The US has averaged 1.5% annual productivity growth for 125 years, despite transformative technologies like electrification and the internet. The speaker expects AI to maintain that rate rather than double it, arguing that no single innovation has ever broken through the long-standing ceiling.
My last thought on AI is the following. Over the last 125 years, the US has averaged about 1.5% annual productivity growth. Not for every individual, but on average, okay? Some didn't get any more productive at all. Some got less productive, but as a group, about 1.5%. That 1.5% was driven by electrification, sanitation, health advances, computers, internal combustion— just go on and on and on of all the things— the internet. And all of those things were only able to generate about 1.5% a year productivity growth. My gut is that AI is going to have definitely productivity effects. I think that's primarily where we're going to get the next 1.5% for the next 20 years. I think that's more likely than our 1.5% becomes, and I'm just making up a number, 3%, because resources are gonna get shifted around, right? And we've already climbed some of those mountains. And the only reason I say that is, my God, electrification, computers, chips, medical advances, all those things, 1.5% a year. So it's not like I'm skeptical on AI, quite the opposite. I'm skeptical that it will supercharge our aggregate productivity growth, as opposed to, thank God we got it, we might not get our 1.5%.
