Artificial Intelligence and Artificial Problems
BERKELEY – Former US Treasury Secretary Larry Summers recently took exception to current US Treasury Secretary Steve Mnuchin’s viewson“artificial intelligence” (AI) and related topics. The difference between the two seems to be,more than anything else, a matter of priorities and emphasis.
Mnuchintakes a narrow approach. He thinks that the problem of particular technologies called “artificial intelligence taking over American jobs” lies “far in the future.” And he seems to question the high stock-market valuations for “unicorns” – companies valued at or above $1 billion that have no record ofproducing revenues that wouldjustify their supposed worth and no clear plan to do so.
Summers takes a broader view. He looks at the “impact of technology on jobs” generally, and considers the stock-market valuation for highly profitable technologycompanies such asGoogle and Apple to be more than fair.
I think that Summers is right about the optics of Mnuchin’s statements. AUS treasury secretary should notanswer questions narrowly, because people will extrapolate broader conclusions even from limited answers. The impact ofinformation technology on employment is undoubtedlya major issue, but it is also not in society’s interest to discourage investment in high-tech companies.
On the other hand, I sympathize withMnuchin’seffort to warn non-experts against routinely investing in castles in the sky.Although great technologies are worth the investment from a societal point of view, it is not so easy for a company to achieve sustained profitability. Presumably, a treasury secretary already hasenoughon his plate to have to worry about the rise of the machines.
In fact, it is profoundly unhelpful to stoke fears about robots, and to frame the issue as “artificial intelligence taking American jobs.” There are far more constructive areas for policymakers to direct their focus.If the government is properly fulfilling its duty to prevent a demand-shortfall depression, technological progress in a market economy need not impoverish unskilled workers.
This is especially true when value is derived from the work of human hands, or the work of things that human hands have made, rather than from scarce natural resources, as in the Middle Ages. Karl Marx was one of the smartest and most dedicated theorists on this topic, and even he could not consistently show that technological progress necessarily impoverishes unskilled workers.
Technological innovationsmake whatever is produced primarily by machines more useful, albeit with relatively fewer contributions from unskilled labor. But that by itself does not impoverish anyone. To do that, technological advances also have to make whatever is produced primarily by unskilled workers less useful. But this is rarely the case, because there is nothing keeping the relatively cheap machines used by unskilled workers in labor-intensive occupations from becoming more powerful. With more advanced tools, these workers can then produce more useful things.
Historically, there are relatively few cases in which technological progress,occurring within the contextof a market economy,has directly impoverished unskilled workers. In these instances, machines caused the value of a good that was produced in alabor-intensive sector to fall sharply, by increasing the production of that good so much as to satisfy all potential consumers.
The canonical example of this phenomenon is textiles in eighteenth- and nineteenth-century India and Britain.New machines made the exact same products that handloom weavers had been making, but they did so on a massive scale. Owing to limited demand, consumers were no longer willing to pay for what handloom weavers were producing.The value of wares produced bythis form of unskilled labor plummeted, but the prices of commodities that unskilled laborers bought did not.
The lesson from history is not that the robots should bestopped; it is that we will need to confront the social-engineering and political problem of maintaining a fair balance of relative incomes across society. Toward that end, our task becomes threefold.
First, we need to make sure that governmentscarry out their proper macroeconomic role, by maintaining a stable, low-unemployment economy so that markets can function properly.Second, we need to redistribute wealth to maintain a proper distribution of income.Our market economy should promote, rather than undermine, societal goals that correspond to our values and morals. Finally, workers must be educated and trained to use increasingly high-tech tools (especially in labor-intensive industries), so that they can make useful things for which there is still demand.
Sounding the alarm about“artificial intelligence taking American jobs” does nothing to bring such policies about. Mnuchin is right:the rise of the robots should not be on a treasury secretary’s radar.
J. Bradford DeLong, a former deputy assistant US Treasury secretary, is Professor of Economics at the University of California at Berkeley and a research associate at the National Bureau of Economic Research.
By J. Bradford DeLong