Confidence in your abilities is usually a good thing – as long as you can recognise when it's time to ask for help. As we build ever smarter software, we may want to apply the same thinking to machines. An experiment that explores a robot's sense of its own usefulness could help guide how future artificial intelligences are built. Overconfident AI can cause all kinds of problems, says Dylan Hadfield-Menell at the University of California, Berkeley. Take Facebook's newsfeed algorithms, for example.

If you're just talking about the time until we see some driverless vehicles on the road, probably not that long. Anthony Foxx, secretary of the U.S. Department of Transportation, went on the record in 2015 with the claim that "we're going to see [fully autonomous cars] within five years," though he allows that that "just means market availability." A more comprehensive timeline assembled by Recode suggests that by 2030, "Automakers will stop manufacturing cars that don't have at least some highly autonomous features." It goes on to predict that by the middle of the 21st century, we'll witness total fleet turnover, at which point virtually all vehicles on the road will be at least partially autonomous. If that's true, it's possible that driving your own car will rapidly come to be seen as a dangerous affectation like smoking