Ah, I see your point is not philosophical. It's that we don't have historical data about the effect of AI. I understand your point now. I tend to be quite a bit more liberal and allow things to play out because I think many systems are too complex to predict. But I don't think that's a point that we'll settle here.