Is your AI even cool if it doesn't "accidentally" hack another company? Also, curious how will they put the AI in prison like early hackers that accessed systems without permission?