There aren’t enough data scientists to go around—unless you automate them.
Demand for statisticians and data experts outstrips supply. The shortfall in the U.S. alone could reach 190,000 workers by 2018, according to estimates by McKinsey & Company.
Michael Schmidt has created an automated data scientist that can take in observations about the world and spit out theories to explain them.
Schmidt showed it could be done in 2009. He then wrote software that could examine raw data and derive laws of physics, like the one that describes the swinging of a pendulum.
Since then Schmidt has refined the software, named Eureqa, so that it can handle more than just physics questions: astronomers have used it to characterize galaxies, and doctors have used it to predict which children will have acute appendicitis. Since 2011, Schmidt has been running a startup, Nutonian, that offers the software to business users who aren’t math experts. It hopes to win over businesses like the retailer Lowe’s, which has piles of sales data and yearns to uncover equations that might help it sell more gas grills or two-by-fours. “The people with the skills are going to Google or SpaceX or to Wall Street, not to home-improvement chains,” says Schmidt. “Our mission is to help with that, and show that you don’t need to be a data scientist to make useful discoveries.”