The phenomena you are describing is technological deflation, and yes it can cause the economy to go down. For example, the Great Depression was caused by deflation. One reason this happened was that the gold standard prevented the amount of money from increasing to match the amount of goods and services. The Federal Reserve no longer uses the gold standard, so what would happen instead is lower interest rates and quantitative easing to offset technological deflation resulting in growing GDP. Theoretically if the Federal Reserve hits the zero lower bound and the government refuses to provide fiscal aid, deflation could still happen. For example the great recession in 2008 was caused by a combination of deflation, the zero lower bound and insufficient fiscal aid. Japan too experience deflation, low interest rates and low economic growth for most of the 90′s-2000′s but due to a high savings rate and aging workforce not due to technological inflation.
In either case, when the federal reserve is at the ZLB the proper response is fiscal aid. This could take the form of a UBI (as in the case of the 2020 recession when the government sent checks to everyone) or in the form of other government spending (health care, defense, bridges to nowhere). it’s generally easier to get governments to spend more than to get them to save more, but as noted (2008, Japan) this is not always the case.
Thanks for this comment! I think this one of the main concerns I am pointing at.
I think somethings like fiscal aid could work, but have people tried making models for responses to things like this? It feels like with covid the relatively decent response was because the government was both enforcing a temporary policy of lockdown, and was sending checks to adjust things “back to normal” despite this. If job automation is slightly more gradual, on the scale of months to years, and specific only to certain jobs at a time, the response could be quite different, and it might be more likely that things end up poorly.
Agreed. I think a big part of the reason why we saw a large fiscal response in Covid but not in e.g. 2008 was because it was agreed that it was “nobody’s fault”.
In this sense, the faster that AI produces unemployment, the more likely we will see a policy response. If tens of millions of middle class educated workers suddenly wake up one day without a job, politicians will respond. If, on the other hand, AI slowly squeezes the lowest productivity workers out of a job over the course of 1-2 decades, there will be calls for “reeducation” or “tough love” or some such nonsense as the economy slowly spirals downward Japan style.
Ironically, this then makes for one of the few cases where “going faster” makes the transition to AGI less harmful. Whereas most AI safety issues are worse the faster the transition is.
The phenomena you are describing is technological deflation, and yes it can cause the economy to go down. For example, the Great Depression was caused by deflation. One reason this happened was that the gold standard prevented the amount of money from increasing to match the amount of goods and services. The Federal Reserve no longer uses the gold standard, so what would happen instead is lower interest rates and quantitative easing to offset technological deflation resulting in growing GDP. Theoretically if the Federal Reserve hits the zero lower bound and the government refuses to provide fiscal aid, deflation could still happen. For example the great recession in 2008 was caused by a combination of deflation, the zero lower bound and insufficient fiscal aid. Japan too experience deflation, low interest rates and low economic growth for most of the 90′s-2000′s but due to a high savings rate and aging workforce not due to technological inflation.
In either case, when the federal reserve is at the ZLB the proper response is fiscal aid. This could take the form of a UBI (as in the case of the 2020 recession when the government sent checks to everyone) or in the form of other government spending (health care, defense, bridges to nowhere). it’s generally easier to get governments to spend more than to get them to save more, but as noted (2008, Japan) this is not always the case.
Thanks for this comment! I think this one of the main concerns I am pointing at.
I think somethings like fiscal aid could work, but have people tried making models for responses to things like this? It feels like with covid the relatively decent response was because the government was both enforcing a temporary policy of lockdown, and was sending checks to adjust things “back to normal” despite this. If job automation is slightly more gradual, on the scale of months to years, and specific only to certain jobs at a time, the response could be quite different, and it might be more likely that things end up poorly.
Agreed. I think a big part of the reason why we saw a large fiscal response in Covid but not in e.g. 2008 was because it was agreed that it was “nobody’s fault”.
In this sense, the faster that AI produces unemployment, the more likely we will see a policy response. If tens of millions of middle class educated workers suddenly wake up one day without a job, politicians will respond. If, on the other hand, AI slowly squeezes the lowest productivity workers out of a job over the course of 1-2 decades, there will be calls for “reeducation” or “tough love” or some such nonsense as the economy slowly spirals downward Japan style.
Ironically, this then makes for one of the few cases where “going faster” makes the transition to AGI less harmful. Whereas most AI safety issues are worse the faster the transition is.