Podcast Episode 2: Stop the Slop: How We're Keeping AI Out of Our Blind Spots | Ep 221

Description

AI slop isn't a writing problem; it's an effort problem that shows up when we outsource our thinking to the model.On this episode, the team walks through what AI slop actually is, how we're keeping it out of own channels, and practical tools like a viral AI Writing Policy and skills our team uses internally to help you stop the slop.Jon kicks off with version updates including v19.4 in beta with AI agents coming very soon, Rock IQ, and v20 development. He shares his hot take on AI slop framing it as a human problem rather than a technology problem.Nicole walks through the Write-Like-Me skill, which is an internal tool used to speed up drafting communications in your real voice while stripping generic AI output. It's refinable over time and we're giving it away for free.Emily brings in Extreme Ownership, applying Jocko Willink's stewardship principle to Rock admins. There are no bad teams, only bad leaders, and there is no AI slop problem, only a review process nobody is enforcing. Download the resources and listen to this episode to learn how you can apply them.Visit the show notes to download all the resources talked about in this episode. Don't forget to join the new Rock Cast Rocket Chat Channel to see what other churches are saying about this episode.

Transcribed Content

Welcome back to ROTCast. I think today's episode is such a good one that everyone needs to hear if you use AI to any capacity, whether that's chat or Gentic. And today, we're focusing on how to stop the swap. Alright. I know I'm guilty of it, but the tech giants are sick of it in recent news. And we're gonna be sharing how one company's policy that they rolled out for the engineers was so good, they distribute it to their whole team to use internally. And then we're gonna share about a skill that our team created to keep the slop out of our own channels. And then lastly, we're gonna wrap up with a powerful mindset that is gonna make all this stick together. So, John, can you get us going with some version updates? Yeah. We'll stay on the AI theme. 19.4 is currently in beta. Although by the time you get this, it hopefully will be actually fully released. And just as a reminder, that is the biggest feature or biggest reason to get this update is the new AI features. So those have been inside of 19 but quietly hidden, and so they become unhidden in nineteen point four. And so I want to unpack a few more things about RockIQ just to help bring more clarity and more details into all of that. So, RockIQ, again, is the first service of Rock's new connected services. And so, again, that's a feature set that we've created so that you can get more advanced features in Rock without having to provision a lot of advanced stuff. More to come on that, but the first service is RockIQ. So inside of RockIQ, you'll sign up, you'll pay for your tokens through this service, and it links right to the same credit card you use for Rock Shop purchases. So that's really helpful and convenient. Now, you don't have to pay to use all the AI features. If you want to use the AI features in your own harness, Claude or OpenAI, you can still use that. MCP is free. , there's no there's no cost to AI within Rock. However, if you're gonna use tokens, there's cost to tokens. So we're just helping you pay for those tokens. But if you wanna keep it outside of the native Rock chat, you're fine. You can do that at no cost. So, again, MCP is free. There's even more features coming in 20, so, , don't don't worry about if you don't have the skill that you want. We have a lot of skills. You can do a lot of things, but version 20 will have twice as many skills. So most of those are administrative and some really cool ones. There's two plans that you can pick when you sign up for RockIQ. There's Standard, which we highly recommend, and then there's Economy, which is actually gonna use cheaper models. But when you think cheaper, think dumber. So Okay. It's it's true. , if you wanna save some money, go ahead and and try economy. I think you're not gonna the results, and don't judge the tool by the economy mode, please. But find that for those of you who are out there that, , you wanna, , save some shekels, there there you go. Now, what are the models that we're using behind the scenes? We're not really talking too much about that because they change all the time. So we're doing a lot of research, it's not just a onetime research. It's constant research. In fact, we're we're creating an agent that's gonna be out there helping us detect changes in new models and changes in pricing. So even a a model that's been around for a while, the pricing is changing all the time. Super interesting. OpenAI and Cloud are both doing price wars within the pricing within OpenRouter, which is the technology we use to handle all this routing. And they do that for some very interesting reasons. How they report their books to Wall Street is dependent upon what their usage is in the industry. And OpenRouter actually is one of the places that they go to to measure who's using what. And so they're discounting their models on Open Router to help us, encourage us to use those models, which makes them look better to Wall Street. So again, this is all more difficult than it seems. But we are picking our models based on quality. We want good models. Price. We want as cheap as we can get models. Speed. We don't wanna be waiting around for all your tool calls, and privacy. So those are the attributes that we're using. Now, every model we use will not be using your data to train. So that's off the table. Your data will not be used for training. However, some models, especially the frontier models that you'd want us to use, do have a policy or a rule that says, , some of the prompts will be used and logged and stored for a short amount of time not to do anything with other than to do some guardrails, in case they have to go back and adjust guardrails. It's really just risk mitigation. So they're just trying to say, hey. If we get in trouble, we need to have these logs. There is something called out there, if you're if you're familiar with all these technology terms, there's something called ZDR, which is zero data retention, which means, hey. Our models won't even do that. , absolute zero data retention. The problem is that sounds good, but the best frontier models don't support that. And if you're really into this, you might say, but wait, there is a way to do that. Yes. If you wanna use some of the frontier models with ZDR, you can go through, , Azure to do it because they have a zero data retention. And you can still get those same OpenAI and Claude models through that. However, you're trading their reliability. In fact, in our research on this, Azure's uptime on those models is about 80%. Oh. Yes. That means the downtime is 20%. Right. And if you're familiar with uptimes, you wanna see 99.999. , 80% is Ouch. Horrific. And we actually had that turned on, and we were actually seeing, , our AI just not responding. , we'd be typing and nothing was happening. And we're , oh, no. We have a bug. And then we went in and looked at the uptime and was , oh, that's it. And as soon as we switched off that and it went back to using OpenAI as a as a true source of the model hosting, everything was fine. Oh, interesting. So, yeah, ZDR is great, but, , right now, it's if you want the best models, it's it's not an option. And again, they're not training on this data. They're just logging in. And what I would say is just everything else in the cloud, , your email is still being logged. You're , everything is being logged. So, , to to use a different metric on this or different level kinda doesn't make sense. PII is all over your cloud. So you you have to know you have to trust the cloud and what their their terms and conditions say. And and we're following the terms and conditions. We're not just, , going blindly and saying trust them. We're just saying, okay. This is a person they're not gonna do, and so we have to trust that. You have to do it even with ZDR. , you don't have access to data center. You can't tell if they're truly doing ZDR. So at the end the day, it's a trust. Read the read the terms and conditions, and then do you trust what they're saying in the terms and conditions? So, anyways, that's a little bit of an update of how we're picking those models, but we're constantly tuning them because they're constantly being changed. Switching gears a little bit. V 20 is coming along. So there's some really big features in v 20. Here's what I'd say. It's you need to come to RX to hear about these features. Not only you gonna hear about them there, but we're also gonna be talking about them in in greater detail that you're you're probably not gonna get anywhere else. So I would say, and I have been saying it to lots of people, our product road map for the future in v version '20 and even more so in '21, which we're gonna be talking about at the conference, is unlike it's ever been. , I'm more excited about the roadmap now than I ever have been. And I've been excited about the roadmap at all times. So we have some really just great stuff coming. It's exciting. And for those who haven't been to the conference before, they may not realize just how critical that experience is to their entire Rock experience all year. It it opens up the next level of Rock, really, by understanding the vision, plugging in with peer churches that are doing incredible things, collaborating and working together so that everyone's not reinventing the wheel in their own sphere. Mhmm. It's just a really critical experience. Yeah. And I think for those people who might say, well, let's I can just watch it afterwards. I would say a few things. One is it's not the same. Mhmm. You get to ask questions. You get to engage. And then you're really missing out on what happens in between those sessions. There's this constant hallway conversations, not only with the Spark team, but with other people in in from other churches. It's not even close. Mhmm. I'd also say for most people who say that, the amount of content that they actually watch is , there's the intend to watch and actually watch. , and that's not that's not that's all of us. Right? That's me too. I intend to watch every session afterwards. Yep. But the reality is you get back, your calendar's full, and you're going. This intentionally forces you to stop. Mhmm. Connect. Go deep. The content subscription, I think, is best used as, , your follow-up notes and your refresher of the things that you want to remember and grab all the details on because it's really hard to get that in the moment. So that's kind of the way I use that. Yeah. And we have the metrics to show it to. , I go through and see, okay, who's watching what sessions? It helps us know from a content perspective. But again, I think it dramatically shows intent versus, , follow through. And life is just difficult. You get back to your office and you have a million priorities. Yep. I think that in the conversations that I hear from people about Rx and in the follow-up messaging and stuff, I think that those hallway talks are one of the biggest takeaways for people. , we, , do our absolute best to deliver excellent content from the stage. But just those interactions they get with the churches, , just shoulder to shoulder, I think is one of the most impactful things they walk away with. And we also have a content subscription waitlist. So Oh, wow. If you want be notified when it gets released, you can sign up for that. We'll probably leave that in the show notes. Yeah. The hotels think we're crazy because, , literally We do. There's, , hundreds of people milling about till late in the night. , sometimes 2AM, the conversations are still going. But even till 11AM, there's hundreds of people just all talking. And, , you're just not gonna get that through a content subscription. No. You sure aren't. And you don't even realize how many hundreds of friends or, , partners in arms you have out there in the trenches of doing ministry through technology until you come to the conference. And they're just such a great group of minded people doing really incredible things. Yeah. That's so true. So keeping going on the topic of Oh, yes. Let's keep going. Let's talk about sloppy AI. What? There's sloppy AI out there? No. That's what the big word is. Everybody's using the word AI slop all over the place, which, , is true. However, I think it's the wrong word. I feel when we say AI slop, it's really human slop. We're just kind of putting the AI name on it, right? AI cannot have slop because AI doesn't take what you do and just give it to somebody else as the input. You tend to do that, or you give it permission to do that. But if you don't go talk to AI and tell it to go do something, it won't do it. And thank goodness our token budgets would go way out of whack if it just decided to do whatever it wanted. So, yeah, AI slop is just really human slop. And I think the first step in understanding a problem is admitting that it's you're the problem. And we all need to do that, , all of us. , we're all guilty of sending something that was, , we just So, pitched it over the , thinking through this, I love what Cal Newport said in Deep Work, that writing is a form of cognitive fitness. And I think that term cognitive fitness, it was kind of that phrase was new to me, researching for this. I love that concept. We often think of fitness in our physical being, going to the gym. Lots of people go to the gym. In fact, that's a big thing right now. But mental fitness, cognitive fitness, I don't see a lot of people, , working on their cognitive fitness. In fact, I think it's actually regressing holistically to the culture. But but yeah. So I I love that. And constantly, we're context switching right now, especially with AI. , the context switching within AI because it's it tends to be a little slower. So while you have that response going I usually have, , five AIs going at once, and the contents switching between that is painful at some It's exhausting. Yep. I go home more tired than I have ever been. And so that context switching sometimes gets us in a mental model of not putting cognitive deep thought into things. And so what I would say about Kal Newport, I love that he says that writing is a form of cognitive fitness, but I'd push back a teeny bit and say that's probably too narrow. I didn't think it's writing in of itself is that I just think intentional thinking. Writing is one way of having intentional thinking, but I don't think it's necessarily the only way. But I think holistically, we're just not spending time and just thinking. And so that kind of leads into, , the outcomes of slop. But I think writing is shifting because I do a lot of writing or have always done a lot of writing. I mean, in the early days, all the documentation I wrote. So and I've written articles, , for my whole career, even all the way back into when I worked just at the church. But I think it's shifting. I think it's changing a lot. And in some ways, that some might be somewhat controversial, but I think more time is needs to be going into the formation of thought and and your ideas and the creation of the outline, and less time is gonna be on the exact word choice. Obviously, we need to own the exact word choice. But the selection of the exact words and the order of those words is becoming less important. But more important now is the focus on the critical thinking, the deep thought that goes into that. And so I would tell people, and again, this is a controversial side, , it's okay to have AI do the writing. It's not okay for it to do any of the thinking. So if you take some of the stuff I've written lately and you put it through an AI detector, it's gonna say 100. Yeah, it's AI. But 0% of that thinking was. And the wrestling of that idea, , the actual last draft of it, I'm reading through it a course 10 times, sometimes a 100 times of longer or bigger articles, but I'm wrestling with those ideas for And one of them, did it for months. So And that's okay. I think it's okay that your AI Your text can be put into AI and says, yeah, this is somewhat AI generated. That doesn't mean every word was. , sometimes you're going there wordsmithing something. I I don't quite that. That's still too strong or a little too weak. So we but we really need to put the more thought into less about the focus on the exact words and more on the focus of the exact thoughts, and don't let it outsource that. So, , I I think it kinda depends on how big was your prompt. , my prompts tend to be, , you can't , you have to scroll up and down to see one prompt. And I love the the feeling of just writing that prompt because that's where the thinking is happening, and I'm editing that prompt before I even send it to the first time. And it it's just you're , I think some people are talking about, , it's a stream of thought. You're just thinking and then, , keep thinking, keep thinking, which causes other thoughts, and you just keep going. And it's you could never send that to somebody without seeming you're insane. But that stream of thought is really important because it's you're getting all your thoughts out. But I love the , talked about there's the folks at Clay worked on, , their anti SLOP kind of rules, and I think they're very practical. And somewhat, I I had I had changed some of these and added to them a little bit, so it's not ex exclusive of what they said. I think the first thing to think about is the a a time rule of thumb that you should spend between three to 10 x the time writing it than it is to read it. And I think it depends. , if it's just a quick email, it's probably, , three times. You should spend three times as much writing it as you do the person spends reading it. But if it's a longer article, it's easily 10x. Might even be 100x. If it's an important thought provoking article, if it's a quick article or an update, maybe 10x. But also too, this is the part I love, that they say you have to have ownership of every line of every word, and you have to stand behind it on the spot. , you cannot say, oh, I'm sorry. Yeah. AI wrote that part. It's you cannot say that. You own it. If you sent it, you own it. You can change your mind, but you can't blame AI for that. Be thinking about the fact that you think about the math. One person's creating the message, but 10 to 100 to 1,000 people are consuming that content. That's right. So the efficiencies of if you spend more time making your your thoughts more concise and logical and right, correct, you're saving a lot of people a lot of time. I think a lot of times when we do AI slop, read human slop, we're we're choosing to be lazy and then forcing a whole bunch of people to have to work extra. Mhmm. I think if we keep that in our minds, , we should have empathy for that. Length is a virtue. I think with AI in the beginning, we're all , oh, cool, look, I just typed this, and I got all of this. Well, that's actually a bad thing. Yeah. , you have to make it as concise and readable as possible, and AI just makes it very easy to pad it. And I think I think you can kind of see the maturity of someone's use of AI through this. New people, you see it. You get long things. Things are bolded that shouldn't be bolded. You see the em dash, which who, , kind of at a certain point, who cares? But to me, when I see an em dash, it's , okay. Well, someone obviously doesn't know that take it out just because Yeah. No one wants to see that. I feel so have people who actually used Emdashes. I didn't know anybody who used an Emdash before, but Yep. They're all in trouble now. But you can just read it. It's , oh, thank you for the AI slop. Let me put that through my AI so I can actually understand your slop. Mhmm. But I also think transparency is fine. It's okay to say this is sourced by AI. There's nothing wrong with that. We don't we shouldn't feel ashamed of that as long as we were the ones who are guiding it, giving it the principles, giving it the thought. So, , I heard this on a podcast, that's that's super helpful. Yeah. It definitely is helpful. Yeah. So I'm sorry. Emily, can I throw you a curveball? Okay. Alright. So if a staff member handed you a heavily AI written thing that was, , borderline sloppy, how would you handle that? Or how would you, , help them improve that writing? That's a good question. , I think it you shouldn't even focus on the AI at all. You should just focus on , does this sound you? And a lot of times people that might send you something, I think, John, it's what you mentioned, they might not be as mature in their use of AI and they may not realize that what they're putting out there sounds generic or has these telling patterns and things that you see in a lot of AI generated content. So they aren't sounding themselves. So I would focus on , this doesn't sound you. Here's how we can help that. Yeah. That's The a great other day as someone sent me a lot of thoughts And we're a community, so I always wanna get back to as many people as I can. I can't get everybody, but I felt ignoring this would not be right. But I had a choice. I mean, at this point, it's probably 09:00 at night I'm looking at this. I'm , Well, I can either choose to ignore this or I can choose to respond, but I'm gonna have to use AI to help me because I do not have time to write something that would be good enough or complete enough to what they had sent me. So again, I just typed in for fifteen, twenty minutes, stream of thought, said go, and then I had to wordsmith it. Mhmm. But the choice for me was don't use AI and don't respond or respond and use AI to help me in the productivity of that. So, , what is the more relational thing to do? Some people say, well, it's not a high relationship to use AI. Well Okay. So then they get nothing. Yeah. But instead they got something, and it goes to my true thoughts, and it was wordsmith. And I have a whole bunch of just two or three letter phrases I add, because sometimes I wanna adjust the tone to say kind but firm. I love the word kind, Especially when telling AI, it's not nice. Nice is , Walk all over me. Kind is , Well, you can't walk all over me, but, , I want to be kind and Christian, but yet firm in the thought. In that case, it wasn't firm. It was more just understanding. That's an interesting perspective for using AI. Yeah. I mean, because people say that it's not relational. Okay. Well, then you get nothing. Yeah. Would you rather have my thoughts that's not my exact words? Mean, I hopefully, that led into that was an opportunity to lead into another conversation or something. Yeah. I mean, I felt kind of bad about it. I was , well, this isn't very relational, but it's , better than nothing. Did you say AI sourced at the bottom of your message? No, but it probably should. But in that context, it wouldn't have made sense. Mhmm. Okay. Well, thanks for walking us through that. I think it's gonna be a great resource resource for our churches, And we'll have that in the show notes. So to our listeners, let us know if you do end up using it or adapting it within your own church. I would love to turn us to a skill that we've created on our team to help us keep the slop out of our channels. I know I do use this quite frequently. I use it all time. Yeah. It's been very helpful. It's called the Write Me skill, and it's designed to draft communication in the author's real voice, polish it against quality standards so that it strips out all the generic AI output while maintaining the integrity of the writer. Very cool. It's comprised of two skills. The first is a Build My Style Guide. So when you go to set it up in your co workspace, it'll ask you a series of questions in an interview format. So some of them are , what's the message that you wrote to your team member to follow-up on something? Or it'll ask for a sample of something that you wrote that was for, telling a church or a client or someone about bad news that you have to push back a deadline. The key to this is putting in things that were not written by AI. , so we're not giving AI more AI. We're giving it our actual written thoughts and our own style so that it can create a style guide. And I thought it was so good. I printed mine out so I could share some So with my style of writing is that I'm, my warm beat is specific, not generic. So I ask , Hey, I haven't heard from you. How's it going? And then another good one is that I try to reframe disagreement rather than defend it. And a lot of these times I'll continue the conversation with a question. So another one is that I lead with context before asking something. So tell somebody, Hey, I'm out of office. Do you think we can prioritize this? Something that. The second part of this skill is the company style guide. So we have created a guide that further helps us take out the AI. So we have, , no embashes or no overly used AI words. Feel streamlining is one of them or utilizing. This is one I love. And here's why that matters. Yes. The first time I read that, I'm , oh, I that. But then you start seeing it everywhere. It's , okay, now I feel I'm being talked down to. So whenever I hear that, I feel I'm being talked down to. Or that pattern that says, not this, but this. Oh, yeah. There's just a lot of different patterns. Yeah. And you can collect them. We're constantly adding to that company wide one. Yeah. Oh, here's a new pattern. Let's get rid of that one. Yeah. So, what's the best part about this skill is that you can continue to adjust it and refine it over time. So you can feed it more writing and it'll pick up on more of your nuances and to make it sound better just you. But I think it's important to note that, , this is the difference between publishing something that sounds Rock and publishing something that any anybody could have written. And it's not that we're trying to take out the human judgment part. You absolutely need that. Okay? This isn't an excuse to skip over that. But it does make the drafting the communications easier. But yeah, you still have to have the final judgment call at the end. I think it's a two parts. what I was talking about is the thinking that goes into it. And then you're talking about how do you actually style it into your voice. And so while we say the company, it's really the person's voice. The company part is , Hey, just take out the Em dashes, take out the We're definitely not saying write the company. And that's why I the interview, which is what created that. Those are tailored to you. Mine would be very different. But I find it works really, really well. Yeah, I that it's improvable over time too, because you can get so much out of, , a multistep series of prompts. But over time, if you're , we're repeating this too often, or this is maybe more of an edge case for me, you can keep feeding things in and it just keeps getting better. Yeah, and inside of Cloud, you can just publish this skill to our marketplace. we have a Triumph marketplace of skills, which is basically just backed by a GitHub repo. So as soon as we push it into the GitHub repo, one person pushes an update, it just gets deployed to everybody. So it's super easy to keep all going. So you have the anti AI words and phrases, and then you have the personalized to me, as you were saying, John, section that says, I'm taking out the things that sound AI, and now I'm making it sound me. Yeah. And news this week too, and another big part of the news was Anthropic is talking about how they're watermarking the text outputs, and it's unpreventable. And I kind of heard them , Well, first of all, I'm not paying you to do this. And this is all from EU rules and stuff, which is unfortunate. But first of all, it came made me a little frustrated. I'm , I'm not paying you to watermark my stuff. But at the end another point is , I don't really care. I don't care that someone knows it's it was the words the exact words were AI. I it was all my thoughts. I reviewed every word, and I wordsmithed a few that I needed to do. But that makes me so much more productive and I can reach more people with more thoughts. I don't know. Feel as a communication specialist, I feel this should I should be upset by this. That's all going down to , do you want higher productivity and fewer exact word choices or exact word choices and lower productivity? I think that's the trade off. And I don't think there's a right or wrong. I think, , Seth Godin, I would hope he doesn't use AI to write his books, but I'm actually paying for that. And, , I think the economics work differently. But if you're one person and all your stuff is free, , and if everybody in the world said, hey, I only want one article or one thought or one I only want you to reply to one person a week. I'll be , okay. I can do that. If you want everybody wants a reply or not. And again, I can't do everybody. So if you've ever teams me or I'm sorry, rocket chatted me and I didn't get back to you, sorry. But I think you'd be surprised how many people get a response. Mhmm. Mhmm. I don't know if you guys are have any kids that are in college, but my boyfriend is going through, , GCU's statistics class. And it's so interesting to hear how they use , prompt the students to use AI. Be , put this prompt into AI, tell us what you got, and then break it down for us. Which is interesting. , they're they're really trying to help you hone those skills. That's good. I think critical thinking is the most important skill on the planet right now. And our schools generally don't teach that. It's good to see that that's, at least in the college level, that's changing. Yeah. Yeah. All right. Well, so we talked about how to stop the slop with this skill that you can use, and we will put that in our show notes. So you guys will have another resource in your tool belt for that. But Emily, can you zoom out for a minute and share, , about the mindset behind this all that applies so much more to just to then just AI? Yeah. I mean, I think you're right. I think we're talking about mindset here, right? We're talking about critical thinking. We're talking about making something sound us. All of this is kind of related to whether or not you take ownership of your output. So there's a great book called Extreme Ownership, How US Navy SEALs Lead and Win by Jaco Willink. And one of the central stories in his book is pretty dramatic. So I wanna tell it, and then I I'll tell you how I think it relates to everything we're talking about here and to Rock Admins. So they were in Iraq. He was leading a team commanding a SEAL task unit, and they were in some really up close, very intense battle. And they were fighting, and this incident happened. And, under friendly fire, they took out and and killed a soldier on a friendly force, and one of the people on their own command was seriously injured. And this is a major fail. Right? , this is the worst that can happen. It's probably worse than having enemy fire. ? This is what should not happen. So big fail. You have to go back and figure out what happened in order to prevent it in the future. And during the debrief, several of his men started stepping up and saying , oh, this was my fault. I should have , I saw movement over there. I realized it was our people. I should have said something. Someone else is , no. I was in charge of watching, , the the places where everyone was supposed to be. I should have helped with that. And they were all kind of taking responsibility. But Jocko was the commander. He stood up, and he said, no. There's only one person responsible here. That's the commander. That's me because I'm in charge of making sure we have the right preparation, the right training, and even the right communication, , in the midst of battle. So he owned it to himself, his team, and his superiors. So that's, , that's a big, big story if you think about it. And it can be , well, okay. That's cool. But how does that apply to Rock admins inside churches? , what does that mean for us? And I would say, , it's very easy to look at the overall structure of your church and what you're in charge of and what you're not and think , I don't really control anything. , I'll just go along with the flow, do what I need to do, but I can't affect change. So therefore I don't have ownership over an area. But that doesn't really mean that's not really accurate because in any situation, don't own everything is different than I don't own anything. And sometimes around here, we to say instead of ownership, we say stewardship. Ownership kind of feels a little grabby. Stewardship feels responsibility. So when I'm saying ownership in the context of this book, I'm thinking stewardship. I'm thinking responsibility to to provide for a group, not necessarily grabby its mind. Right? So and I think this does apply to Rock admins. You're stewarding quite a few things that are very important to the whole. So don't wait to be told to fix it. You might have to wait to get approval for, , a system overhaul from your leadership because of the budget, but you don't have to be told by anyone if your data view and report structure inside Rock is falling apart. , there are tools right inside of it to tell you the last time something was run, looking for duplicates, maintaining your systems in that way. That's something that's in your purview and in your lane. , maybe take a look at something the attributes in your system. They compound over time. If they're used incorrectly, they can impact performance. So these may not sound big, huge things, but keeping clean, well organized data is actually the foundation for personalized digital ministry, which is something we talk about a lot. And that's totally in your lane. Staff training, how do you roll out updates? I bet you could have a lot of impact on that, a lot of intentionality behind it, and the kinds of things that help level up your whole organization even if you don't own staff training over everything. And so those are some super detailed ways, , that you can actually take ownership or steward your areas of focus. But there's one other story that I think is also applicable, and it's called decentralized command. So this is also out of the book Extreme Ownership. And, Jocko, who's the author of that book and was the commander, has an argument that that no leader of any complex or large organization can be in charge of every minute decision. Right? , there's just no way the senior pastor can make a decision about, , your email carrier. , these things are not possible. So in a good organization, the job of leadership is to push the decision to the person who's closest to the actual system itself because they understand it. The key here is that that person also has to understand the overall mission of the organization. And this is again where I think, , this is super applicable to all of us because understanding the mission of the organization helps us make good decisions. So we can apply the systems thinking that we're talking about, a structured way of approaching the systems that support our whole organization. And instead of just taking, for instance, orders from ministry teams that come in and say, I need this fix right here, pausing for a minute, thinking about the impact over all ministries, the impact over time, and creating sustainable solutions, that's the job of the Rock admin. That's not the job of the senior pastor. So that decentralized command, I think, also really applies to the idea and the mentality of, , what's in my purview? I wanna do the best I can to steward these resources for the good of everyone. And these decisions are right in your lap. Right? So leaders don't give you a play by play, and your job is to try and understand what it is they're trying to accomplish at the strategic level they operate in, and then try to support that in the areas that are in, , in your lane. And I would say here's the key principle. Look for areas where you have some autonomy and then make sure your decisions are aligned. So it's where that autonomy and alignment intersect that you can have the biggest impact right where you are today no matter what your position is. And if you fight the good fight on the fronts we're talking here, , fix the things you can, make aligned decisions in your lane, Then when the major initiatives come from leadership on big direction changes or or system overhauls, you're ready to move really fast. So you're making good long term decisions, and you have good short term, , just the benefit of clean functional systems and lower risk. I think some people say, Well, I don't have a seat at the table, I don't even know what the priorities are. And I'd push back and say, You don't have a staff meeting? And if you don't have a staff meeting, you don't have a sermon every Saturday. And you don't have a state of the church. So you can see what matters to the leader, align yourself with that, and you're naturally , I don't know a leader who goes , Yeah, I know you're trying to help me exactly what I'm trying to achieve, get away. It's more when you come and say, Hey, I know what you wanna do, and let me push my own motivation into that. That's You might have a problem. But if you get an alignment, I mean, I think that's the story of my beginning at CCV is this trying to figure out alignment. And once I had alignment, it was easy to push things. Yep. Because I wasn't pushing my own thing, I was pushing their thing. And I just got on board. Yeah. So you can apply extreme ownership right where you are. Stand up. Take responsibility for the things that are in your lane. And when you see those things, realize you have some autonomy there to improve and make things better. Just make sure you're aligned in the decisions you're making. Mhmm. Yeah. That's great. I love that. One of the, , takeaways you talk about in that book is when the leaders take ownership. So when Jaco basically stepped in and was , actually, this is my responsibility. He ended up earning the respect of everybody around him. And it was an effect that rippled down throughout every layer of their organization down to, , the junior role. It's something that impacts everybody just by standing up and saying, Hey, that wasn't the best move. Here's how we'll fix this. Yeah. Yeah. It's pretty refreshing to hear that story and how everybody was taking their own responsibility. And we also talk about task ownership versus outcome ownership. And hearing that was very insightful for me because, know, we don't just want to check off boxes and call it good. Know, we do want to be responsible for the overall outcome and how it does align with the organization or push the mission forward. So, ownership here isn't about blame. It's about deciding you're the one responsible for the outcome. Yep. Okay, to wrap up this episode, I just want to close out with some announcements that our second hotel block is 70% full. So we'll actually, we'll be pulling it off our website soon and putting up our last and smallest hotel block, which means that it's gonna fill very fast. So if you're from out of town, if you're on our East Coast flying in, make sure to get your hotel rooms booked. And we didn't save the best for last. Right? So the last one is probably the one that we the least. Right? So get into this one. Yeah. Don't wait. Don't wait. Just because it's the third one. That's the one you don't wanna be. I'm in the third choice. That's it. One sounds great. I mean, they have complimentary breakfast. There you go. The Westin does. Yes. Yeah. Okay. So the next announcement is that Gold Circle Awards are due on September 18. That is less than a month away. It's coming up quickly. So if you haven't yet, submit your innovative digital ministry project so that we can review it. And then if you haven't yet checked the schedule and say the last week, you should give it another look because a lot has been added. Things are being changed. So give that a review and get ready for conference. We're so excited. I feel the buzz is picking up around here. Hopefully people feel that too. Okay. We'll see you guys on the next episode.