I remember feeling it was just such an opportunity to think so differently about the way we construct those projects. And I remember feeling really uncomfortable and unsettling to us to try just being able to set it up that way. And I think that's one of these complicated things about the STEM piece of it, right? The STEM itself is supposed to be so data-driven and clear. And so then trying to put a bunch of people who have been trained to think in this very, data-driven way into a context that is itself so amorphous, I think the tension's really uncomfortable.