Pretty sure he has a guilty conscience. He wants to build an autonomous self driving car, which will have to get pretty close to general AI and he wants to put

Pretty sure he has a guilty conscience. He wants to build an autonomous self driving car, which will have to get pretty close to general AI and he wants to put it in multi-ton robots. He knows this could be dangerous and scary. At least he is giving money to research the problem rather than just forging ahead with the potentially dangerous part. Personally I think the robot uprising concern is at least 50 years premature. But the "what if someone could take the controls" concern is a concern we should have been already been considering yesterday.