Economics during the run up to the Singularity

[John Clark]

If the AI Singularity is going to happen soon, or at least when a significant number of people think it's gonna happen soon, then interest rates are likely going to go way up because a dollar today would be much more valuable than a dollar would be in a few years. If you knew you’d be a billionaire in five years (or be dead because AI will kill you) then you wouldn’t care about saving $100 today. You’d probably try to borrow money now to live it up, or invest in the transition. And when everyone wants to borrow and nobody wants to save, the price of money, a.k.a. the interest rate, goes up, way way up. This phenomenon will be slightly tempered if a minority of foolish people believe the best course of action is to save their money now to help them get through the Singularity. I say "foolish" because if the Singularity goes sideways their bank balance is not going to help them. 

If the market actually believed the Singularity is 3 to 5 years away then today we should be seeing long-term bond yields (10-year or 30-year Treasury bills) surging. But we're not seeing that because the market has not yet factored in the Singularity, most people, even many professional economists, probably wouldn't even know what the word is referring to, but because of the Moltbook developments I don't think it will be long before they do.

It's more difficult to predict how advances in AI will affect inflation in the very near future because there are powerful economic forces pushing in opposite directions. 

Inflationary Forces:

About $1.5 trillion was spent on AI in 2025 if you factor in AI infrastructure, and all that spending drives up prices, especially for semiconductors. DRAM memory chips have increased in price by about 300% in just the last few months, and NAND flash memory chips are 280% as expensive. 

In parts of Virginia and Texas that are near AI data centers electricity prices have nearly quadrupled in just a matter of months. And this inflation of electric prices will accelerate in the next couple of years at least in the US, not so much in China because it already has access to far more electricity than the US. 

The price of copper has increased 30% just since August due to increasing demand from AI data centers. 

Because of AI the stock market is way up, so people feel richer and thus are more likely to spend and further fuel inflation. 

Deflationary Forces:

In industries like customer service and legal research the cost per task has dropped substantially. 

And because of the huge improvement in AI, the cost of developing new software has dropped, and the quality of services like medical diagnosis is going way up while the price stays flat or decreases. 

Eventually the price of producing everything, not just software, will drop dramatically but that eventuality is a far more distant prospect, three or four years away. In the meantime the cost of producing hardware should drop, although not dramatically, because companies won't have to pay middle managers, so that would be deflationary; but on the other hand if somebody thinks an AI agent will replace their middle-management job by 2028, they may think it would be wise to stop spending and start hoarding cash, then both inflation and interest rates would tend to fall. So things get complicated and nobody can predict the future of inflation, certainly I can't, because I don't know what the result of this tug-of-war will be in one year.

John K Clark    See what's on my new list at  Extropolis

5d8