Michael Duva/Photonica/Getty By Celeste Biever Read more: “Take the visual Turing test” HOW intelligent are you? I’d like to think I know how smart I am, but the test in front of me is making me reconsider. On my computer screen, a puzzling row of boxes appears: some contain odd-looking symbols, while others are empty. I click on one of the boxes. A red sign indicates I made an error. Dammit. I concentrate, and try again. Yes, a green reward! Despite this small success, I am finding it tough to make sense of what’s going on: this is unlike any exam I’ve ever done. Perhaps it’s not surprising that it feels unfamiliar – it’s not your average IQ test. I am taking part in the early stages of an effort to develop the first “universal” intelligence test. While traditional IQ and psychometric tests are designed to home in on differences between people, a universal test would rank humans, robots, chimps and perhaps even aliens on a single scale – using a mathematically derived definition of intelligence, rather than one tainted by human bias. What’s the point? The idea for a universal test has emerged from the study of artificial intelligence and a desire for better ways to measure it. Next year, the most famous test for gauging the smarts of machines will be widely celebrated on the 100th anniversary of the birth of Alan Turing, its creator. The Turing test is, however, flawed. To pass it, a machine has to fool a human judge into believing he or she is conversing with another person. But exactly how much smarter are you than the cleverest robot?