In an experiment recently, Facebook chatbots created their own language and had to be shut down. Isn’t that a dangerous sign that we are dealing with something we don’t fully understand yet? Andrew Ng responds: “In the early days, electricity was seen as dangerous. People did PR stunts to show electricity could electrocute an elephant.This kind of fear mongering is completely unnecessary. Ultimately, like electricity, AI will be beneficial to the human society. And electrocution of an elephant hardly ever happens today. One thing that AI can do is that a small team of AI engineers can do such work today which earlier even thousand engineers could not have done.The thing I worry about is that AI software will be in direct competition with a lot of people for jobs.”
Share on Facebook