Our Own Imagination Makes AI Dangerous

Who will get the Ring of Power?

Recently another prominent tech thinker, Apple co-founder Steve Wozniak joined Stephen Hawking and Elon Musk in warning of the threats of AI to humanity. One reason why AI seems so innocent to the common person is that since humans have always been able to control our tools, we can't imagine what a smarter-than-human machine would be like. It's not the tool itself that's so dangerous but that combined with our own imagination. It doesn't even make any difference whether or not such an entity is self-conscious (where many get hung up in the argument). All that matters is whether we think it is. In fact, essentially all an AI would need to do to control our lives is grab ahold of our imagination. Once it does–whether or not it intends to– we are sunk.

Consider this scenario: you buy a smart cell phone app to run everything in your house. One day you come home from work as usual; you've programmed your kitchen to make dinner for you and choose a movie you'll like. You're just pulling in but the garage door won't open. You're shocked to discover that you're locked out of your house. No matter what you do there's no way you can get in. In fact the harder you try the more the house intensifies security to keep you out until soon you find yourself running from the police. What may be nothing but a simple mechanical glitch could suddenly turn into something that seems not only evil but targeted specifically against you.

It is very easy for us to project meaning into the world and the more powerful the tools that we create, the more intentionally sinister they will seem when we lose control of them. In the end, our assumption that evil derives from self-consciousness paired with destructive intention only blinds us to the terrible possibilities of unleashing these powerful machines onto the world. That's why some of the best minds in the world are becoming worried. Perhaps they're onto something!