Is current AI displacing workers without delivering productivity gains?
Core argument: Had the labor force kept growing in line with history and GDP, there would be nearly 5M more workers, down from a gap closer to 7M a year earlier.
If the labor force had continued to grow more or less in line with history and GDP, we’d have almost 5M more workers out there. But we don’t. The gap is shrinking—it was closer to 7M a year ago—but it is still a very large number, and, given retirements, skill mismatches, and aging, it seems unlikely we will close that gap quickly, if ever. Absent a wave of immigration, which creates its own problems, politically and otherwise, and doesn’t necessarily fill the observed skill gaps, something needs to change. Historically, when this has happened—labor became more expensive than capital—economies have responded with automation, so we should expect that again today. We think, contrary to hyperbole, we are already at the tail end of the current wave of AI. We need to look past the limits of current AI technology if we are to break free from the past few decades of tech—enabled automation, where there is high worker displacement without commensurate productivity gains impact—where minimal human flourishing is created.
