to the workers - since 1970, wages have largely remained stagnant even with dramatic booms in productivity. As productivity goes up with the substitution of human labor with machines, companies that install the relevant machines and capital, end up raking in the profits.
In most sectors, it seems as though employment has increased along with increased technology. Agriculture offers a stark contrast. Why is the case?
What happens when the market gets saturated?
One reason could be applying the analogy of the Ford assembly line to agriculture. As technology reduced costs for making vehicles and skyrocketed demand, over the next few decades, employment had to significantly grow to keep pace to the increased demand.
As you can see, in the early 1900s, less than 1 in 10 people in the US had a car. However, after 2000, this number has jumped to 8 out of 10 people, and it is highly unlikely for there to be a sudden jump in US car demand anytime soon (unless for some reason each person now wants 2 cars!)
Since the past 100 years, there has been hardly any change in plant based agricultural planted area, and also with meat production. In the agricultural sector, it seems as though existing production areas got more efficient with the introduction of machines, and there was no boom in agricultural demands resulting in reduced human labor.
In the book Human Compatible, Stuart Russel argues that the change in employment following an industry's adoption of AI would follow an inverted U-shaped curve. Russel looks at a hypothetical example of automation in the painting sector. If the paint brush is very small measuring mm in width, this is an inhuman task and no one paints their house. With a few mm wide, some dedicated murials are painted. At 10 cm, the price goes down and demand goes up. Now coming to slightly larger paint rollers and spray guns, demand saturates, but efficiency increase, so the number of painters drops. Finally, when you have multiple robots equivalent to meter sized brushes, whole houses can be painted and very few painters are required.
Rise of Industrial AI
Currently, automation is limited to highly specialized tasks. Recent developments in AI like ChatGPT have shown us that in many cases, machines can respond just like humans and thus we can have more meaningful interactions with them.
As I've illustrated in another article, think about the potential for having a device (let's call it DeviceX) that can do multiple intelligent tasks based on few inputs. You could say “DeviceX plan me an entire week's family vacation in a European country” could be as good as you or your spouse planning an entire family vacation. It would take into account your monetary status, kids schedules, food preferences, time preferences, etc. and do the necessary bookings.
In principle, entire industries with complex moving parts can be replaced by semi-intelligent “minds” that spearhead everything from design, to process, product, and customer outreach.
How will society cope?
In his book Human Compatible, Russel says that an important class lesson that all of humanity would need to take in the near future would be in “the art of life itself.” This is because soon, when people are replaced by productive AI based machines that don't have nearly as many demands as we do, we need to figure out how to keep occupied, and not waste away as a species. He quotes John Maynard Keynes the famous economist:
It will be those peoples, who can keep alive, and cultivate into a fuller perfection, the art of life itself and do not sell themselves for the means of life, who will be able to enjoy the abundance when it comes.
Iain Banks' beautiful science fictional Culture Series illustrates the life of a super intelligent society (The Culture) that has solved all existential problems. Characters live in a utopia where they are genetically altered to increase lifespan, naturally release drugs, and live lives of pleasure. In The Culture, humanoids coexist with drones and super intelligent “minds” peacefully. Even though minds are much more intelligent than the humanoids - their existence centers around their purpose in the society, and often these minds are in charge or running spaceships that can travel space at the speed of light.
Final thoughts
Since the 1960s, as technology has been introduced - there has been a growth in both productivity and employment in 79% of the years. Nowhere is this better illustrated than in the automobile industry, where automation reduced costs, increased demand, and hence created more jobs. Only in 12% of the cases have we seen a growth in productivity and a decline in employment.
While this may point towards a trend of increasing employment with the adoption of industrial AI, over the long term the trend could be the opposite. As demand saturates and AI powered automation gains new grounds, the number of jobs could drastically reduce.
If the economics works out, cheap, scalable, and efficient AI could work harder and better than humans at a fraction of the cost. In this case, there might be a reduced need for human labor, and reaping the benefits of AI might mean setting up universal basic incomes (UBI) for much of society. An important lesson we would need to learn then is:
How does one lead a fulfilling life without necessarily contributing to society's basic functioning and leaving that to the machines? How will we make progress without frittering our lives purely on momentary pleasures?
I expect a gradual mindset change - changing from a 40-hour work week to maybe 20 hours. Some jobs might be more redundant than others - to give us a sense of purpose. And more UBI recipients for those who are finding it hard to land a job.
If you enjoyed this post, please share on social media or even just one person you think might enjoy holistic perspectives on the interconnections between technology and modern societies. Feel free to also post any comments in the post discussions on the cyber-physical substack page. This is a small, but growing effort and I hope that I can share in my journey in understanding and building resilient societies.