My work at OpenAI reminds me every day of the scale of socio-economic change that is coming sooner than most people think. Software that can think and learn will do more and more of what people are doing now. More power will shift from labor to capital. If government policies are not adapted accordingly, the situation for most people will deteriorate.
We need to develop a system that will reimagine this technological future. This system will have to tax assets that will make up most of the value in this world (companies and land) in order to distribute some of the future wealth fairly. Then social processes will cause less disagreement, and everyone will have the opportunity to express themselves in social achievement.
In the next five years, computer programs capable of thinking will read legal documents and provide medical advice. In the next decade, they will work on conveyors and may even become human assistants. In the decades to come, they will do almost everything, including new scientific discoveries that will expand our concept of "everything."
This technological revolution cannot be stopped. And the recursive cycle of innovation will accelerate the pace of the revolution, as these smart machines themselves will help us create smarter machines. This leads to three important conclusions:
This revolution will create phenomenal wealth. The price of many types of labor (which determines the cost of goods and services) will drop to zero as soon as a sufficiently powerful AI "becomes part of the workforce."
The world will change so quickly and radically that it will take an equally radical change in policy to distribute this wealth and enable more people to live the lives they want.
If we do both right, we can improve people's lives more than ever before.
As we are at the beginning of this tectonic shift, we have an amazing opportunity to influence the future. This turnaround cannot simply solve current social and political problems. It must be done in such a way as to promote the development of a radically different society in the near future. The plans of politicians who do not take into account this inevitable transformation will fail for the same reason that the organizational principles of pre-agrarian or feudal societies do not work today.
Next, we will talk about what awaits us and get ready to navigate the new world.
Part 1: The AI Revolution
On a smaller scale of time, technological progress follows an exponential curve. Compare what the world looked like 15 years ago (no smartphones), 150 years ago (no internal combustion engine, no electricity in the house), 1,500 years ago (no industrial machines), and 15,000 years ago (no agriculture).
The coming changes will center around our most impressive skill: the phenomenal ability to think, create, understand and reason. To the three great technological revolutions - agriculture, industrial and computing, we add a fourth: the artificial intelligence revolution. This revolution will create enough wealth for everyone to have what they need if we, as a society, manage it responsibly.
The technological advancements we make in the next 100 years will be far more important than anything we've done since we first took control of fire and invented the wheel. We have already created artificial intelligence systems that can learn and do useful things. They are still primitive, but the trend lines are clear.
Part 2: Moore's Law for Everything
Broadly speaking, there are two ways to ensure a good life: higher earnings (which makes the person they grow richer with) or lower prices (which makes everyone richer). Wealth is purchasing power: it reflects how much we can get with the resources we have.
The best way to increase community wealth is to lower the cost of goods, from food to video games. Technology will quickly help us bring prices down. Consider the example of semiconductors and Moore's Law: Over the decades, chips have become twice as powerful at a constant price every two years or so.
Over the past couple of decades, the prices of televisions, computers and entertainment have dropped in the United States. But other spending has risen significantly, most notably housing, health care, and higher education. Redistribution of wealth alone will not work if these costs continue to rise.
AI will lower the cost of goods and services as labor is a determining factor at many levels of the supply chain. If robots are able to build a house on land you already own from natural resources mined and processed locally using solar energy, the cost of building that house is close to the cost of renting robots. And if these robots were created by other robots, their rental costs would be much less than if they were created by humans.
Likewise, we can imagine AI doctors who can diagnose health problems better than anyone else, and AI teachers who can determine what the student is not understanding in order to explain to them.
"Moore's Law for Everything" should be the watchword of a generation whose members cannot afford what they want. It sounds utopian, but technology will help to achieve this (and in some cases they have already done it). Imagine a world in which, for decades, everything - housing, education, food, clothing, etc. - will become half the price every two years.
We will discover new jobs (they always appear after the technological revolution) and thanks to the abundance, we will have incredible creative freedom in relation to what they will be.
Part 3: Capitalism for All
A stable economic system needs two components: growth and inclusiveness. Economic growth matters because most people want their lives to improve every year. In a zero-sum world with little or no growth, democracy can become antagonistic as people seek to take money away from each other. Distrust and polarization follow from this antagonism. In a fast-growing world, bickering can be much less, because it is much easier for everyone to win.
Economic inclusion means that everyone has a reasonable opportunity to receive the resources they need to live their lives the way they want. Economic inclusion is important because it is fair, creates a stable society, and can make the biggest slices of the pie for most people. As a side benefit, it will give more growth.
Capitalism is a powerful engine of economic growth because it rewards people for investing in assets that create value over time. It is an effective incentive system for the creation and distribution of technological advances. But the price of progress under capitalism is inequality.
Some degree of inequality is normal. In fact, it is critical, as all the systems that have tried to ensure absolute equality have shown. A society that does not offer sufficient equality of opportunity to advance everyone will not last long.
Usually this problem was solved by taxation methods. It didn't work for a number of reasons. This will work much, much worse in the future. While people will still have jobs, many of them will not create more economic value in the way we think of value today. Since AI will produce most of the world's basic goods and services, people will be able to spend more time with loved ones, care for others, appreciate art and nature, or work for the good of society.
Therefore, we should focus on taxing capital, not labor, and we should use these taxes as an opportunity to directly distribute property and wealth among citizens. In other words, the best way to improve capitalism is to empower everyone to directly benefit from it as a capital owner. This is not a new idea, but its implementation will become possible as AI develops, because there will be more wealth. The two dominant sources of wealth will be 1) companies, especially those that use AI, and 2) land with a fixed supply.
There are many ways to introduce these taxes and many thoughts on what to do with them. After a long time, it may be possible to abolish most of the other taxes. Next, we will discuss the idea in the spirit of the beginning of this text.
We could create something like the American Equity Fund. It will be capitalized by taxing companies that exceed an estimate of 2.5% of their market value each year. This value will be paid in shares transferred to the fund and tax at 2.5% of the value of all private land payable in dollars.
All citizens over the age of 18 will receive an annual distribution in dollars and company shares into their accounts. People will be instructed to use the money according to their needs or desires - for better education, health care, housing, starting a company, etc. Rising costs in government-funded industries will come under pressure as more people choose services in a competitive market.
As long as things continue to improve in the country, each citizen will receive more money from the Fund every year (on average; economic cycles will continue). Thus, every citizen will increasingly enjoy the freedoms, powers, autonomy and opportunities that come with economic self-determination. Poverty will be significantly reduced and many more people will have the chance to live the life they want.
A tax paid on a company's shares will align incentives between companies, investors and citizens, whereas a tax on profits will not. Incentives are exceptional strength, and this is a critical difference. Corporate profits can be disguised, deferred or offshore and are often independent of the share price. But everyone who owns Amazon stock wants its price to go up. As the individual assets of people grow with the assets of the country, they are literally interested in their country to succeed.
Henry George, an American political economist, introduced the idea of a tax on the value of land in the late 1800s. This concept is widely supported by economists. The value of the land will increase due to what society will do around it: the network effects of companies operating around the plot of land, public transport making it accessible, and nearby restaurants, cafes and access to nature - all of which will make the land desirable. Since the landlord will not do absolutely all of this work, it is only fair if this value is shared with the wider community that did it.
If everyone owns a share of American wealth, everyone will want America to do better: collective justice in innovation and in the country's success will align our incentives. The new social contract will be a platform for all in exchange for no ceiling, and it will share the belief that technology can and should improve social welfare. (We will continue to need strong leadership from our government to ensure that pushing higher stock prices is consistent with environmental protection, human rights, etc.)
In a world where everyone benefits from capitalism as an owner, the collective focus will be on making the world "better", not "less bad." These approaches are more different than they seem, and society does a lot better when it focuses on the former. To simplify: if we want to do more good, then we try to bake a bigger cake. If we want less bad things, we try to more equitably divide the available pie. Both can raise people's standard of living once, but continuous growth only happens when the pie grows.
Part 4: Implementation and Challenges
The amount of wealth available to capitalize on the American Equity Fund will be significant. The value of US companies alone, measured by market capitalization, is around $ 50 trillion. Suppose that, like the average over the past century, it will at least double over the next decade.
The United States also has about $ 30 trillion in private land (excluding infrastructure on that land). Suppose that this value also roughly doubles over the next decade - somewhat faster than the historical rate - but as the world really begins to understand the shifts that AI will bring, the value of land, as one of the few truly finite assets, should increase at a faster rate.
Of course, if we increase the tax burden on land ownership, its value will decrease compared to other investment assets, which is good for society, since it makes the main resource more accessible and stimulates investment, not speculation. The value of companies will also decline in the short term, although they will continue to perform reasonably well over time.
It is reasonable to assume that such a tax would lead to a 15% drop in the value of land and corporate assets (which will take only a few years to recover!).
Based on these assumptions (current value, future growth, and declines in value due to new taxes), ten years from now, each of America's 250 million adults will be making about $ 13,500 a year. Income could be much higher if AI accelerates growth, but even if it doesn't, $ 13,500 will have a much higher purchasing power than it is now, because technology will significantly reduce the cost of goods and services. And this effective purchasing power will grow dramatically every year.
It will be easier for companies to pay tax every year by issuing new shares that make up 2.5% of their value. Obviously, companies will have an incentive to avoid paying the American Equity Fund tax by going offshore, but a simple check that takes into account the percentage of income generated from America can solve this problem. The bigger problem with this idea is the incentive for companies to return value to shareholders instead of reinvesting it in growth.
If we tax only public companies, companies will also have an incentive to remain private. For private companies with annual revenues in excess of $ 1 billion, we can allow equity tax to accumulate for a certain (limited) number of years until they go public. If they remain private for a long time, we can allow them to pay the tax in cash.
We will need to develop a system so that people are not constantly chasing a lot of money. A constitutional amendment defining acceptable tax ranges might come in handy. It is important that the tax is not so large as to constrain growth - for example, the tax on companies should be much less than their average growth rate.
We will also need a reliable system for estimating the actual value of land. One way is to organize a corps of influential federal assessors. Another option is to allow local governments to conduct assessments, as they do now when determining property taxes. They will continue to receive local taxes with the same assessed value. However, if a certain percentage of sales in any jurisdiction in any particular year falls too low or falls below the local authorities' valuation of property values, then all other properties in their jurisdiction will be revalued up or down.
In theory, an optimal system should only tax the value of the land, not the improvements built on it. In practice, this value may be too difficult to estimate, so we may need to tax the value of land and land improvements (at a lower rate as the total cost will be higher).
Finally, we cannot allow people to borrow, sell, or otherwise pledge their future fund allocations, or we cannot solve the problem of equitable distribution of wealth over time. The government can simply make such transactions illegitimate.
Part 5: Transition to a new system
Creating a great future is not difficult: we need technology to create more wealth, and politics to distribute it fairly. Everything you need will be cheap and everyone will have enough money to afford it. Because the system will be so popular, politicians who adopt it early will be rewarded: they will become extremely popular themselves.
During the Great Depression, Franklin Roosevelt was able to put in place a huge social safety net that no one would have thought of five years ago. We are now in a similar situation. Thus, a movement that stands for both business and people will bring together an extremely wide audience.
A politically viable way to launch an American Equity Fund that can mitigate the transition shock is through a law that gradually moves us to a 2.5% rate. The full rate of 2.5% will come into effect only after the GDP has increased by 50% since the adoption of the law. If we start with small payments, they will both motivate and help people get used to a new future. It looks like it will take a long time to achieve 50% GDP growth (it took 13 years for the economy to grow 50% to 2019 levels). But once AI comes along, growth will be extremely rapid. Going forward, we can probably reduce many other taxes as we tax these two main asset classes.
The coming changes cannot be stopped. If we embrace them and plan things out correctly, we can use them to create a fairer, happier and more prosperous society. The future can be almost unimaginably great.
Thanks to Steven Adler, Daniel Amodei, Adam Beibat, Chris Bazer, Jack Clark, Ryan Cohen, Tyler Cowen, Matt Danzeisen, Steve Dowling, Ted Friend, Lachi Grum, Chris Hallacey, Reed Hindrumoffman, Naygrutuyt Naytuer Konstantinidis, Andrew Cortina, Matt Krisiloff, Scott Krisiloff, John Lattig, Eric Madsen, Preston McAfee, Luke Miles, Arvind Nilakantan, David Oates, Cullen O'Keefe, Alethia Power, Raul Puri, Ilya Suzkeverni, and Voychel Waltz for review drafts of this text, and also to Gregory Koberger for its development.