Inverse reinforcement learning could be a useful technique for national governments and tech companies to set objectives, according to experts. Rather than manually designing objectives, experts suggest presenting real people with different scenarios to pick which would improve the world, in turn allowing issues to be addressed faster and more efficiently.