Why the Fear of AI Taking on Humans s a Nonsense

Lately you may hear a lot of talk about AI going wild and deciding to destroy humans and how dangerous that can be. This is such nonsense, and here is why.

• First of all, AI must raise a question about humans. Only then should it make a decision to do something with it.
• From what we currently have, our AI systems have never asked a question. With a current state of technology it just runs a piece of code we ask to be loaded into a computer memory, chunk by chunk, and performs an output according to it.
• Second, we know animals that can “talk”. Like parrots or Koko the gorilla. She can interact with humans. But even these animals have never asked a question. They answer/react to our questions only when asked, like in the case of a Koko or when you ask your dog to go outside and it gets excited.

Every time I hear that AI can harm a human, I wonder if AI is already capable of:

• Deviate from output and ask a question without direct input to do so, by its own “will”, bypassing blocks of code loaded into a memory.

Without it, I don’t see how it can harm anyone if a question can never be asked.