This weekend I watched a great video that explained why we should really be scared of AI, thanks in part to Mrs. Gergen. It was by a Robotics Professor at Brown University by the name of Peter Haas. Haas started off by asking the audience who is scared of AI. Many raised their hands and I definitely would have as well. Haas then continues to try and prove that AI can be scary. First he shares the story in which an AI is asked to identify what type of dog is shown in an image. The picture is of a husky that is in the snow. The AI identified the animal as a wolf and not even a type a dog. This was because it saw snow in the background and had bias written in code to assume that when there is snow the animal has to be a wolf. The researchers didn't even know why the AI made the decision to call the animal a wolf. This is what scared Haas the most. The AI researchers had no clue that this was happening. He even explains a situation in which an AI program will continue to do its job no matter the circumstances. Haas says that every program has a system that tells itself some form of the following quote. "The system must continue". Haas then goes back to tell us about a story when he was driving to Salt Lake City in Utah. He was driving in the rain when the rain turned to snow and eventually the snow becomes so thick that he can't even see outside the window and he gets scared that he might get his by a car.
Haas goes on to say that AI is like rain right now and will turn into snow and may even turn into a blizzard one day. If we don't keep checks and balances on things like this then we eventually be overrun by the vary robots we created. So I come back to the question I started with in the beginning. What is the real reason to be afraid of AI? The real reason is us. We should be scared of ourselves because we control out own destiny and our future.