We’ve spent years feeding neural nets vast amounts of data, teaching them to think like human brains. They’re crazy-smart, but they have absolutely no common sense. What if we’ve been doing it all wrong?
With deep learning, an artificial intelligence system ingests massive amounts of data to identify patterns – but only for the data it has learned about. This article dives into several ways researchers are trying to build AI systems that can develop common-sense understanding and apply their knowledge to new contexts.
Read the full article at Wired.