This article first appeared on the American Enterprise Institute site.

So not great timing for this quote by Treasury Secretary Steven Mnuchin at an Axios event: “I think that is so far in the future—in terms of artificial intelligence taking over American jobs—I think we’re, like, so far away from that [50 to 100 years] that it is not even on my radar screen.”

Not only are there seemingly daily reports on advancing automation and potential job market impacts—good and bad—but there is this new study from Daron Acemoglu and Pascual Restrepo: “Robots and Jobs: Evidence From US Labor Markets.”

First the summary:

As robots and other computer-assisted technologies take over tasks previously performed by labor, there is increasing concern about the future of jobs and wages. We analyze the effect of the increase in industrial robot usage between 1990 and 2007 on US local labor markets.

Using a model in which robots compete against human labor in the production of different tasks, we show that robots may reduce employment and wages, and that the local labor market effects of robots can be estimated by regressing the change in employment and wages on the exposure to robots in each local labor market—defined from the national penetration of robots into each industry and the local distribution of employment across industries.

Using this approach, we estimate large and robust negative effects of robots on employment and wages across commuting zones. We bolster this evidence by showing that the commuting zones most exposed to robots in the post-1990 era do not exhibit any differential trends before 1990.

The impact of robots is distinct from the impact of imports from China and Mexico, the decline of routine jobs, offshoring, other types of IT capital and the total capital stock (in fact, exposure to robots is only weakly correlated with these other variables).

According to our estimates, one more robot per thousand workers reduces the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent.

And this from later on in the paper:

We also document that the employment effects of robots are most pronounced in manufacturing, and in particular, in industries most exposed to robots; in routine manual, blue collar, assembly and related occupations; and for workers with less than college education.

Interestingly, and perhaps surprisingly, we do not find positive and offsetting employment gains in any occupation or education groups. We further document that the effects of robots on men and women are similar, though the impact on male employment is more negative.

I would guess this is going to be a pretty important paper. First, the “rise of the robots” has been a hot topic in the news. Second, this paper seems like a natural follow-on to the “The China Shock: Learning From Labor Market Adjustment to Large Changes in Trade” by David Autor, David Dorn and Gordon Hanson (ADH). Their work on the costs of trade was perhaps the most influential bit of economic research last year, with its impact continuing. Indeed, Autor’s trade work is frequently mentioned in the Acemoglu and Restrepo paper.

Related: Could a robot run for president in 2020?

Just as ADH aren’t against trade but highlight the need to focus on its downsides, this new paper is certainly not arguing we should stop technological progress. But the conclusions are very different between this paper and the China paper.

In the conclusion to the latter, ADH write that the “great China trade experiment may soon be over, if it is not already. The country is moving beyond the period of catch-up associated with its market transition and becoming a middle-income nation. Rapidly rising real wages indicate that the end of cheap labor in China is at hand.”

But the automation challenge is only beginning, as Acemoglu and Restrepo conclude:

Because there are relatively few robots in the US economy, the number of jobs lost due to robots has been limited so far (ranging between 360,000 and 670,000 jobs, equivalent to a 0.18- 0.34 percentage point decline in the employment to population ratio).

However, if the spread of robots proceeds as expected by experts over the next two decades (e.g., Brynjolfsson and McAfee, 2012, especially pp. 27-32, and Ford, 2016), the future aggregate implications of the spread of robots could be much more sizable.

James Pethokoukis is a columnist and blogger at the American Enterprise Institute.