We all have heard about artificial intelligence. It is the AI that powers a lot of things around you already, whether it is facial recognition or your car identifying pedestrians close to it. Nonetheless, the problem or rather the limiting factor of artificial intelligence as of right now is that it cannot evolve. SciShow is exploring the idea of whether we can teach artificial intelligence to evolve by designing it like a human brain.
Artificial intelligence can perform a wide variety of intricate tasks but falls short when it comes to basic abilities of a toddler such as recognizing different faces. The problem is known as Moravec’s paradox. It has been named after the scientists who first articulated the problem.
Hans Moravec wrote, ‘it is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility.’
One might wonder that why is it so? Why can we teach a machine how to solve even the most complex problems but are stumped when it comes to teaching them the basic skills that a toddler enjoys? That is because the basic three-year-old abilities are a result of evolution whereas the AI takes its learning from the coding. What Moravec suggested was that we should apply the idea of evolution to artificial intelligence and have it mimic the human brain so that it is able to grow.
Will this approach actually work? Can we actually make artificial intelligence evolve? If yes, then where does it end? Will artificial intelligence evolve like Ultron and bring about the doom of the world? Maybe we are getting ahead of ourselves. We would leave you to check out the amazing video by SciShow, a YouTube channel, that discusses the principle in detail. Check it out and do let us know what you think of it!