Post-AGI Economics (Part 2)
Part 2 - Post-AGI Economics
This is part 2 of the article titled; Post-AGI Economics; What will happen to the Job market and the economy when Artificial General Intelligence (AGI) arrives and starts dominating the workplace.
In the first part, we discussed the traditional human-labor categories and energy utilization in the past centuries, including the three Industrial Revolutions. We also gave an overview of various characteristics of the Post-AGI economy, in which, it is anticipated that human-labor will have a very negligible role.
In part 2, we will discuss The ‘Post-AGI ‘Ownership of Production’ 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'.
So let’s get started;
First let’s define what ‘Radical Abundance’ means. It is the idea that AGI and advanced nanotechnology will create a world where everyone has access to the resources they need to live a fulfilling life. This includes things like food, water, housing and shelter, clothing, education, and healthcare. AGI has the potential to solve some of the world's most pressing problems, such as climate change and poverty.
Overall, the concept of radical abundance in the era of AGI is a hopeful one. It suggests that we have the potential to create a world where everyone can thrive. However, for humans to realize this promise, we need to make sure that AGI is used in a fair and ethical way, and we can ensure equality so that everyone can benefit from the fruits of technological progress.
Now let’s take-up something that is at the center of many debates – ‘The Ownership of Production’.
The basic question is; Who owns the machines and technology that own the means of production in the AGI economy?
That is a mouthful, but the question is essentially asking who has ownership and control over these AGI-driven means of production. It highlights concerns about the distribution of power and wealth in such an economy.
If a small group or entity owns and controls the AGI technology, they could potentially have significant influence and control over the production process, potentially leading to economic inequality and issues related to the distribution of benefits and resources.
This question is central to discussions about the societal and economic implications of AGI, as it raises important issues related to ownership, access, and governance in a future where advanced technology plays a central role in the production of goods and services. It prompts discussions about how to ensure that the benefits of AGI are distributed equitably and that the technology is used for the benefit of society as a whole.
If things continue the way that they are, we run the risk of accelerating the existing trends of capital concentration and even regulatory capture. Regulatory capture means the top players and companies calling for regulation of AI already have the advantage of market presence and dominance and are allowed to participate, and all newcomers will be ‘regulated out of existence’. The fear is that only a handful of top players are allowed to play in the AGI production space. That would be extremely problematic for the rest of humanity.
Even without regulation, one other problem which we will face in the AGI driven economy is the extremely high-bar for entry into the production ecosystem itself. There are three distinct pieces that make up this AI ecosystem today. These building blocks are;
Number 1. Talent,
Number 2. Computational Resources, And,
Number 3. Data, specifically Training Data.
For example, today most of the data used for deep-learning and training AI Models is under the control of companies like Facebook, Apple, Twitter, Amazon, Google, IBM and a handful of others. While most of the compute power is with N-Vidia and the others mentioned above as well.
Control over any of these inputs could allow a firm to gain excessive power over AI markets – especially in the generative AI sector.
Possible anti-competitive practices by incumbent firms may include discrimination against competitors' products. Network effects and platform effects could allow a firm to gain and entrench a dominant position in generative AI markets. Open source models could help democratize access to generative AI, but may also enable misuse if protections are removed.
As AI continues to evolve rapidly, vigilant regulation and antitrust enforcement will be needed to maintain competition and innovation. Beyond the AI algorithms and training data-sets ownership, we will need more sophisticated models of ownership of all other assets as well. For instance co-ops and collective ownership for real estate and resources. If corporate payrolls shrink and many of us lose jobs, how then are we going to redistribute the benefits obtained by any promising technologies? Who is going to reap the rewards?
That is why a robust Redistribution mechanism is essential. Whether or not we change ownership because obviously many people will still want to maintain private property and certainly private property will exist, but not everyone is going to be a landowner. Not everyone is going to own a Robot Factory or a Cloud Computing infrastructure with GPU’s and TPU’s. So, we are going to need some kind of Robust Redistribution policy. That could come in the form of Universal basic income or more sophisticated models.
Unfortunately, money today is like water. It tends to flow towards some kind of an accumulator. In our world, those accumulators tend to be Federal governments and the capitalist class in form of shareholders and corporations. Money tends to flow in those two directions. The future redistributive policies need to distribute the money evenly and fairly and in a controlled policy. The Arthur Laffer's Supply side or ‘trickle-down economics Theory’ just won’t work.
Let’s look at a few examples of collective ownership. Some of them are at a national scale. One example is the Alaska Permanent Fund, which is basically that every Alaskan gets an annual check for ‘Collective ownership’ of the oil reserves in Alaska. The Norwegian Sovereign wealth fund is similar, where, all citizens of Norway benefit. It is a trust fund that is used to fund public services, and it is funded by their oil reserves.
The MonDragon Corporation is an owner-operated Corporation in Spain. The Green Bay Packers are an NFL team that is unique among the NFL as it is collectively owned by 360,000 shareholders.
Sam Altman, the CEO of OpenAI, wants to do something on these lines. He has explicitly said OpenAI, which owns the means of production in form of the Large Language Models (or LLM’s) such as GPT4, GPT5, 6 and 7 etc, and, whatever other future models they come up with, he wants to create a trust fund that is then used to distribute that surplus of wealth.
This is a part of one of his ideas of launching their ‘World coin’ cryptocurrency. Experts however think that this entire project should NOT be piloted and mediated by a Silicon Valley CEO exclusively. Silicon Valley CEO's don't exactly have a good track record in terms of being fair and not allowing power to corrupt them over time. They think that as sincere Sam Altman is, he is still in a position of a lot of privilege and power, and, therefore healthy skepticism and scrutiny is absolutely mandatory.
The point is that collective ownership doesn't necessarily have to happen on a national scale or a state scale. It can happen on a local and smaller scale such as in the form of OpenAI's trust fund, which is similar to land trusts. There are plenty of examples of collective ownership that exists today in this form.
One primary complaint and a pushback about redistributive policies such as Universal Basic Income (or UBI) is the fear of inflation. And there is a valid reason and some merit to this. UBI can actually cause inflation as we did see a small but not negligible effect from the stimulus checks in the USA during Covid 19. But in reality, aggregate supply and aggregate demand and other supply chain issues were a far larger reason for the current inflationary Paradigm that we're seeing globally.
The stimulus checks did cause some inflation in America because they created higher aggregate demand, which is basically that consumers overall have more money to spend, therefore prices go up. Another thing that can drive inflation is inelastic supply. As Global wealth goes up due to AGI, as productivity goes up and demand goes up broadly, there are going to be goods and services and other supplies that do not scale. For instance one of the most limited resources on the Planet is Rare Minerals like Lithium, Cobalt, Praseo-dymium, Cerium, Lan-thanum, Neo-dymium, Samarium and Gado-linium.
No amount of AGI is going to make more Cobalt unless of course it can dig deeper from the Earth's crust or AGI’s land on Asteroids and mine it from there, which is entirely possible but not in the next few decades.
Deflationary pressure is something that lowers prices. Deflation means that your dollar goes much farther. The job losses that can be anticipated from AGI are going to reduce consumer demand until that is somehow fixed or replaced, because, with job losses people are going to hoard their money a little bit more. Job loss results in deflationary pressure because people are spending less money and driving demand down.
Market saturation is another cause of ‘Deflationary Pressure’. For instance some goods and services will continue to go down in price. For example, if you are a writer and get your book professionally edited, you wind up spending several thousand dollars for that. But now with ChatGPT4 and Claude-2 and soon many other LLM’s, which have much larger context windows, you could get your book professionally edited for just fifty dollars. That is a 100x drop in price. It is really difficult for people to imagine the possibility that many goods and services are going to be a thousand time cheaper or even tens of thousands times cheaper in the coming years due to Advanced AI technologies.
This downward pressure due to the abundance of AI is going to be akin to a default raise for everyone. In the case of medicine, the total cost of Health Care in America is 4.3 trillion dollars per year today. Imagine if within a few years, AGI is able to reduce the total cost of Health Care in America by a factor of a hundred. Just imagine if your health care costs drop down to as low as $100 for the best medications, tests and advice. And that's not with socialist policies or single-payer systems where the government foots the bill. This is the actual cost of Health Care which could collapse that much in the long run due to advanced AI. That is a tremendous deflationary pressure.
Another thing to consider is that in many cases there are elastic supplies. Imagine the derivative effects of AI enhancing Supply chains at every level and reducing the need for middlemen and reducing the need for expensive humans and human labor. All the goods and services that you get, even if an AI is not directly used in creating that final good or service, it's still going to help the entire economic process. Which means that the entire economy is going to be a lot more elastic which will also drive down prices.
Now we get to the hardest part of all of this. There is just so much economic incentive to drive down prices. The inevitable conclusion is that there will be massive job losses. Most of us are going to lose our jobs. Conventional jobs are simply not going to exist anymore.
One of the most common human sentiments is that a job or ‘work’ is an essential pillar of a fulfilling life. There are many reasons for this which include; cultural reasons, ethical reasons, religion, spirituality, and even familial history, identity and gender norms. All kinds of stuff such as their self-sufficiency, individualism and self-reliance pushes people to identify with their job. Careers also do a lot for our sense of purpose, excellence, and, confidence. Careers are a way to acquire a sense of independence and accomplishment.
People have deep-seated fears about the prospect of job loss in the wake of advanced artificial general intelligence. Two primary concerns arise: the fear of being unable to provide for oneself and loved ones and the fear of losing personal agency and autonomy to external powers, particularly the government.
One potential solution to these fears is the ancient Greek concept of 'Erete,' which embodies the pursuit of excellence. It involves feeling competent and achieving a sense of mastery and challenge, driven by social rewards and incentives rather than financial gain.
For elites, such as the aristocracy and capitalists, personal competence and achievement often stem from non-career-related activities like academic achievements, independent competitions, participation in public life, and family well-being. This pursuit of excellence extends to activities like sports, adventure trips, and acquiring status symbols, providing a sense of recognition and achievement.
Moreover, tribal belonging, such as cheering for a sports team, fosters a sense of belonging and excellence, demonstrating that fulfillment can exist beyond conventional careers. It is important to note that adapting to life after leaving a traditional career may be challenging initially, as individuals seek ways to feel productive and meaningful.
Even without a conventional career, people can find outlets for pride, mastery, and challenge. Physical fitness serves as a familiar model, where individuals take pride in looking and feeling good, despite no financial incentives. Gamers also experience a sense of excellence when they win or receive recognition for their skills.
The more challenging aspect to address is autonomy and self-determination, which can lead to burnout, rebellion, and despair when lost. Economic instability can result in citizens rebelling against oppressive systems. In societies like the United States, personal autonomy is closely tied to financial power, emphasizing the link between income and freedom.
With post-AGI economics, there is hope that optimal redistribution can increase individual autonomy. This involves both an abundance of free time and a decrease in the cost of goods and services, effectively increasing financial power. However, trust in the government's ability to implement these changes is low, requiring government action to address the issue.
Measuring success in post-AGI economics requires new metrics. Traditional indicators like labor force participation may become less relevant. Instead, measuring economic productivity and individual well-being through means like a well-being index based on Self-Determination Theory, a theory that focuses on internal sources of motivation, including a need for personal growth and fulfillment, becomes essential. This index would consider autonomy, mastery, and connection as key components. Autonomy would measure free time and social mobility, mastery would assess competence and achievements unique to individuals, and connection would focus on social relationships.
Additionally, Maslow's hierarchy of needs could offer a valuable framework to create new Key Performance Indicators (KPIs) for post-AGI economics, ensuring that basic needs are met for everyone. This holistic approach considers the multifaceted aspects of human well-being beyond just financial wealth.
This concludes the series. If you wish to see the Video series please join me on these Links:
Last updated