Value-First Delivery - Jan 21, 2026

[00:00] **Introduction** Chris Carolan: Good afternoon, LinkedIn friends, Value First Nation. Welcome to uh a new series Value First delivery with Erin Wiggers. Uh, how we doing Erin?

[00:09] Erin Wiggers: Awesome. Yeah, I'm super excited for this.

[00:11] Chris Carolan: Me too. Me too. Uh, it's been since June, uh, is when we first talked about this. Um, and uh, now that I, I've learned a lot since then, as as you can imagine, um, but very recently, as I start to try and do more and more with AI and start to understand, um, why it's not as easy as I would like it to be sometimes, I've really started digging into all of the traps and how just all these status quoes that we talk about being problematic. Um, this managed services trap is showing up, uh, and it's not hard to like show like why it's such a problem. Um, and so I'm really excited to dig in. I I'd like to define it, uh, for folks first. So I'll pull it up on the screen. Uh, the manage services trap, uh, so this is the counter to Value First delivery, which we'll spend one episode on talking through this and then the rest of the series will be talking about how you avoid um, this trap. Uh, but it really comes down to, uh, you know, outsourcing has created dependency. Uh, what begins as a practical need for technical expertise, evolves into spiraling dependency on external specialists to manage systems that shouldn't be complex in the first place.

[02:53] Chris Carolan: And so this comes from, I think, a pretty clear like, you know, trajectory, uh, starting with uh, HubSpot, or not HubSpot. [inaudible] Chris Carolan: Salesforce. Erin Wiggers: Right. Chris Carolan: Um, introduced the lead object, uh, in 2000. And this allowed for a separation of, you know, leads and contacts or, uh, like the humans. Erin Wiggers: Mhm. Chris Carolan: So it started a dehumanizing process and now all of a sudden we need to get a bunch of leads. Well, how do we do that? We need to create, we need to create ads. We need to create conversion points. Uh, make it easy to get people's email addresses, even though there's a bunch of garbage leads coming in. Well, we've got thousands of leads now. Uh, and I think there's data to show like Salesforce is creating like millions of leads in their own, in their own pipeline. Well, we know it's not all good. We got to qualify them now. So let's bring in a whole SDR team, uh, to do that. Uh, that team uh, needs a place to work and we need to understand all this data. So we got to create these really complex systems, but we don't know how to do it. So, we've got to get consultants and agencies that can handle it all for us. And this whole manage service industry was created and has been thriving for, you know, 25 years now. Um, and the reason I want to call this out is because there's a ways to go.

[05:10] Chris Carolan: Like customers have not enjoyed this, but they haven't really had a lot of options to avoid it. Erin Wiggers: Right. Chris Carolan: And now AI is here and it will help them avoid it whenever they figure it out, right? So even if you think you can build a profitable business doing this, like like good, good luck. Um, that's the, that's all I'll say and there's a bunch of us that are going to be on the other side, making sure people know how to use AI uh, and to create the capabilities internally. Again, it comes back to uh, dependency. It just doesn't, you just don't have to have it anymore, yeah? Erin Wiggers: Sure.

[05:53] Erin Wiggers: Yeah, I mean, I totally agree and I think, you know, it kind of stemmed from this idea of, you know, it's, it's easier to upsell an existing client and then cheaper to upsell an existing client than it is to, uh, go, go get another one. And so people in agency world or, you know, in this managed service kind of space, decided that the way to do that and keep people in kind of a pool that you can upsell to, is, you know, just sell them a retainer to maintain the thing that you built. Um, and I think I just never really got into that as a concept and I really want to blame my ADHD for this one because it's kind of like once I've finished a project, I want to hand it off. Like, I want to hand it off to you and I want you to maintain it. That doesn't mean I don't want to do other cool stuff for you. And so like, the upsell opportunity is to build on what you've already created, not to just maintain it. Um, because then, you know, you end up with, with technical debt, um, because eventually the maintenance is going to hit a wall. Uh, you're going to realize that, you know, you've scaled as much as you can scale with this thing. And, you know, then the question becomes, do I want to go through this whole process again where I'm locked into a vendor for this, you know, extra long amount of time, just because I don't know how to maintain it myself. Um, and I think the the unlock here has really, especially with AI, is it allows people like me to do things that would otherwise probably get overlooked, like documentation and enablement and training and making the systems accessible to the people that they were built for. And, um, so I I think as we kind of move into more and more AI native companies and operators, that piece gets easier and easier. And so I think AI is kind of setting us up to to wean ourselves off of this dependency mindset, um, and, and get more into that value first delivery mindset.

[08:42] Chris Carolan: Yeah. No, that's exactly right. Um, but at the same time, like, AI will, we'll stick with the plan, uh, if, if you don't, uh, want to be enabled. And that's what, like, even with HubSpot, this has always been a challenge where if you're used to IT project work and Salesforces and WordPress, and like, you're used to being like hands off. I don't even want to try. It's like I'm not a developer. So HubSpot is just another one of those. But like it's not, right? And so doing everything in your power to just show people how easy it is to create properties, change properties without breaking the whole system, like it's really a mind switch, uh, when people see that for the first time and like now it's like I'm, I'm superpower mode as far as like that it only takes one of those moments with AI to realize like, oh, all of a sudden, I have no barrier to entry on doing this stuff. And guess what? I know my business better than anybody else. So I'm probably the best person to do this. Um, because I can act faster and and, but what I'm experiencing now, um, is like, uh, and this is just over the past week, like I'm trying to drive home this AI native concept of like there's getting help with from AI to write emails. And then there's like transforming your business infrastructure so that your humans cannot have to write emails anymore, right? And one is just thinking of AI as a tool. The other is like AI native thinking and not, not even AI first thinking, right? Because I think that language is problematic. Um, because it's either or, AI first or human first, right? Instead, it's like we need you to think about AI first, but not at the expense of like the high value work that humans could be doing. Erin Wiggers: Right. Chris Carolan: Right? And it's been interesting, like there's been two spots where I have not been able to overcome the natural training program of AI and that is one, like HubSpot and what it is and its purpose, right? And then this exchange with Claude, right? Every AI system has been trained on millions of pages where HubSpot equals marketing and sales CRM. That mental model is so deeply embedded that explicit documentation. So any custom instructions, any skills, any agent triggers, like anything I do to try and not have to correct the record every time. He tells me, uh, that explicit documentation gets treated as interesting exception. Erin Wiggers: Mhm. Chris Carolan: Rather than an operating reality and the bias reasserts itself the moment there's any ambiguity. Right? So what it's saying is 99.9% of the world understands HubSpot as the marketing and sales system. So who the hell are you, Chris, to suggest that we can do like ERP level functionality because nobody else in the world is saying that. So, right. Chris Carolan: Understandably, it's like I got to prove it. And since every conversation is like a new conversation, right? That's been one spot. But this other spot of funnel thinking, calendar based pacing, like phases, timelines, week one, week two, uh, prioritization, obsession with quick wins. This is all stuff that I've known I've not liked for a long time and it's at the heart of some of the problems that we have to deal with. Um, and uh, like here's an example from something I'm working on like right now, as we speak, right? Where so, now again, this is in the face of like I've never been so, like I've never created so much value in my life. So this is not to complain about AI at all. This is to help explain the problem that we're trying to solve, right? So uh, working on building some more show prep and making content easier in the website, right? And I don't know if I can make this any bigger. Um, but either way, uh, like, okay, I've created an implementation plan and it's like, and I even, like he brings in the skills that are supposed to overcome these things as he creates the plan, right? Uh, the workspace aligns with VF platform principles, capability building, content pipeline integration, multi-host collaboration, HubSpot Unified View. Cool. Four phases. Target, February 26, March 26, April 26, Q2. He just loaded a skill that says, don't ever give me timelines, basically because you're an super powerful AI that's going to get this done in 30 minutes. Right? And what I found that does, especially when it relates to quick wins and the avoidance of overwhelming humans, basically, and or helping them feel in control, architectural get dis, architectural decisions get made based on maybe we don't get to this Q2. Like what happens if we don't? We got to build in all these fallbacks and contingencies and then we have to clean up later the the dummy data that was added because that's the the right way to do it. Like, and it's just been driving me uh, bonkers. Because literally, oh man. Like in what we're, and this is an easy fix, right? It just, Claude, put your AI native hat on. It's not going to take you three months. You get it done in 30 minutes, so just do it all right now. Right. Chris Carolan: It'll get done, but what I'm experiencing now as I get more again, increase my capability of how development works and how the layers and where the gottchas are. Erin Wiggers: Mhm. Chris Carolan: When it comes to making architectural decisions to get to the quick wins first or to um, like contingency, this future where we might not get to the thing inside of the 30 minutes that it's going to take, like that's when the gottchas show up as far as um, like that the bugs show up because those decisions are made in that way and that's how deep, you know, this problem goes. I'm curious if you've had, if that resonates or if you've had any success like just finding ways to overcome that.

[17:49] Erin Wiggers: Yeah, no, I I definitely run into that a lot, especially, um, doing a lot of like spec driven development where I'm giving it a lot to work on at one time, uh, and it kind of just defaults to this, oh, well a human must be doing this. Um, I found that if I, uh, if I when I create the specs, if I tell it this is going to be executed by an AI coding agent. Sometimes I'll tell it specifically which one. It tends to help if it's like, I'm doing this for Replit, uh, because there's some like idiosyncrasies there, or like I'm doing this for Codex. So it kind of knows the environment that it's going into. Um, and that tends to help. It will still phase things out sometime in like a weird way where like it puts in placeholders and like then like subsequently forgets that those are placeholders and like hard codes things that it subsequently forgets was supposed to be like dynamically pulled. Um, so yeah, I definitely think try when when creating the plan, reminding it that, you know, this is for AI to execute, not me, is helpful, but it still kind of comes back to your point where like it was trained on billions of data points that, you know, are to the contrary. And so at the end of the day, all of these models are just prediction engines and so it has much more data to support that particular prediction, um, then, then something else. So, you know, maybe as they start training the models more on or even just as the context window gets bigger and big enough where it can just remember and, and prioritize some of those memories. I think would be helpful, but for now we're kind of just like stuck reminding it of things.

[20:35] Chris Carolan: Yeah. Yeah. Um, good question from from Ryan here. Uh, how do you think about the differences as it relates to environments of low versus high risk? And that this is a fair pushback. And all I'm trying to accomplish is that I'm okay with like contingency planning and like architectural decisions based on real risk, but that I'm not okay with it when it's based on like the the human tendencies of, uh, let's try to get the quick win now and then if we can get to the rest later, great. Uh, what if this thing breaks? Well, you're an AI, like you're going to be able to fix it. So let's not make decisions based on that. You've got the API key, every system is working. We don't need a bunch of fallback hard coded data like to support this area, right? So that's where, um, it can be very hard, especially if you're not a developer, you start using these, it can be very hard to to tell the difference. Um, and so that's, that's, uh, and I do come off as as fairly like, let's not worry about the risk as much. Um, I know more than most people, but I have discovered like when I asked to add security, like to start thinking about security for the Value First team website, I did put the qualifier in of I'm not enterprise. I'm not, I don't have to be SOC two compliant. Like I'm not ISO compliant, like because if I don't do that, he's going to come back a bunch of best practices and just load the system up with, you know, again, these requirements that have come from complex systems managed by humans not meant to be managing complex systems. Erin Wiggers: Right. Right. Chris Carolan: Um, so we're going to get back to the channel. trying to make this entire series about AI. But it's also the reason we're here, right? Like AI, it like as organizations struggle with complex implementations, they don't have to struggle as much anymore. If if you know how to use AI, right? But historically and still increasingly turned to manage service providers creating a compounding problem. So dependency on external experts, uh, to manage the system complexity developed by those external experts. So like all the feelings of like, man, why does WordPress have to be so hard? Like they always develop it this way and then I can't just fix a forum. Like, well, the system incentivized them to build really complex things that you couldn't handle. So then you keep paying them like month over month. Um, system supposed to simplify operations, become unmanageable, uh, also known as the SAS trap. Uh, each layer of expertise creates new dependencies. Uh, new dependencies require even more expertise to manage and then spiraling costs with diminishing internal capability.

[25:15] Chris Carolan: This part is just crucial, right? And people are just wholly, businesses are wholly unprepared to use the resources that they already have. The humans that know their business inside and out, know the customers inside and out to just point this capability in the right direction because they're so used to thinking that they cannot do it. Erin Wiggers: Mhm. Chris Carolan: Right? Um, as you've been, uh, traveling in the, uh, HubSpot ecosystem and agency land, have you seen, can I hear like any stories that come to mind like related to, uh, you know, this impact that we're talking about.

[26:11] Erin Wiggers: Yeah, for sure. I think, um, especially since I've done a lot of, you know, CMS development, um, where theoretically the modules should all be very intuitive to use. Um, but people coming to the HubSpot CMS from other platforms where it's not as intuitive, uh, there's like an immediate kind of pushback on really any kind of like, why you can edit this going forward or like you can create your own landing page. You kind of get this like, I mean, I'm not a developer, you know, or like, uh, I'd really rather have an expert do something like that. Um, and you know, at this point, especially with, you know, if you have a system that's set up correctly, the risk that the average user is going to blow anything up by pushing buttons is very minimal. If it's not minimal, you're working with the wrong people. Like like if there's a chance you could nuke something by accident, you're in, you've got the, you've got a whole other issue. Um, so yeah, just people are, I don't know, it's, it's like, it's almost like they have imposter syndrome where it's like they haven't quite gotten over that hump of this is figutable for me, for me personally, who's not a developer, who's not a, you know, XYZ. Um, and I think that that's the the unlock is that is a deep belief that it is figurable. Um, I think that is what has gotten me to where I am in my career. Uh, just because I I don't know that barrier of like, oh, I'm not a, I'm not a softwa, at some point I was not a software engineer and I was like, well, I'm going to figure it out. Um, and yeah, so I I have always kind of felt like part of my job when handing off these systems has been to empower people and like build up that confidence of like, you don't need me. I don't need to hold your hand. Like you have got this. Um, and it's been and like that's the rewarding part for me and why I think everyone should kind of embrace this because when you build something, maybe it is complex. Maybe it's complex by by necessity. Um, but the average person who's in it every day has no problem using it. That's the gold standard. Um, and yeah, and and I think the other thing that I, I always kind of encourage across the board is everyone needs to, everyone on the team, regardless of like rank or, you know, role, everyone needs to stay as close as possible to the systems. Um, we kind of touched on this a little earlier where it's a lot of people have that kind of like, well the system is over here and I'm going to stay in my world and like work on my stuff and it, it operates in the background. No, no, no. It operates right under your feet is what it is. Um, like it's not, it's not something that can be in the background because you need it. Like you these are the rails that you're kind of rolling on and if you're not aware of the systems and how they work, you're going to miss things and you're going to, you're going to do more manual work than you probably would need to. Um, and and then you're not going to see the types of friction and and inefficiencies that you can solve with AI. Um, so I just, I think it's super important to empower people to just get in there, get your hands on the system and and really work to understand it. Um, and that's like a step that gets skipped all the time.

[32:16] Chris Carolan: Yeah. Uh, and we need leaders to to do better in my opinion. Um, because I've seen so many situations like even people that are forced to use the system every day, like they know all the problems, they know how hard it is, they care so much about the customer that eventually they just settle for the situation. They stop asking for help because they don't get the help that they need or, uh, we got to get permission and we need, all right, we did all this work. Oh man, this is so frustrating to to get the approval. I did all the work to put the budget together and then we got all the people together and everybody thought it was a good idea and we got the approval and budget time comes in three months and it's like, oh, we have to do more approvals. We have to reassess like and then there's eight levels of approval. It's like, no, I'm not even doing like the first part anymore, right? Uh, and that's like if you're, I don't want to call it lucky, but in the less risky situations, you have people that comply and use the systems. In the higher risk situation, uh, they go find other tools and they bring them in and they start using them on their own accounts and now it's connecting with your data layer and especially now that AI is involved, like so easy to plug in systems, like with AI because generally if you're in this mode, you probably don't have like a great AI policy in place. So people are free to go and find like when we talk about like human potential flourishes through enablement rather than control, um, they like frontline teams especially, they do what they need to do to get the job done when they like if if serving a customer is their role, they will serve that customer and that's where it's like we've seen posted notes, like copy and pasting into PDFs and then we're editing the PDFs and they're just doing all of this crazy stuff because that's what I got to do to make sure the customer knows what's going on for me to feel good about actually having finished the the problem or fixed solved the problem. Right? And like, it's so easy to just talk to them and say, all right, what's in the way? Like, but if you don't ask that ever, it turns into this cycle. This is the concern of many when it comes to replacement uh, by AI, right? And this whole concept of you won't be replaced by AI, you'll be replaced by somebody who's using AI. If you try to use AI with this mentality, like you will not like be one of those, um, because you've never asked like what's possible and can I do this. And um, I remember like back in, you know, 2021, just this, this, uh, you know, this loop of understanding that we've asked our teams to use these systems, sometimes with like five-year-old laptops where HubSpot somehow takes 30 seconds to load and these random like all the things are popping up loading separately. I've never seen anything like it. But the first question was like, how old is your laptop? Oh, five years. Oh, when you've ever asked like for a new one? Uh, like, oh, yeah. Yeah, I stopped asking though a while ago. And then when you understand how leaders understand the value of those people, it's like, oh, they're they're yeah, they're data entry. Like we probably can't ask them like to do a lot more and it's this cycle, right? Meanwhile, those people are the unlock for like so much value for customers, right? And that's one of my favorite parts and also one of my like devastating parts when I got had to move on from those situations. I was like, oh no. I was going to help all these people. Right. Chris Carolan: And now they got to stick in their ERP. Right. Chris Carolan: Oh, man. So I'm again, that's why I'm so passionate about this situation and, and really, I think one thing we should be able to count on AI for now is to create the space to actually think about this stuff, right? That's so that's what we're trying to do. Let's just highlight it, right? And like name it, label it. Like you've we've been feeling all of these things, rising costs, increasing coordination, longer timelines, mounting frustration, growing a disconnect between agencies and the goals of the business, right? Like that's been felt for a while, creates all these hidden costs and just um, it's so obvious to everyone. And it's been interesting to see some break like in the clouds, but you still have to get very specific with the language you use because there's all this baggage related to everybody knowing exactly what we're talking about. And if they're not in an enterprise situation where they have lots of budget, like every dollar is scrutinized like based on like the trust in my opinion that has been just yeah, I I won't get to negative. Let's try to wait for that. Um, are you seeing anybody out there like start to come at this from a different perspective? Have you, um, uh, like what are other options in terms of like it can be so easy to build a business and say, all right, I'm in the manage services category and this is how billable hours work and this is how you build it out. Like, what kind of alternatives have you seen? Have you been talking to, you know, the community about in terms of what's like maybe some better ways to to think about this stuff.

[42:34] Erin Wiggers: Yeah, I think the the main thing that people can really try to work towards is to just snap out of this maintenance mode. Um, the idea that you will ever have a period of time when your tech stack is just being maintained is not a reality. Um, so you're either iteratively improving your system or you're watching it become technical debt is really is what it comes down to. And so it's not necessarily about let's complete this and then move into like wait and see mode. Um, you know, it's more about once we've completed this, what's the next logical step to scale that? And, and be more in the mindset of, okay, based on the feedback we get from what we built, how can we turn that feedback into improvements to the system so that we're not maintaining it. We're constantly improving it. And part of that is building in those signals directly within the implementation, being able to track things like, you know, uh, who's looking at the enablement materials and like who's following the process? Who is um, you know, where are people getting stuck? Um, so building in the mechanisms that are going to allow you to figure out what's working, what's not and how to kind of adjust as you go. Like that's the opportunity for manage services to continue um, on like a retainer basis after project implementation that doesn't feel icky to me, you know, like it's not like I you're just paying me to like watch the error logs or whatever or like, you know, come in and push some buttons when when y'all don't necessarily know what to do. Um, and again, I think it goes back to just my ADHD and being like so super like, okay, we finished this. I'm going to move on to this and like I've got other things. Um, that yeah, I think there's a lot of people who do this well. Um, oh, and it's it's a lot of people like that are in the value first kind of collective like it's it's the people that you see on LinkedIn who are genuinely helping other people, um, providing solutions and and not gatekeeping, who kind of who kind of have that um, that aspect to to what they're bringing to the table. Um, so yeah, it's uh, it's definitely something that I feel like is catching on slowly but surely. Um, Well, it's definitely getting harder and harder just because it's harder to defend billable hours. Which is at the heart of all of this. It's so easy and it's still we're not out of this yet because I've even seen some like people in AI space, right? That understand doesn't make any sense at all to think about any project work in terms of hours. Yet by the end of the conversation they're asking me for an hourly rate. And it's like where have we? That's again, their training data like is just built in that and of course that's easier to look at the balance sheet and understand your budget for the month, but at the end of the day that puts both both sides in a position to not like serve like emergence, which is what you really have to build scopes for and like just everything is changing. So why like there's no world where it makes sense to have these very black and white, you know, scopes anymore. And this is where like as soon as you can make that switch. So you're going to hear us talk about the phrase manage partnerships. Uh, I I for me that makes sense as just a simple switch going from services to partnership because just because we want to be done with that thing, doesn't mean we want to be done with you. Right. Chris Carolan: You have lots of ideas. So let's develop the next thing. Let's continue to build. Um, and it just changes the the whole like perspective. Uh, right? And all of that I think we can try to give some decision structure. It's what we're going to try and do in this series, just to think about because a lot there's going to be a lot of overlap in what we talk about, but the intent is to get granular because right now we're we're presented with these situations where this is the the way it's always been done. And like how do we transform from like a whole infrastructure, but also understanding each human in the process, each process related to this. As we go through this, I'm going to, I'm going to go through all seven commitments and then that's the rest of the series is going to be about. Um, it's like all in the name of enabling natural flow and helping people, you know, transform instead of optimize. Like. So number one, we will focus on outcome transformation, not just output, uh, completion. Uh, I'm going to talk a lot about the unified views framework, um, as a surprisingly tangible outcome that is a yes or no measurable thing, uh, versus your 90-day checklist, your onboarding checklist, your quick win checklist, whatever, right? Uh, And to your point about the hours, like that's this is really the crux of it because at the end of the day, how many hours you spend does not tell me what you've got, you know? Like it doesn't tell me anything about, you know, the the outcome. It tells me the output. It doesn't tell me the outcome. And if you're focused on the output, like I could hypothetically work on something for 100 hours and it still not work. But I promised you 100 hours. So there you go. Oh man, I wish that was like some some make believe like example that you just gave. It's like when we, this is what's behind those stories where people do three years of implementation, spend a bunch of money, a bunch of hours and you end up with a system that nobody uses. Right. Chris Carolan: Right? That's real. Like that's so real. And that's where when it doesn't matter what tool comes next. It doesn't matter what AI comes next. As long as you're in that bucket, it's like, no, I don't want to we hate the system, but we don't even want to think about going through that again. Like, and that's where, you know, these 600-hour projects that's like, oh, do we, did we complete the 600? Okay, like, I guess we're done. It's like, no. Um, that's not going to not going to stand anymore. Uh, we'll enable natural value, not just manage handoffs. We will adapt to emerging needs, not just adhere to initial plans. Oh man, I'm so excited to talk about this one with you because you've been working on an app, right? And just you've always been thinking about it differently than I see others think about it in terms of like, yes, we have to understand what we deliver and, and roughly how much effort it's going to be and who needs to be involved and who's going to be in charge of what. But how do we build for emergence? Like and there's this a scope, the uh, some articles I started, started drafting with Claude about this concept of what this boils down to is because of the pace of change right now, because of AI, it's making the out of scope conversation like inevitable and super painful, right? Because everybody around you is starting to experience this, but like it wasn't in our scope. So either it's going to cost more or take longer and that's like and it's you see these situations, hey, that's the contract, right? Sorry, like I wish you could help like no. Um, so I'm excited to dig into that one, uh, with We will build on distinctive strengths, not just apply standard methods. Uh, that's a partnership for you. We will create continuous learning, not just post delivery reviews and back to capability building. We will enable transparent progress, not just status reporting. So much of this is reliant on just visibility. That's why I'm so excited about the unified views. It's literally just built into the name. Like unified customer view. Like that's the visibility everybody wants when you get that, all of a sudden alignment and trust like happens. Um, And we will make value tangible early, not just at completion. Uh, I'll need help with this one for sure, um, because this this starts to get into the heart of we want time to value, right? Of course, as quickly as possible. Like you're spending a bunch of money up front, but where that has taken us is is quick wins, right? And um, there's there's some some more traps and like all this, all these outcomes that we don't actually want just because we need to feel the value and that takes us back to what we were talking about Claudy earlier. There's so many consequences of how you make this specific decision, right? That um, uh, I just look forward to digging in there too. Uh, So, value first delivery and approach that transforms implementation from a mechanical fulfillment process into a dynamic collaboration that creates lasting value and enables continuous transformation. Like if we can just get everybody on this path, because AI is right there with you if you want to collaborate with it, uh, we'd love to help you get there. Um, any of those specifically, uh, that you're looking forward to or anything that popped into mind as I went through those.

[50:05] Erin Wiggers: Oh gosh