Post-AGI Economics (Part 1)

Part 1 - Post-AGI Economics

Post-AGI Economics can also be termed as Post-Human Labor economics. In order to talk about AGI Economics, we first must review the history of labor-based Economics in human civilization.

Human labor and our physical output, the energy humans use and physical force that we can apply to do things in our world, has been the biggest driver of productivity.

From our oldest ancestors, hunter-gatherers, who had not yet figured out how to harness the energy of domesticated animals such as horses, elephants, camels and cattle etc., use of human muscle was the only way to get things done.

When our agrarian ancestors started to use domesticated animals such as Horses and Oxen as well as basic machines like the plow, water wheels, levers and carts, the use of human labor, began to decline.

You may argue that a ‘Lever’ and a ‘Cart’ don't actually add energy to the Equation, but you do have to agree that they allow you to multiply your own Force many folds. They were the first forms of a ‘Force multiplier’.

Then, the first two Industrial Revolutions happened. First in 1760 and the second in 1860. These had a monumental impact on almost every aspect of human civilization. We saw steam engines, railroads, steamships, cotton mills, telegraphs, electricity, automobiles, airplanes and so much more. The second Industrial Revolution lasted until the early 20th century.

The Third Industrial Revolution (also known as the Digital Revolution) began in the 1950s. It is characterized by the shift from mechanical and analogue electronic technologies to digital electronics. Some of the key technological developments that led to the Third Industrial Revolution include:

· The invention of the microprocessor in 1971.

· The development of the internet in the 1980’s.

· The rise of cloud computing in the 2000’s.

· The development of artificial intelligence and robotics.

Now, at present, in the 21st century, we are in the midst of the fourth Industrial Revolution – a revolution whose primary focus is ‘Force amplification’ or ‘Force multiplication’ at an extraordinary scale and pace. This is primarily due to maturing Artificial Intelligence Architectures and advanced deep learning algorithms supported by massive storage, mountains of data and mind-boggling compute power.

In essence, the Fourth Industrial Revolution is the trend towards automation, technologies and processes which include cyber-physical systems (CPS), cognitive computing, and artificial intelligence. This phase of industrial change is fast moving towards blurring the lines between the physical, digital, and biological worlds.

Looking back at each Industrial era, humans first started inventing machines that could use thermal energy from various fuel sources such as wood, coal and oil and they were able to output more energy and more work than we could with just human muscles and large animals combined.

The use of machines gave us the advantage to metabolize fuel at a far higher rate. For example, a gallon of gasoline has about 35,000 plus calories of energy. This is about how much an average human eats in two weeks. A large machine can use a gallon of gasoline in a matter of minutes or hours depending on what that machine is doing. The large 8-cylinder automobiles in the 1950’s to 1970’s were called ‘gas guzzlers’ precisely because they consumed enormous amounts of gasoline in just a few minutes. One large tractor or a harvester on a farm does more work than dozens of humans. The point is that the amount of work being done and the amount of energy used was the big game changing factor for the Industrial Revolution.

But as everyone knows, a tractor is not particularly smart. It pulls a plow which is not any different than what a human or an oxen can do. Tractors are just bigger, faster and stronger. So that is where the Industrial Revolution changed labor-based economics. But you still needed an intelligent human driving that tractor.

The past industrial revolutions of course was a Marvel of Mechanical Engineering, Metal Sciences, Electrical Engineering, Chemical Engineering and of course Electronics among others. But it is important to point out that all this still comes down to the application and use of ‘Energy’.

The human population growth over the last two centuries has been tied directly to the amount of energy we had access to. And that is directly correlated to the amount of food we can produce and the amount of water we can transport, purify and drink.

Ultimately, labor is about what humans can do with our muscles, hands, bodies and what we can do with the use of machines. But, we're also still 100% dependent upon the underlying sources of energy.

In the long history of human labor until today, there are three major categories of labor.

The first is manual, unskilled labor. This includes moving heavy things using your hands and body – such as something that you pick-up and move. For instance, one of the jobs in this category is working in a warehouse to restock the shelves. This does not take much brainwork but it takes a lot of muscles to move heavy things from point A and place them at point B, and, then repeat the process over and over again.

The next level above manual unskilled labor is ‘skilled labor’. This is crafts such as carpentry, electrical work, Ironsmith, Plumbing, etc. This also could include hospitality and food services because it's still very Kinetic. It’s still very much oriented towards the use of the human body. But crafts and other skill require training and many of them require certifications. The one thing all ‘skilled labor’ has in common is that it requires the use of tools and materials.

There are jobs such as a doctor or a nurse that are also very kinetic but those require years of training and sometimes many certifications. So while they are still embodied work, they fall under the third category, which is ‘professional labor’ or ‘knowledge work’. Besides doctors and nurses, this includes professions like Architects, Lawyers, Programmers and many things that you do on computers. Typically, professionals work at a desk, in an office or in front of a computer. However, many of these professions such as the medical profession and lawyers, interact with people all day every day, and that requires a tremendous amount of human labor.

In the past two decades or so, we have gradually transitioned from a largely unskilled labor and skilled labor economy to a ‘knowledge-based economy’ or a ‘services-based economy’.

Significant changes in employment over time have taken place due to the Industrial revolutions. Let’s highlight two trends: one is the agricultural share of total employment, which has decreased from over 30 percent to around two percent in the last century. The other is the non-agricultural employment, which has increased from around 20 million jobs to over 140 million jobs in the past century. This just illustrates how industrial automation, mechanized farming, and economies of scale transformed the agricultural sector, leading to a dramatic decline in farming jobs. This example underscores the broader impact of technological advancements on employment in various other sectors as well. Jobs have gone but there are going to be entirely new categories of jobs that come up.

What people are hoping now is that today all the service jobs that we all hate, all the boring jobs, the desk jobs that we hate today, are going to follow this trend the way that Farm jobs have. It would be an interesting future if there are millions and millions of new jobs created by AI, BUT THAT IS A WRONG ASSUMPTION.

Many AI experts are skeptical that there are going to be millions of millions of new jobs created by AI. There are many reasons for this skepticism;

First, Knowledge work or Service Sector jobs are rooted in what's called ‘Cognitive Labor’ or’ Knowledge Work’. As farm jobs were crashing, knowledge work was rising. The majority of work today is knowledge work or cognitive labor. This also includes ‘Emotional Labor’. Emotional labor’ is a form of cognitive labor that is more common in nurturing and care-giving roles.

AI Tools in form of Large Language Models and Generative AI, like ChatGPT and its successors are going to continue to invade the domain of cognitive labor.

We have already seen ChatGPT being used for tutoring, used for getting medical advice, for relationship advice and people have even automated companies using the automatic CEO concept. It is pretty obvious to everyone by now that these Large Language Models are getting more and more powerful each day. Models like ChatGPT4 are capable of cognitive labor and it would be naïve to assume that this will not continue.

So that leaves physical labor, skilled labor, unskilled labor, and certain areas of cognitive labor. But if AI is invading cognitive labor first, what are we going to do? Do we go back to manual labor? Do we all pick up arts and crafts and others semi-skilled and skilled labor? That is not necessarily what most people want to do. If there is a future state and economy where we are not going to get paid by anyone, but an AI is going to come through and basically take our jobs, we certainly won’t go back to shoveling dirt. And in any case, there will be machines that are better than humans at arts and crafts and even shoveling dirt.

For all the above categories; skilled, unskilled, and cognitive labor, machines are better, stronger, faster, cheaper, never get tired or sick and don’t take any days off, and moreover, they are getting better by each passing day.

This raises a very real question - What are humans going to do for a living in the post-AGI world?

We can view the ‘AI versus Humans’ debate from the ‘Demand Side’. What kind of jobs and work will people demand or want for only humans to always perform. For example, people might always prefer to have a human for child care, elderly care, hospitality jobs, personal instructors, therapy such as massage therapy, influencers, sports coaches and instructors, tour guides, and other performance oriented things, such as singers, comedians, theatre and drama, movies, and so many other jobs where there is no substitute for the ‘human element and touch’.

Many people want to see a real human in their interactions, they want to see real human facial expressions, their intonation, and they want to develop a parasocial relationship with the people that they interact with or get influenced from. Many people do not like an AI face or an AI voice.

The desire for that parasocial relationship and human connection is something that most humans are going to want, no matter how good AGI gets. That does not mean that there will be people who don't really care and just want the information and entertainment. That’s fine, but as human beings, the human element and personal touch will always prevail – at least in the first one or two decades of AGI emergence or until our DNA starts accepting otherwise.

This leads to one final question. Will there be enough of those human exclusive jobs to go around when AI and then AGI take over? No one knows for sure. We might be able to have some, where, maybe we can spend more time with our families. Maybe we can spend more time with child care and education. But that’s not enough to drive entire economies.

Now let us examine what ‘Post AGI Economics’ could look like and briefly go over the main characteristics.

The first is Automation Dominance. This is the idea that Automation in the form of AGI Robots and other kinds of automation is going to take over jobs, including cognitive labor.

The second aspect of post-AGI Economics is the ‘Decreased Economic Value of Human Labor’. Basically humans could not compete on an economic scale with tractors. Why should we then expect that humans are going to be able to compete at any level on an economic scale with AGI, especially when AI’s power keeps increasing?

The third characteristic of post-AGI Economics is ‘Redistributive Policies’. Redistribution is not necessarily a collective ownership. It may be in form of Universal Basic Income (UBI) or a negative income tax.

Number four is ‘Ownership Shifts’. This is the closest thing to Socialism or Marxism and the centralized management of the economy. Centralized management of the economy is really not a good idea because there is more wisdom in decentralized market forces as the past has proven. But what you can anticipate is that the ‘Ownership Class’ might shift and we might shift more towards decentralized ownership models or Collective Ownership models such as co-ops. This is not something new. These models exist today. But with the help of AGI, it might be easier and frictionless for ordinary people to come together and collectively purchase resources, property, and other capital goods that allow them to participate in the means of production in the AGI economy.

This is certainly not going to totally replace all corporations, but if we watch the trends in modern Information Technologies like blockchain, fractional ownership concepts and decentralized autonomous organizations (or DAO’s as they are called), then we can expect super intelligent AGI CEO’s managing our resources for us. Fractional ownership in Real Estate, and, once out of reach masterpieces of collector’s art, is already a popular investment vehicle in blockchain models.

The fifth characteristic of post AGI economics will be Price Deflation. What does this mean? Well, as AI gets better at producing more goods and services, the marginal cost of those goods and services is going to collapse. And it's not all goods and services, some things are going to remain more expensive than others. But other things are going to get very inexpensive.

The sixth is that New Economic Indicators will be needed. This means that GDP, employment and unemployment rates etc. are not going to be quite as meaningful in the future as they are today. Today, total employment and unemployment rates are important indicators for how well the economy is doing. But in a post-AGI economy, employment rates won't matter as much. So we might need other metrics in order to measure how well the economy is doing. One example is Gross National Happiness Index or other metrics that reflect overall well-being and satisfaction.

Next, and number seven, is that there will probably be Changes in Consumption Patterns. For instance, we're going to have more leisure time and less requirement to work. That means that we'll probably change where we live, how we eat and access to some services like healthcare will probably decline because we'll be living healthier lives.

One of the toughest and most challenging is actually going to be number eight. This is about ‘Reshaping our Identity and Self-Worth’. Education would be focusing on Reshaping Identity and Self-Worth and no longer about getting degrees for careers and training workers for the job market. Education would be around nurturing creativity, improving self-worth, empathy, and emotional intelligence etc.

Finally, number nine is Regulatory Changes. These will have to be addressed as we transition from a human labor economy to a non-human labor post-AGI economy.

This concludes Part 1 of the series.

In part 2 of this series, we discuss The ‘Ownership of Production’ in the Post-AGI economy and other ideas like; Universal Basic Income (or UBI), The debate over Inflation and Deflation, and, the concept of Robust Redistribution in an era of 'Radical Abundance'.

Last updated