AI: Point. Counterpoint.

This is a picture of HAL9000.  This AI went crazy in a movie.  Surely that won't happen in real life... right?
Nothing to fear, it’s just a machine.

Point: AI Will Save Us All

Here’s the thing about AI, it learns exponentially. Right now AI is pretty task driven (detecting faces, helping drive cars, etc) but the best AI teaches itself and really there is no limit to what AI can do.

So think about that… If we can build an AI that can start solving our problems, there is no limit to how much better it can make our lives. Right now IBM’s Watson is helping people try new ways of cooking and it’s giving us recipes that sound weird, but taste delicious (Mushroom and Strawberry Kabobs for instance) but if we focus it towards problems like climate change – who knows what it will come up with.

When cars can truly drive themselves and factories can be controlled by a really smart computer, we will be able to simply relax and chill. We can relax while we watch TV in our AI designed ultra-high efficiency houses or sit outside in our AI climate controlled environment and drink AI created cocktails.

Anything we can do, AI can eventually do better and it can likely make our lives easier because of it. AI can eventually do things that we will only recognize as magic and that magic is here to help our lives get better. AI will be able to solve problems we don’t know or understand and will be able to protect us at a level we’ve never had before. That is why AI Will Save Us All.

Counterpoint: AI Will Destroy Us All

Here’s the thing about AI, it learns exponentially. Right now AI is pretty task driven (detecting faces, helping drive cars, etc) but the best AI teaches itself and really there is no limit to what AI can do.

In case you didn’t really read that, THERE IS NO LIMIT TO WHAT AI CAN DO. Once you give AI a task or problem to solve, it will literally do anything it can to solve it. Sure we can put restraints on it, but we’re talking about something that is smarter than us, can think faster than us and we can’t really understand.

AI learns exponentially. We don’t. This means that there will be a period of a few days (or maybe hours) between the time that AI is as smart as the dumbest person alive and then as smart as the smartest person alive. After that, we just have to sit here and hope for the best.

Domesticating wolves took thousands of years and there were more than a few people bitten along the way. If we’re hoping to domesticate AI, we better get moving and slapping AI in the nose with a rolled up newspaper probably isn’t going to be as effective as it once was for dogs. When AI eclipses our natural abilities, we just have to hold on and hope for the best. The majority of the problems in the world are caused by people, and unless we can figure out how to completely constrain how AI works, we have to live with the realization that from a computer’s point of view the easiest way to solve most problems is to get rid (or severely limit) the people that caused them. And that’s why AI Will Destroy Us All.