Time Regained

In 1992, in my junior year at Northwestern, I sat in Roger Myerson’s game theory class in the Mathematical Methods in the Social Sciences program and could barely follow the formulas. Professor Myerson would later win the Nobel Prize in economics. Across my years at Northwestern and the University of Chicago I had the good fortune to study under seven Nobel laureates, the seventh being Joel Mokyr, the Northwestern economic historian honored last year for identifying what sustained growth through technological progress requires. When I wrote my book, The Scholars’ Treasure, there were six, and, as I confessed there, not one of them gave me an A.

I would rather have been reading Shakespeare. Northwestern let me do both: history and philosophy beside an intense program in mathematical methods, economics and, above all, game theory. Then, from 1997 to 2000, the University of Chicago’s business school added the models of finance: investments, securities, options, valuation. Harry Davis became my mentor there. I earned an A in Steve Kaplan’s famously demanding Entrepreneurial Finance course, in the classes that sent many of my classmates on to run hedge funds and sovereign wealth funds. I left with a toolkit of nearly two hundred mathematical models, each one a key to a particular kind of lock.

Then I put the toolkit away. For almost thirty years I built companies in the world of expert networks and expert witness referrals: starting them, growing them, selling them, buying them back, and writing books about it all. There is an irony in that career I only saw later. Much of my working life was spent introducing clients to the very people who had written those models, the professors whose theorems I had sweated over in Evanston and Hyde Park. I knew exactly where the keys were. I rarely turned one myself.

This year two births brought it all back. My son Solomon arrived, and with him a month at home. And the AI models arrived, good enough at last to work beside me. In college I dreamed of becoming an investment analyst, a fund manager or a financial writer. At fifty-five, I decided to stop dreaming about it.

Proust’s great novel, which many of us first met as Remembrance of Things Past, was one of the most profound books I have ever read. It can seem to have no plot at all: thousands of exquisite pages of memory, a madeleine dipped in tea and a whole childhood rising out of the cup. The plot is there, but Proust hides it until the last volume, Time Regained. Only there does the narrator understand that the years he thought he had wasted were the material for the work he was meant to do. The time was never lost. It was waiting.

That is how these notes feel to me. A segment on CNBC, a story in The Wall Street Journal or Barron’s, a remark from Jim Cramer, and a model I learned in 1992 comes back the way the madeleine brought back Combray. Proust wrote that the only true voyage of discovery would be “not to visit strange lands but to possess other eyes.” A good model is a pair of other eyes. Game theory sees a price war as a truce that holds until it pays to break it. Option pricing sees the value of waiting. A stock and a flow see how a rising margin can hide a falling bank balance.

So, yes: AI helps me write these notes, and the line at the foot of every one says so. I do not apologize for it. I did the math the long way. I put in my ten thousand hours thirty years ago, on the problem sets, the exams and the papers, and then I parked it. Now I get to work with the most capable AI models in the world, and I have built a research desk on top of them that carries the same library of models I learned, catalogued and explained on this site for anyone to read. The idea, the question and the judgment are mine. The machine is a tireless research partner: it reads the filings, checks every number in code, and argues with me until the draft holds up.

Sherlock Holmes warned that “it is a capital mistake to theorize before one has data.” That warning is the method. We start from a question, choose the model that fits it, and test it against everything already published. Then we write down, plainly, the data that is still missing, and go and get it from the people who know: former executives, engineers and specialists who have seen the thing from the inside. That is what I mean by primary research in the age of AI. The conversation with an expert still matters most. It now comes last, aimed at the one fact that would change the answer.

Solomon wrote that “it is the glory of God to conceal a matter, and the glory of kings to search it out.” I now have a son named Solomon, a library of models, and a partner that never tires of searching. The time was not lost. It was waiting.

Russ Rosenzweig