# The Cost of Capital in the AGI Age

> Projects and Policies for a Successful Transition

**Published by:** [Continuations](https://continuations.com/)
**Published on:** 2026-08-09
**Categories:** capital, twac, agi
**URL:** https://continuations.com/the-cost-of-capital-in-the-agi-age

## Content

A few years back we experienced the infamous vibecession. Kudos to Kyla Scanlon for this terrific term which captures a moment where the public feels bad about the economy at a moment where traditional economic indicators are not showing a recession. At the time there was a huge amount of commentary on where the “bad vibes” might be coming from. Lots of theories were advanced, but then, as is always the case, the news cycle moved on. Therefore an important paper published in early 2024 about what was going on received relatively little attention. The paper is quite descriptively titled “THE COST OF MONEY IS PART OF THE COST OF LIVING: NEW EVIDENCE ON THE CONSUMER SENTIMENT ANOMALY” As it turns out one of the co-authors of the paper was Larry Summers – which may have harmed subsequently publicity for the paper. The critical point being examined in the paper is that “[t]he cost of money is not currently included in traditional price indexes.” What does this mean? The numbers that the government and economists are tracking are what you pay if you are buying the product without borrowing. So consider a car that drops in price from $30,000 to $28,000. The car has just become 6.7% cheaper. Suppose though that you have to borrow to buy the car and the interest rate goes from say 7% to 10%. This could be due to a change in overall interest rates or because you personally go from being a prime borrower to being a subprime borrower. Here is what the change in the price of the car looks like for you over four years (1.10^3 $28K) / (1.07^3 $30K) = 1.464 $27K / 1.331 $30K = $40.99 / $39.32K = 1.0425. Instead of 6.7% cheaper, the car has become 4.25% more expensive for you! Now when the authors add the cost of money in, what they find is that their “[...] alternative measures of inflation [...] include borrowing costs and can account for almost three quarters of the gap in US consumer sentiment in 2023.” This is a massive effect. It basically says that as a first approximation the vibecession is explained by the cost of borrowing! By the way, this exclusion was an intentional change in how the consumer price index is calculated made in 1983! This insight is crucially important as we think about policies for the age of AGI. People who need to borrow to pay for cars, houses, etc. experience a fundamentally different economic reality from those who have the capital to make these purchases. Many people will face a two fold pain as AI becomes capable: there will be increased pressure on wages and employment while at the same time interest rates will go up. Why the latter? Because capital is becoming more productive. In 2013 Thomas Piketty published Capital in the 21st Century. The “capital” in the title refers primarily to financial capital (as distinct from physical capital as in the title of my book). Piketty traces what happened to wealth distribution over a long period of time. One of his core concerns is summarized by the inequality “r > g” – the idea that financial wealth could be growing faster than the economy resulting in increased wealth concentration. Now it has been debated whether this condition could ever hold for the economy as a whole over a prolonged period of time. Many in the “let AI rip” contingent believe that the economy will somehow adjust by itself so as to eventually make everyone better off. “Eventually” does a lot of work here, as history has shown repeated prolonged periods of where things got worse for large parts of the population. Daron Acemoglu’s book Power and Progress documents many examples of this. So what then are the relevant policies? It is therefore urgent that we figure out how to address this or we will face the potential of a vibepression! This is why Gigi and I recently signed on to the We Must Act Now statement, and had previously signed on to the AI Pledge for Humanity. There are two policies that absolutely stand out. The first is competition. Only competitive markets turn productivity gains into price reductions. In markets with a lot of market power these result in increased profits (*) instead. “Profits” here carries an important asterisk because in the modern managerial economy these don’t necessarily accrue to shareholders, but often show up as excessive managerial compensation. Nevertheless it is essential to have competitive markets if we want AI productivity gains to result in pervasively lower prices. At present we are far from this across much of the economy. The second policy is some form of basic income or negative income tax. This is essential for two reasons. First, in the absence of a floor we will wind up with insufficient automation as labor will be too cheap (other interventions have way more distortive side effects). Second, income is the basis for demand, which in turn is crucial for preventing an economic depression. Many of the jobs that will remain the purview of humans (qua their being human), such as hospitality and caring professions, cannot support themselves in the absence of paying demand. Essentially as Andy Stern put it in his book on Basic Income we should put a “floor” underneath everyone. This is radically more pro market and creates way more opportunities for adjustment than programs such as Section 8 housing and food stamps. These two policies are crucial not just for the transition period but even matter in a full AGI scenario as you can see in this general equilibrium model (you can find the code on Github). Now Gigi and I have done a lot more than sign on to pledges. Instead of waiting for governments to act, we have started a slew of initiatives to pull forward a positive, high AI future into the present. Two of our initiatives deal directly with the cost of money. The first is our HudsonUP Universal Basic Income Pilot. We have been long time proponents of UBI, see Gigi talking about it nearly a decade ago. There is also an entire chapter about it in my book. If you have access to a steady stream of financial capital, even a relatively small amount, then deflationary technology will make you better off. The second are our steward ownership projects. In these projects we buy and renovate buildings. We then make these buildings available to small businesses or residents at the carrying costs. This means the lease rates only have to cover taxes, insurance and maintenance. There is no current payment to us and since these are 100% equity financed there are no interest payments on a mortgage either. Furthermore, we cannot sell or mortgage the buildings without the consent of the lessees. Finally, we have agreed to share any capital appreciation 50:50 with the leaseholders. The net effect is substantially below market rate leases and a true sense of ownership through both decision making and upside participation. Steward ownership simulates in the here and now a world in which physical capital becomes super cheap to construct. We encourage others who have access to capital today, to also commit some portion of their capital to similar projects. It is not enough to tell people that it will all be fine simply because AGI and robots will make everything cheap. We must show in the here and now what that world might look and feel like. This will get people excited about the policies that are necessary for a successful transition into the AGI age.

## Publication Information

- [Continuations](https://continuations.com/): Publication homepage
- [All Posts](https://continuations.com/): More posts from this publication
- [RSS Feed](https://api.paragraph.com/blogs/rss/@continuations): Subscribe to updates
- [Twitter](https://twitter.com/albertwenger): Follow on Twitter