{"id":41618,"date":"2017-12-01T13:07:47","date_gmt":"2017-12-01T21:07:47","guid":{"rendered":"http:\/\/world.edu\/?p=41618"},"modified":"2017-12-01T13:07:47","modified_gmt":"2017-12-01T21:07:47","slug":"teaching-machines-teach","status":"publish","type":"post","link":"https:\/\/world.edu\/teaching-machines-teach\/","title":{"rendered":"Teaching machines to teach themselves"},"content":{"rendered":"

\"\"
\nAre you tired of telling machines what to do and what not to do? It\u2019s a large part of regular people\u2019s days \u2013 operating dishwashers, smartphones and cars. It\u2019s an even bigger part of life for researchers like me, working on artificial intelligence and machine learning.<\/p>\n

Much of this is even more boring than driving or talking to a virtual assistant. The most common way of teaching computers new skills \u2013 such as telling apart photos of dogs from ones of cats \u2013 involves a lot of human interaction or preparation. For instance, if a computer looks at a picture of a cat and labels it \u201cdog,\u201d we have to tell it that\u2019s wrong.<\/p>\n

But when that gets too cumbersome and tiring, it\u2019s time to build computers that can teach themselves, and retain what they learn. My research team and I have\u00a0taken a first step toward<\/a>\u00a0the sort of learning that people imagine the robots of the future will be capable of \u2013 learning by observation and experience, rather than needing to be directly told every little step of what to do. We expect future machines to be as smart as we are, so they\u2019ll need to be able to learn like we do.
\n