Roll the dice, raise the drama

MIT, he says, was a mind-blowing experience, beginning on day one. “In high school, we learned calculus from a textbook by MIT professor George Thomas,” he recalls. “The first class I attended at MIT was taught by George Thomas. I was in awe. He was like a rock star to me.”

Thomas introduced Shiller to a three-step approach to tackling challenges: to begin, formulate an initial solution to the problem; then draft a general theory from that solution; and finally challenge your conclusions with outlier cases. “Learning is about struggle and discovery and making mistakes,” he says. “It requires curiosity and an open mind.”

Shiller played violin with the MIT Symphony Orchestra for four years—three of those as concertmaster. After graduation, he worked as a professional musician for three years and then shifted into computers and technology. Over the next 20 years he formed and sold an accounting software company and ran another company that published CD-ROMs. All that time he continued to think about math and math education. 

In the late 1990s, inspired by what he saw his children learning at a Montessori school in Princeton, New Jersey, he teamed up with the Princeton Center for Teacher Education to develop ShillerMath—a math curriculum for ages three to eight based on Montessori principles and designed for home use. First published in 2000, the curriculum brings the advantages of Montessori schools—which promote hands-on, self-directed learning—to families who might not be able to afford the schools’ tuition. “I think kids are naturally mathematical and philosophical,” says Shiller, who expanded the ShillerLearning curriculum to include language arts in 2010. “Even a two-year-old knows when something is true or false.”

Now living in Texas, Shiller continues to pursue ventures in IT and data—the most recent of which is a startup that aims to use metadata (the digital labels attached to data) to make internet and AI searches more accurate. “One of the reasons AI hallucinates is because the data it is fed has not been accurately labeled,” he says. “We want to create a metadata-driven infrastructure ecosystem. It would make AI more grounded. We’ve had a lot of interest in this.”

When Shiller isn’t performing on violin, attending IT and data conferences, or developing curricula, he can often be found playing backgammon, a game whose blend of complex mathematical strategy and pure chance he’s enjoyed since his MIT days. At tournaments, he both competes and provides play-by-play commentary on YouTube. He is a veteran broadcaster—years ago he hosted a radio show in Connecticut, where he says he “learned to make a fool of myself in public”—and he likes to insert drama into his broadcasts. “Most commentators talk about probability and strategy,” he says. “I like to talk about what might happen if a highly unlikely but catastrophic roll of the dice comes up. I call it the exuberant American approach. It’s not for everybody, but it’s a lot of fun.”