TED Talk "Don't Fear the Super Intelligent"; Does the potential benefit of AI outweigh the risks?

TED Talk "Don't Fear the Super Intelligent"

https://youtu.be/z0HsPBKfhoI

I absolutely believe that the potential benefit of AI outweighs the potential risks. As technology and specifically artificial intelligence advances, we, as humans, will always have the fear that our machines will one day outsmart us and topple our society in some way, shape or form. Hollywood and the media continues "stoke the fire" of these fears through movies like "Terminator" and "iRobot". This feeds the worry that one day in the future, humans will be enslaved or destroyed by their own creations. As Grady Booch had mentioned, AI will eventually learn human values. I agree with this statement. Instead of being at constant war with our own creations in the future, I believe that AI will coexist with us. AI will be able to help automate certain jobs with their faster processing power. This may make certain human jobs obsolete, but it will create an entirely new industry of jobs. After all, someone will most likely have to maintain and trouble-shoot our new bionic friends so they can continue to function at optimum capacity. With the invention of the Roomba, the "lawn mowing" Roomba and others, we're moving toward a space in time where normal household chores can be outsourced by machines. I put little googly eyes and a tail on my vacuuming robot to give it more of a pet like appearance. AI already exists, both in machine and human appearance. People can already order (to say in an appropriate way without getting flagged) AI "Love companions" that respond to conversation based on the personality selected by the users. These AI already have the ability of semi-autonomous mobility (can walk on their own, but need to be led by the user.) In my opinion, the way the mouths move and the way they move in general are not very close to humans and it's almost a little creepy, but overtime, I believe they'll eventually be almost indistinguishable from humans. While, this may not be the best approach to using AI, I think it could be a step in the right direction. Eventually these AI could be used to help people with mental illnesses like depression. I'm not a psychiatrist, so I won't begin on speculating how it could work, but a mental image I have in mind would be like Baymax. Personally, I'd love to have a robot companion that I could play chess, shogi or pai sho with on a regular basis. Ah, one day. It might be little scary to think about how their advancements are going to continue, but overall, I think the utilization and integration of AI is actually going to boost civilization, not destroy it (Assuming we can figure out how to make their movements a little less creepy and a little more natural.) I am incredibly excited for the future of AI and I truly believe that we should not fear it. They'll be designed to further the human race. Like Grady Booch also stated, worst case scenario, "we can always unplug them".