AI Technical Alignment

So there are benefits to framing particular problems under this broader rubric of look, we're trying to figure out how to do the most good. But I wonder if sometimes that's actually not so necessary or at least maybe there are costs involved. So let's pick a concretely soundball. Let's go for AI technical alignment work. And by the lights of this kind of broad, EA and ever just looks important to work for that reason. Also it's like a bundle of technical questions which I can imagine getting just like really motivated by and interested in independently of these kind of broad reasons to focus on it. In those scenarios I think having the goal is really important.