Yeah, yeah. I can give you probably three quick answers. So one would be one of the biggest struggles people have, and I can super relate, is you're like, what is the hype? This thing writes terribly. So if you're struggling to get AI to write, well, you're like, What do I do? Because it also can't focus. It doesn't have a memory. It can't remember that you've trained it yesterday on everything about your organization. So how do we overcome that? And that's why it's really important to build funder context, who you're pursuing the grant money from, and fundee context, you the organization, and the fundee is going to get you 80% of the way to a really good draft the funder is going to be final 20% that gets you a great draft, but it takes 80% of the effort. So in those two pieces, when fit together, will get you a great writing assistant, but most people don't know how to do that, and so they get frustrated because you expect AI to remember what you thought you told it already, right? So that would be probably one of the biggest, biggest mistakes I see people have. And then there's their frustration. Another example would be AI can't do novel and the thing about grants is that often what's most funded are new and novel concepts. So if you're a, I don't know, like a nonprofit leading the way in some form of science innovation, and that science has never been done before, there's nothing that AI can read and help like rework, because it's trained on information in the past that's publicly available. So in there's a variety of like National Science Foundation grants or even SBIR Small Business Innovation Research grants. Like, really you're better off doing it healthcare innovation grants. You are better off doing it the old school way, writing by hand, than wasting your time trying to get AI to do it for you. Also were learned through hard experience, right? Like, learned this trying to write a grant for a really innovative healthcare company in Europe was how we learned that, like, literally, it was a disaster. Had to write it by hand, is the only way. So that would be another one. I'm sure I could give you more examples, but that's like a good starting point in terms of, like, what it can and can't do. I mean, I think the biggest bit of what it can do is it is your writing assistant. It is your buddy, it's your thought partner. It can help you and with citations, like it can help you with research. But the oh, this would be the other major mistake I wanted to share, which is, you think one tool can do it all. And I think 90% of people, when you think of AI, you think of AI, you think of ChatGPT, because it was certainly like the loudest, you know, largest tool we think of. But the reality is, it will make up citations. It will make up data that does not exist. So that's why we love Perplexity, because Perplexity is like far better than Google search, and it will give you really, really good data. So now I need to learn, how do I plug those two together, so my writing assistant can get good data and blend those into like, into being an actual helpful thought partner. So it's sort of like thinking about, how do I employ multiple assistants that are specialists, versus thinking you're going to have one generalist?