The chaos theorist Doyne Farmer has led an eventful life. He and fellow schemers beat the casinos by inventing a concealable computer that could predict (albeit with low accuracy) where a roulette ball would land. He got a doctorate in physics in 1981 and took a job at the nuclear weapons laboratory in Los Alamos, N.M., where he did “nonlinear studies,” not weapons research. He co-founded a successful quantitative trading firm, the Prediction Co. And, of course, he worked on chaos theory. Now 72, with an ample white beard, he splits his time between Oxford University and his 46-foot sloop.
This is not a look back on Farmer’s illustrious career, because he is as busy as ever. He recently co-founded a company called Macrocosm that’s using insights from chaos theory and a field called complexity economics to fight climate change and other knotty problems. He also just wrote a book, “Making Sense of Chaos: A Better Economics for a Better World.”
Farmer is the director of complexity economics at the Institute for New Economic Thinking at Oxford’s Martin School, which operates separately from the U.S.-based institute of the same name. Both have received major funding from the financier and philanthropist George Soros.
In “Making Sense of Chaos,” Farmer writes that in mainstream models, the economy always tends toward equilibrium, like a rocking horse that rocks when it’s whacked with a stick but eventually settles down (to use the analogy of the Swedish economist Knut Wicksell, who died in 1926).
In fact, Farmer argues, a lot of the economy’s ups and downs come from internal forces, not external ones. As a result, the economy never settles down into a quiet equilibrium. It behaves chaotically.
Chaos to Farmer isn’t a messy sock drawer. It has two qualities: sensitive dependence on initial conditions, such that a tiny change in how things start off can lead to dramatic differences in what happens later; and endogenous motion — the notion that the system never settles because it generates its own churning.
Importantly, Farmer says, chaos is not synonymous with unpredictability: “On one hand, chaos imposes fundamental limits to long-term prediction; on the other, it means that data that otherwise look random can be predictable in the short term.”
Economic forecasters who ignore chaos theory err by working from the top down, trying to predict where the complex system is headed by looking at aggregates such as employment and price levels, Farmer argues. He uses a different approach, which is to build a model of the economy from the bottom up out of “agents” representing individual consumers and businesses. Then let those agents interact and see what behavior emerges. It may be something you never could have guessed.
Weather forecasters do it right, Farmer writes. They gather data from millions of probes measuring the local temperature, humidity and wind speed and feed it into a model that’s a virtual representation of the globe’s atmosphere. Weather forecasts have gotten much more accurate, while economic forecasts have, um, not. (Recall 2008, when Queen Elizabeth II asked during the financial crisis, “Why did no one see it coming?”)
Farmer told me he’s trying not to antagonize economists. He said he went through a draft of the book removing language that he thought would set them off. The best economists are the least threatened by what he’s up to, he said. Lawrence Summers, the former Treasury secretary, gave the book a nice blurb. Farmer leads off one chapter with a quote from a 1996 book, “The Self-Organizing Economy,” by Paul Krugman, my colleague at The Times.
Still, it’s hard not to think of Farmer as some kind of rogue sea dog besetting the economic establishment. That’s reinforced by a video he did for Yale University Press in which he heralds a “fundamentally different way of doing economics” from the deck of his sloop, sailing off the coast of Spain while dressed in a shirt with blue and white horizontal stripes.
Agent-based models like those Farmer is working on have found niches. They’re useful in traffic planning, for example. Farmer thinks an agent-based model his team built may have helped minimize a British housing bubble, although he can’t be sure: “Our results were presented to the U.K. Financial Policy Committee and shortly thereafter they implemented the policy. (We were not privy to the committee’s deliberations and don’t know whether our model influenced their decision.)”
Climate change is for now the main focus of his new company. Macrocosm has only seven employees but will double that, he told me, with a new grant from the National Science Foundation to build an agent-based model to simulate global power and energy systems. He also has a team of about 15 at the Institute for New Economic Thinking at the Oxford Martin School.
Last year I wrote about research by Farmer and three of those colleagues. Using an agent-based model, they concluded that the world could save money by investing heavily now to fight climate change, because the big effort would accelerate learning by doing, thus bringing down costs in the long run.
I suggested to Farmer that the go-fast approach might not work so well for individual companies, because the expertise they develop could go out the door and benefit rivals. He acknowledged the possibility. He said he’ll be studying “how you set things up so there’s enough incentive in a company to push things down the learning curve, but let some of that learning out into the public so it benefits everybody.”
Elsewhere: Nvidia Will Help Teach A.I. in California
Nvidia, which makes chips for artificial intelligence, is teaming up with the government of California to train 100,000 professors, students, developers and data scientists in A.I. skills, and it’s starting with the state’s 116 community colleges. Among other things, students will learn how to use A.I. hardware such as Nvidia’s Jetson, which runs robots, Louis Stewart, the head of strategic initiatives for Nvidia’s global developer ecosystem, told me. Nvidia has similar agreements with Australia, Indonesia, Luxembourg, Thailand and the United Nations.
Quote of the Day
“Through a mutual friend, Larry Page and Sergey Brin became our tenants. I wasn’t particularly interested in what they were doing as long as they remembered to take out the recycling and pay the rent! But, over time, I got to know them more and began using their search engine, Google.”
— Susan Wojcicki, who became the 16th employee of Google and later the chief executive of YouTube, in an interview with the Female Founders Fund (Sept. 18, 2019). She died of cancer on Friday at age 56.
The post Is Chaos the Key to Better Economics? appeared first on New York Times.
