certainly an oversimplified logic, tell me what you think:

AGI is an enormous increase in efficiency that could lead to the typical technological deflation. the same products and services could now be offered much more cheaply. As mass unemployment is also looming, so privat demand will fall for a transitional phase, which will further reduce prices.

Central banks combat deflation by lowering interest rates. And if interest rates are already at zero, with quantitative easing. Central banks may even become very orthodox and buy large portions of the stock market to inject money into the economy and combat AGI-deflation.

So I assume a zero-interest phase after AGI.

@Jan53274 I think this is sound.

@Jan53274 you're thinking of nominal interest rates, but this question asks about real rates (factoring out inflation/deflation). Example: if the nominal interest rate, which is the number typically discussed, and set by the banks, is 5% and inflation is 3%, the real interest rate is 2%. In a deflationary scenario, if everything gets 25% cheaper and the nominal interest is 0%, the real interest rate is 25%.

If you expect that AGI will make things significantly cheaper, on the order of tens of percentage points per year, that would require a sh*t-ton of quantitative easing to counteract the deflation. Personally, seeing how many hundreds of billions to trillions of dollars of QE it took to respond to the financial crisis in 2008, and expecting AGI to cause much larger deflationary pressures than that, I don't think the world's central banks will have the appetite for putting that much debt on their balance sheets just to keep an inflation number above zero. That level of monetary stimulus will likely cause far more problems than it solves. In the financial crisis, the problem was "not enough credit", so "increase the money supply by issuing bonds and such which make credit markets more liquid" was a response likely to affect the cause of the problem. In this case, "goods are much cheaper so flood the economy with money" could in theory keep that one number in the desired range, but it's likely to cause a whole bunch of distortions as ginormous amounts of newly created money slosh around the economy.

@equinoxhq Thank you! English is not my mother tongue, so I got the terms mixed up. In this case, I am also assuming rising real interest rates.

But I think central banks have no choice but to massively increase the money supply to prevent deflation. That is their political mandate. And their balance sheets are theoretically irrelevant.

I suspect that in a number of countries there will gradually be state financing by central banks,. this is still a taboo but transformative AI could shakes up the labor market to such an extent that extremely expensive welfare programs such as unconditional basic income will have to be financed. But I'm drifting off track 😄

@Jan53274

But I think central banks have no choice but to massively increase the money supply to prevent deflation. That is their political mandate.

But political mandates can change. That political mandate is based on an understanding of how the economy works that says a little inflation is good, but too much, or deflation, is bad, AND we know how to keep inflation in the "just right" range, without causing too many problems in other areas of the economy. If we for example had some kind of bioweapon used where our food supply was reduced by 75%, the political priority would be to make sure as many people as possible got fed and the social order was maintained, and the fact that food was now much more expensive, and so the inflation number is "a lot", would be fine with the government of the day. If suddenly everything costs a fraction of what it did because of massive productivity improvements, the political priority will be how to distribute the stuff and secure the means of this new productivity, and what's happening with inflation or deflation would be secondary.

And their balance sheets are theoretically irrelevant.

That's complicated. But my general take is, we're all playing by a shared set of rules, buying into a shared set of fictions like "democracy" "McDonalds" and "central banks", which is what allows our societies to function as they do. Start breaking the rules by which the shared fictions are assumed to operate, and the system breaks down. So: there's a range within which "central banks can do things with their balance sheets that effectively create money from nothing" is true, and outside of that range people sit up and go "hold on a second, you can't triple the size of the money supply in a month by doing a database update in your financial software, I don't know why that won't work, but that's not how things are supposed to work". Probably there are stakeholders within the financial system who depend for their financial success on the assumption that something like that won't happen, and their response acts as a constraint on what a central bank can do, I'm not exactly sure, but I'm pretty sure if everyone wakes up one morning and it's glaringly obvious that some fiat currency is just purely a shared fiction which a central bank is now messing with in unexpected ways... the results will not be good. At the very least, people's expectation that their bank accounts will be worth something in a week will be much less certain, if central banks start doing things that are way outside of the expected range... and the effects of that uncertainty will be bad.

I suspect that in a number of countries there will gradually be state financing by central banks,. this is still a taboo but transformative AI could shakes up the labor market to such an extent that extremely expensive welfare programs such as unconditional basic income will have to be financed.

I think however a state finances a UBI, it has to be backed by something other than central banks turning the money supply into an increasingly obvious fiction. I'm not sure how it would work out, in the case that some new technology makes most necessities cheap enough that they're easy to just give to everyone, but I don't think "central banks create more money backed by nothing" is the answer. Maybe some of taxes goes to buying enough robots and AIs to provide all the basic necessities of life to everyone, and by that means a UBI of sorts is achieved. Maybe some people get rich enough that they can just give everyone universal access to the necessities of life out of their spare change, and because they're nice people and the cost to them is small, they choose to do so. Maybe taxing a small percentage of the huge gains from AI-powered productivity provides enough revenue for a UBI through cash transfers. Maybe a nation nationalizes an AI company, and the revenue from said company funds a UBI. Maybe some future quadrillionaires (inflation/deflation adjusted) set up a philanthropic foundation with enough productive resources that its output provides the necessities of life to everyone. Maybe something else. But "central banks finance governments by making database updates wherein the number of dollars the government has in its account goes up, and everyone just shrugs" doesn't seem like the way forward.

@equinoxhq The balance sheet is just a record of money issued to date. If the balance sheet were lost, it would have no effect. Central banks create money out of nothing and don't have to pay it back, so in that regards: the balance is irrelevant.

"making database updates" sounds like cheating. But in fact, the Fed and the ECB are extremely reliable and predictable based on the inflation rate and the markets generally trust them. And if we get tranformative AI and central banks continue to play by their mandate - i.e. by the rules - they would have to inject money into the capital markets, the state coffers or even transfer it directly to the citizens (helicopter money). But the increase in the money supply would remain closely aligned with the inflation target. This would not be as arbitrary as you fear.

And it sounds much more realistic to me than quadrillionaires financing the whole of society.

Edit: ah the f- real rates.

https://marginalrevolution.com/marginalrevolution/2015/04/tyler-cowens-three-laws.html

@jgyou It's a bit odd to poll on real rates, but I should have been more careful.

This is in reaction to this column: https://www.bloomberg.com/opinion/articles/2024-03-27/ai-could-have-a-surprising-effect-on-interest-rates

And I flip-flopped between real and nominal myself (hence Rana's post and the following chain). Doesn't help. 😬

The raise or lower will depend on the starting position and where the economy will be.

Assuming the decision about interest rate is only due to ASI being achieved, then capital will have much more value, real interest rate will go up so neutral nominal rate needs to go up.

.

I don't want to get fancy with attribution or counterfactual starting states. Otherwise, this devolves into endless methodological hairsplitting. Hence the coarse measurement instrument.

I am happy to take suggestions or redirect traffic to a competing, better-targeted market.

@jgyou Tyler is talking about*real* interest rates - nominal minus inflation. Thesis is that demand for capital will rise. Rising nominal rates could just reflect inflation.

@jgyou ignore the unedited version

@MichaelSavage Does this sound like a good authoritative source to you? https://fred.stlouisfed.org/series/REAINTRATREARAT10Y

@traders I updated the data source. If you feel like changing your position, you can sell out of your position within the next 24h, and I will reimburse any $M losses.

@jgyou yes, I think that's absolutely the right bet.