Think fast and slow to build a great AI product

First for product builders is the idea that you have to think fast and slow about your objective. And the objective of the algorithm really is really your most important job as a product leader. You should figure out what. Because that's the purpose of the product and. And it is the most important job. And it's actually the hardest job, in my opinion, because it has all these nuances of what you want to bring into it. And the reason I'm saying think fast and slow is I think you start fast with your kind of intuition of what would make it great. If somebody's spending time reading through, it must be a good experience, but we all know it's not. So then you apply the slow thinking to it, which is what could go wrong? Even if you're successful in, like, making your metric go up and to the right, what could unintentionally happen? And there's so many unintentional aspects. There's spammers, bad actors, kind of filter bubbles that could happen. So when you want to start to bring those into play, what you're really finding is you have a composite objective that has a purpose. It kind of builds the right guardrails to make sure it's a quality experience, at least for us, that's the way we think about it.