How Have Your Opinions About AI Safety Changed?

Could we have AI-labeled training data for AI with some quality control mixed in? Could we set up this training set so that it wouldn't pollute the model and degrade performance but would instead perhaps expand performance, improve performance? I would bet that you could do interesting things here. It's well worth noticing that people in the 1950s were claiming that oh within a generation we're going to get human level AI and we're not concerned about safety at all. Now we have some examples of simple AIs. Have you understand how have your opinions changed about how we can make these models safe?"