This essay reviews seven books from the past dozen years by social scientists examining the economic impact of artificial intelligence (AI). These works offer valuable insights—AI as cheap prediction, architectural barriers to adoption, data as an economic asset, implementation challenges. However, they offer little guidance when it comes to the transformative scenarios considered plausible by many AI researchers. Economists have made great progress in explaining how to use AI within existing production functions, who benefits, and why; what remains needed is rigorous advice to policymakers concerned about rapid increases in labor churn, scientific development, labor–capital shifts, or existential risk. (JEL C45, C80, D83, O31, O36)

Eric A. Posner of University of Chicago reviews “The Problem of Twelve: When a Few Financial Institutions Control Everything” by John Coates. The Econlit abstract of this book begins: “Explores issues that arise when a small number of actors acquire the means to exert outsized influence over the politics and economy of a nation, focusing in particular on the two late-twentieth-century institutions of index funds and private equity funds.”

On the fiftieth birthday of my Foundations of Economic Analysis, a deluxe edition of it was embalmed in the German Klassiker der Nationalokonomie series alongside of Adam Smith, Eugen von Bohm-Bawerk, Irving Fisher, and many other illustrious suspects. With it, as customary, was published a slim volume in German, a Vademecum, with review essays by Jurg Niehans, Carl-Christian von Weizsacker, and a foreword by the editor Bertram Schefold. By invitation, like Tom Sawyer at his own funeral, I provided for German translation my own recollections under the title "How Foundations Came to Be." Here is the English original, slightly abridged; for some technicalities, readers are referred to the full German text. I remembered much, and, with the perspective of time, learned not a little.

AHEORY of public finance remains unsatisfactory unless it comprises both the revenue and expenditure sides of the fiscal process. The classical (RicardoMills-Edgeworth-Pigou) tradition of a taxation-only view neglected this axiom. Holding expenditures unproductive, or disregarding them altogether, the task was to arrange taxes so as to impose equal (or least total) sacrifice. As a theory of taxation, this approach collapsed with the old welfare economics; and as a theory of public finance, its exclusive concern with taxation bypassed the central problem of how to allocate resources for the provision of social goods. Subsequently, various attempts were made to combine the revenue and expenditure sides in a more satisfactory system. We shall note these briefly, and then consider how cost-benefit analysis fits into the picture.

philosophical (i.e., conventionalist) criteria but rather they, too, are empirically based, hence can be expressed in instrumentalist terms: Simpler means requires less empirical knowledge (the word initial refers here to the process of generating predictions with something like modus ponens). More fruitful means more applicable and more precise [4, p. 10]. The possibility of a tradeoff is discussed. Friedman explicitly rejects the necessity of requiring the of substantive hypotheses before they are used simply because it is possible. But here it should be noted that his rejection of is partly a consequence of his use of the word testing. Throughout his essay always means testing for (in some sense). It never means testing in order to reject as most of his critics seem to presume. That is, for Friedman a successful test is one which shows a statement (e.g., an assumption, hypothesis, or theory) to be true; and, of course, a minimum condition for a successful test is that the statement be inconsistent with empirical evidence (see [4, pp. 33-34]).14 Appreciating the success orientation of Friedman's view is essential to an understanding of his methodological judgements. For Friedman, an instrumentalist, hypotheses are chosen because they are successful in yielding true predictions. In other words, hypotheses and theories are viewed as instruments for successful predictions. It is his assumption that there has been a prior application of modus tollens (by evolution, see [4, p. 22]), which eliminates unsuccessful hypotheses (ones that yield false predictions), and which allows one to face only the problem of choosing between successful hypotheses. In this 13 Note here, although Friedman uses conventionalist criteria, it is for a different purpose. For a conventionalist the criteria are used as status substitutes; in conventionalism one finds that theories are either or worse. In this sense, Friedman can be seen to pose the problem of choosing among theories already classified as better in his sense (successful predictions). 14 I stress, this is the view Friedman uses in his essay. In recent correspondence Professor Friedman has indicated to me his more general views of in which success might be either a confirmation or a disconfirmation. But he still would question the meaningfulness of testing in order to reject. Although Friedman seldom uses the word truth, it should be noted that throughout he consistently uses the word (by which he always means at least not inconsistent with the available facts) in the same sense that truth plays in modus ponens seemingly while also recognizing that modus ponens is assured only when applied to truth in the absolute or universal sense (i.e., without exceptions). Technically speaking his use of the word may lead one to the incorrect identification of truth with validity. In this regard, applications of Friedman's methodology are often confused with orthodox conventionalism. This confusion can be avoided by remembering that is a (but sufficient) condition of empirical truth-hence, validity and are identical-and by recognizing that someone can believe his theory is true, even though he knows he cannot prove that it is true. This content downloaded from 40.77.167.14 on Tue, 20 Sep 2016 06:14:18 UTC All use subject to http://about.jstor.org/terms 512 Journal of Economic Literature, Vol. XVII (June 1979) sense, his concentrating on successful predictions precludes any further application of modus tollens. And similarly, any possible falsity of the is thereby considered irrelevant. Such a consideration is merely an appreciation of the logical limitations of what I called reverse modus tollens (above, Section 1.2). And since he has thus assumed that we are dealing exclusively with successful predictions (i.e., true conclusions), nothing would be gained by applying modus ponens either. This is a straightforward appreciation of the limitations of what I called reverse modus ponens. Knowing for sure that the hypotheses (or assumptions) are true is essential for a practical application of modus ponens, but such knowledge, he implies, is precluded by the absence of an inductive logic [4, pp. 12-14]. By focusing only on successful hypotheses, Friedman correctly reaches the conclusion that the application of the criterion of simplicity is relevant. He says there is virtue in a simple hypothesis if its application requires less empirical information. One reason a simple hypothesis can require less information, Friedman says, is that it is descriptively false [4, pp. 14-15]. (For example, a linear function requires fewer observations for a fit than does a quadratic function.) This raises the question of descriptions versus necessary abstractions. Friedman explicitly recognizes that some economists (presumably, followers of Lionel Robbins) hold a view contrary to his. For them the of a theory is considered to be a direct result of the descriptive realism of the assumptions. But Friedman claims the relation between the significance of a theory and the realism of its assumptions is almost the opposite. . . . Truly important and significant hypotheses will be found to have assumptions that are wildly inaccurate descriptive representations of reality, and, in general, the more significant the theory, the more unrealistic the (in this sense)....

A FLAGRANT ERROR dogged James Steuart, Adam Smith, David Ricardo, John Barton, John Stuart Mill, Karl Marx, and classical writers generally. Modern commentators on classicism are enough tempted by the fallacy to generally overlook it as a vital flaw in the earlier writers. This fallay can be called classically classical-not in the sense of being the greatest error in the classical paradigm (an award that would be hard even to define) but-in the sense that modern scholars ignorant of pre-1900 literature would be little tempted by it. I do not denigrate or patronize a great writer of the past, such as David Ricardo, when I objectively delineate his hits and misses. The fallacy can be simply put. Fixed capitals are prejudicial to wages and the demand for labor; circulating capitals (wage fund items that represent outlays paid to workers at the beginning of a period of production, which are not recouped until the end of that period of production and which of course bear an interest or profit rate during the transition) are allegedly favorable to the real rate and to the demand for labor. Fixed capitals are durable produced inputs that render their services over a number of different production periods. Circulating capitals are produced inputs used up within one period of production; they are relatable to, but distinguishable from, wage fund advances paid to workers at the beginning of a period of production, to be recouped at its end along with an interest or profit return. In the minds of heterodox economists and the lay public generally, a technological change that made machinery newly viable was supposedly the kind of invention that could put people out of work temporarily, reduce market-clearing rates, and in long-run equilibrium at an unchanged subsistence rate call for a significantly reduced population. By contrast, therefore, a new invention that displaced machinery in favor of various raw materials as inputs, would supposedly raise the short-run real and increase the demand for labor. So powerful was the grip of what I shall dub the Classical Fallacy that it was being reminded of it in 1819-21 that appears to have enabled Ricardo to recant, in his famous Third Edition chapter on machinery, his previous boner that every viable invention can be expected to raise every factor's return. From today's wisdomor indeed the wisdom of 1900-that previous position of Ricardo was nonsense. There is no Invisible Hand that seeks out machines or new techniques only if they benefit everyone.I