The danger isn’t artificial intelligence – it’s us

Our most immediate threat is from other humans hacking AI. Cybersecurity is a growing field and will only become more important as we continue to develop AI. Equally important for our protection is to have discussions about how AI is implemented in the first place. By limiting the power AI has, we can guard against misuse if it ever does happen. For example, an AI that can figure out the most efficient way to direct self-driving traffic around a city is great but it probably shouldn’t be in direct control of the vehicles itself. Instead, its output would be used by engineers. Otherwise a hacker could use the AI to cause vehicle collisions.Eric Horvitz, managing director of Microsoft’s main research lab at Redmond, thinks we’re safe from AI if we’re careful in our…


Link to Full Article: The danger isn’t artificial intelligence – it’s us