037 The Impact of AI on Remote Accounting Work

Today, the fallout from Beardgate continues. What about email marketing? Should I be leaning into that? Is it still worth making blog posts on my website? And how do we make AI agents as helpful as possible for our firms? Come on in. Let's talk about it on Jason Daily. Q and A Wednesday. A lot of fallout still from the beard trimming. People asking if I'm actually alright. I made a joke on Twitter the other day that I trimmed my beard to slow down the DMS of people asking if I was okay. And so then a bunch of people asked if I'm okay. Listen, I'm okay. Also, like putting yourself out online. And I think I've talked about this. The whole notion of a daily show, like, if you're gonna turn up on video every single day, buddy, you're gonna have some days that are winners and some days that are losers. If you invite somebody into your living room every single day, there's gonna be days that you're proud of what that living room looks like and there's gonna be days where you're not. And that's just how it is. This has been. I'm actually having way more fun with this than I thought I would. It's a fun opportunity to explore kind of the nuance and more just kind of the things in between all the big things that people always talk about when it comes to running firms like pricing and, you know, all of the normal kind of discussion cycle that we see everywhere. I think this has become a fun place to talk about, kind of the space in between that stuff that has a big, big impact on how we work. And so I've been having a lot of fun with that, been having a lot of fun with the people that I'm meeting throughout the course of doing the Daily Show. I don't know if it'll be a daily thing forever. I do need to kind of optimize for sustainability for me. So it's probably not going to be an every single day thing all the time. There could be days where maybe it's just an audio feed too. Like maybe video is going to be too hard. Maybe I'm traveling, something like that. I still think in those cases I would probably post to YouTube still and it would just be a static image or something like that. But gang, I'm doing great. I'm having a lot of fun. There's gonna be days where I'm gonna look better than other days. But reminder, I got a whole bunch of kids. I got a life. We all got lives, right? That's just that's just how, how things are. We're all just not looking at you on camera every day so you can hide it. Okay. Nathan Sosa. I find a lot of people saying blog posts are not as good as just social media posting. Is that true? So this was on the episode where we talked about SEO and kind of the value of that trending downwards. Are blog posts not as good as social media posting? The two are not mutually exclusive. I would say that if you post something on social media, especially if it's something a little bit longer form, you gotta be posting that stuff on your blog as well. The blog you're gonna have a much more kind of long term approach to or that's an asset that you can send clients back to for years. So many things that we tell clients over and over again that we should have written a blog post about years ago and just send people to that. When it comes to discovery, no new people, I mean, virtually no new people are stumbling into your blog posts on your website compared to the reach that you have on social media. So when it comes to finding new people, the way to find new people is generally not putting a blog post on your website. That's what social media is most valuable for right now. Like that, that is where you draw new people, the folks that are super into what you do, maybe you get them into your blog. Is there maybe still some value in doing a bunch of blog posts for SEO? Maybe, but like the value of that is diminishing by the day. I think the more valuable version of blog posts is to continue doing them, to build your own library of context, your own library of thought leadership, because that is an asset for you when it comes to referring clients out to it to answer questions. But also, as we've talked about with some of this AI stuff, the more documented your thought leadership is, the better. But if your goal is to find new people, pull more people into that bubble, social media is the way to do that right now, in my opinion. Alison Reif Martin, you may have explained this already. Can you explain context? Yeah, I keep talking about context and I've realized I'm using context in two different contexts. Teehee. Okay, so there's context limits within the scope of a language model. And those are some nerdy words, context limits, like, and let's use ChatGPT for example. ChatGPT sits as a chat interface on top of the large language model, that is GPT 3.5 or GPT 4. Those models have context limits. That is, they can only remember so much stuff for any given generation. So when you ask it a question and it returns a generation of text, the things that it takes into account to create that generation of generation of text is limited right now in chat GPT it's limited to 4,000 tokens or about 3,000 words a word. A, the ratio of words to tokens is about three to four right now. So that means within Chat GPT it will recall like about eight pages of conversation. If you assume you have about 500 words on a written page, that's about eight pages of information. And that's why you can't use ChatGPT for really like big data processing operations. And at a certain point in the conversation, it begins forgetting this stuff from earlier in the conversation. That is due to context limits. The fact that with each generation of text it can only take into account so much context to do that. And those contacts context limits have been going up rapidly. And older versions of GPT it was way less. GPT4 has a theoretical context limit of 32,000 tokens, which is a lot, 25,000 words or so. You have to request access to that. I don't know anybody that has access to it. And Chat GPT arbitrarily limits GPT4's context to 4,000 tokens today. So as far as you and I are concerned, if you're, if you're using ChatGPT, the limit is 4,000 tokens, about 3,000 words. So those are context limits for AI models. And I just shared the other day, Anthropics, Anthropic just updated their model to have a context limit of 100,000 tokens, which is wild. I'm doing some playing with that right now. Super cool. You can chuck in huge amounts of information and it can contain, it can take all that into account. When it makes a text generation, it can take a huge volume of information information into account. That is one definition of context. The other situation where we've been using the phrasing of context quite a bit is around crafting your own source of context for a language model to draw on. So for example, when you create your own chatbot, which with GPT4 now it is good enough to, with some prompt engineering, only refer to the context that you give it. So for example, let's say I've written 10 blog posts, I can have the text of those 10 blog posts and tell the chatbot only reply with information that you can glean from these blog posts. And maybe those blog posts are 10,000 words and that's, you know, 14,000 tokens or something like that. Because Chat GPT is currently limited to 4,000 tokens, it can't take all that stuff into account. Practically, what would happen in ChatGPT is it would only remember the most recent 4000 tokens. There are some ways around this context limit stuff. That was the episode where we talked about embeddings and long term memory for AI. That was a pretty nerdy one. But the other situation in which we're using that context framing is when it comes to developing your own context. So if you wanted to have a client facing Chatbot, what is the context that you would have underneath it? What's, what do you want the bot to know, basically? And the fact that these AI experiences are kind of taking over right now and are ultimately really valuable is putting a greater premium on the context that you can develop for your clients. What's the proprietary stuff that you know that they're not going to get from ChatGPT? Right. And so that's where the example we used of Matt Metrus, who's a crypto tax expert, he does talks everywhere. You could pull in transcripts and his blog posts and all these things. And he could have this proprietary body of context that he puts behind a language model. And with a chatbot app like, you know, chat thing or MyAskai, we've talked about, that chatbot could reference his proprietary context that lives behind that chatbot so that we all could chat with chatbot and ask questions and it would answer just according to the context that he provided it. So two different situations where I'm using the phrasing context. One is context limits for a language model. How much information can it hold at any given time and then what the context is, the underlying context. If you have a big long conversation with ChatGPT, that's all context. When it generates the next bit of text. What we're looking at now is like building our own chatbots on top of like a context library. So and that context library could be, I mean, it could be a piece of legislation, could be IRS publication, can be anything. But for us as thought leaders and professional services, it's interesting to start thinking about what that body of context could be that we could develop for our clients and then give them access to that context via a chatbot. So kind of two different framings of context there. This episode is sponsored in part by Client Hub. This week on Tales from the Hub. Yeah, that's a thing. Let's talk about a firm. This firm is called Super Smart Accounting Solutions. They realized profitability of their firm comes down to how efficiently they can get responses from clients. You know, that old chestnut. So they prioritized a practice management platform that would make their client experience so good that the clients would just be foaming at the mouth to get them answers as quickly as possible. Client Hub hey buddy. It was the obvious choice. Practice management with an amazing client portal at its core. The fine people at Super Smart Accounting Solutions. This isn't fictional. They love the idea of a system that is boundary list. Clients are users just like the staff are users and the client portal's like the core of the product. It's not just this afterthought sort of bolt on thing, right? Client Hub Let me tell you, it blew their minds. They loved it. They adopt it super easy. They had that sucker going in a week. A week. When's the last time you changed practice management systems in a week? That's Tales from the Hub. We'll be back with more tales next week. Check out Client Hub at the link in the show notes video description one of those this episode is sponsored in part by LiveFlow. Let me tell you about Live Flow. Okay? LifeFlow is a powerful product that puts your advanced reporting on Autopilot. You know what that means? Autopilot means you just. You set it up and it just goes. Software is designed to simplify the process of creating reports by turning your spreadsheet into a scalable and real time FP and a platform buddy. This means that you can spend less time manually updating spreadsheets. That's a win. And more time analyzing the data and advising your clients. One of the key benefits of Live Flow is its ability to automate the process of streamlining reports from QuickBooks. This means you can easily import all your client financial data into Google Sheets where it can be transformed into a fully customizable thing that can be tailored to meet any of your needs. You little snowflakes even do like cool consolidations, custom dashboards, all that stuff. Anything that's in that QuickBooks file, you sync it back and forth. Bada Boom, bada bing. LifeFlow's gonna do it for you, bud. Learn more Check out the link in the video description. The show notes. You know, check out the link. Robberger asked about email marketing and I was on that episode about SEO and kind of where that slots in. I've never been an email marketing guy like a cold email marketing guy. I'm 100% an email newsletter guy. I think an email newsletter is like the top of your funnel and how you capture new People to continue getting information in front of them. So if somebody sees a cool social post that you do, or if somebody comes to your website and downloads a free resource that's valuable for them, how do you make that relationship sticky and how do you get back in front of them on a recurring basis? And it's got to be by capturing their email and then doing an email newsletter or once you capture that email, putting them on a drip campaign. So some email email marketing tools will let you design drip campaigns where, you know, two days later it'll send this email that you've developed and three days later it'll send this email that you developed and you can even set up like branching logic according to what email gets sent if they didn't open the previous one, or what email gets sent if they clicked on this specific link in the previous one that is a drip campaign. I love those. I think you need to do that because getting that person's email address is how you continue to develop that relationship rather than it being a one time thing. Cold email marketing, I've never been that guy. I know people who have good success with it. Absolutely. To me it feels like a similar to SEO blog writing. It feels like something that's just getting more and more saturated by the day and AI generated text is making that more and more saturated. So it seems like cold email marketing like that value of that to me would seem like it's shrinking. But I'd also admit I'm not an expert on email marketing and, and how to make the most out of that stuff. The more the to me, like the current best paradigm in my opinion is you're publishing helpful content on social media for a very specific type of person with the goal of getting them back to an email list where you then kind of foster that relationship and to ultimately making them a client. Peter, do you think AI will completely take over the accounting profession or completely automate it? Here's kind of how I see this discussion now and I think this framing is helpful. It's easy to understand and I think it's something that maybe we ought to rally around more. The notion that, you know, a profession or a job or a person, that work is a collection of a whole bunch of tasks. Some people in their job do one task all day long. You know, in more enterprise applications you have entire floors of people who are just sitting there completing a task. And if that task got automated, absolutely, that'd be disruptive. Accountants, man, we get stuck with so many tasks. Accounting is like to me is not A task as much as it is like a department in a business that is a collection of a whole bunch of different disparate tasks that are constantly changing through legislation and changing needs of the business and all of that. So like, there is accounting, like, and, and you know, maybe the most similar task is bookkeeping, but virtually none of us just do the bookkeeping like we're doing all of these other things. So I think the better framing for Will AI take Over X is acknowledging that every single task that is on our list today is on a path to extinction. It will eventually be eliminated. So if you look at what accountants did in the year 1900, a lot of those things we probably don't need to do anymore. So every single thing on a more granular level is on a path to extinction. And sure, if all those things got displaced in the next 12 months, then that person is out of work. Reality is though, accountants do very different things from one another. So to me, that is a more helpful framing to think about. The tasks themselves being displaced. And yes, ultimately they will be replaced. And yes, right now stuff is getting replaced faster. I think not. I think it is faster than ever before. So some changes that in the past may have taken a decade, now they're happening over the course of years or less. So I don't think that it will like take over the entire profession. It'll absolutely change the profession. We are, we are all laser focused on accounting, but it's changing everything. I mean, we've seen in the last week a huge number of companies have made, you know, laid off thousands of people, quote unquote, due to AI. And I'm always a little bit skeptical of how much of that is like, are they just looking to assign a reason to needing to make layoffs? But that's kind of becoming in vogue right now, similar to how the recession kicked off a bunch of, a bunch of layoffs. Now AI seems to be kicking off a new wave of layoffs. So AI is impacting the way all of us work. But I, I think it will definitely automate the tasks that we do and accelerate the rate at which we'll be forced to change. Jacob Schroeder was talking on Twitter yesterday about yesterday. You know, we've got so many small businesses and accountants and all that who are so change averse and are still, I mean, they're still on, you know, not leaning into cloud technology yet and really just, I mean, you got firms running on paper still, you know, not doing paperless stuff, right? And there's a ton of those out there. So is it realistic to say that AI is going to blow all that stuff up in the next 12 months when people have been so, so slow to adopt to all this stuff in the past? I do think there will be a, increasing, a greater level of risk to those people than ever before. Like the delta between the high output people and firms and the low output people in firms. It will continue to grow at a faster rate than ever before. So that stuff's happening. Tasks are gonna get automated. We're gonna change the ways that we work probably kind of at like the task level and that stuff is happening faster than ever before. And honestly, like, we don't ultimately know what that looks like. Like, what does it mean for that to change so fast? How quick can the government adopt and our clients adopt and our firms which have initiatives inertia going in a certain direction, how quickly can they adopt and change? Don't know. We just know that it's happening faster than ever before right now. Herman, on agents, that is AI agents that can go out and autonomously do things. Said I had the thought before if it would be possible if you could have a central database like an airtable that stored relevant information in a helpful format that can then be used again as needed for other tasks. So for example, you request and maintain the annual turnover for the tax year once the accounting file has been locked in your database. Yeah, so this is around the notion of, of these autonomous agents. And we're just seeing our first one now being released with pixie. That episode was another week, another killer AI announcement for accountants. How do you make those agents as helpful as possible? Yeah, I think Hector, I think. Or her. Sorry, Herman, I think you absolutely will be storing that information in as helpful way as possible for our AI agents. I don't know if that's stuff that we ultimately do or if it's stuff that like our practice management system does. But using the PIXIE agent as an example, they're using the zero API to go out and fetch the net income figure from zero and the sales figure from zero. And that API is an example of a database where that, you know, that information is all stored in a helpful, easy to access format. So yeah, kind of going back to the conversation of context. What are all the valuable sources of context within your business? From client facing, thought leadership to context. For you as a manager to say, for all of my team members, you have access to this bottom and you can come and ask the spot questions before you need to come and ask me the question. You know, SOPs are another source of context. It's worth thinking about what all those sources of context are within your firm. I've heard from a few firms now that are launching like internal GPT chat kind of experiences. Those will only ever be as good as the quality of the information that you have sitting behind those chat bots. So definitely worth thinking about. How can you organize your information and your thought leadership and stuff that is usually distributed in a pretty disparate way, like, you know, hopping on a zoom call and doing a training that only ever happens once and it's not recorded. And anybody that you hire after that, it's gone. Like there's no permanence to that information or that training that you did. You know, a better version of that is saving transcripts and recordings of those things so that they can be a source of truth well into the future and discoverable down the road. But yeah, how we organize our context and that information is getting more and more important. This episode is brought to you in part by Meester and Company, who's hiring a CAS manager. You can put job listings on this podcast. Yep. Meester Company's a firm in Northern California. They're hiring a remote cast manager. Teeny bit of travel required, but you can generally be anywhere. Let me tell you why Meester & Co. Is better than the firm that you work for for right now, Buddy K. Listen up. A few reasons. Our team is top priority, not revenue. We concentrate on our strengths and aren't afraid to turn down the wrong clients or projects. Our firm success starts with our team members. You sold yet? We believe in work life balance. They probably won't have you running a daily show. No billable hours or timesheets. We focus on results. They specialize in tax, accounting and advisory engagements. Steer clear of audit and attest work because who the heck wants to do that? It is a modern, forward looking firm that is building out their cast practice and you just might be the perfect person to help them do that. Let's just. Let's just zoom out. Of all the firms you could go work for the firm that's advertising on this silly little podcast. How many people listening to this? Oh, not very many. So if you're looking for something new in the cast space, check out this job listing. Put it in the show notes. Never know. Could be. Could be the job of your dreams. Brian Krueger what will prevent unscrupulous individuals from building a chatbot using other people's content, I. E. Download all the transcripts from your videos. Various Social media posts, articles, talks, et cetera, et cetera, and use it to monetize for themselves, or even not monetize it directly, but use it as a free offering to drive users to, say, a website. Nothing right now when it comes to publicly available content like YouTube videos and podcasts and stuff like that. So we're seeing a lot of folks do this around, like the Joe Rogan podcast and the Lex Friedman podcast, building their own kind of landing pages to chat and ask questions of those transcripts. And this is honestly, like the hardest thing with AI is that AI is fundamentally built upon stuff that humans created. And the rules of how, of kind of how that information can be used and scraped are really murky. There have been some cases in the last couple of years that have trended toward making it more okay to scrape publicly available stuff and use it yourself. So, for example, TaxPro talk, is that what it is? The big Tax Pro form that's been out there for years, you could go out, it's all publicly available, all of the posts, all the forum stuff. I mean, a huge percentage of of the GPT models is like Reddit and public, like, discourse places like that where you have rated responses and you can pull the highest rated stuff. So take something like taxprotalk going out and scraping that entire forum and then building a chat bot on top of it to provide more helpful answers to tax questions. From a technical standpoint, you could do it. The legality of all that stuff is really murky right now. And where we see this most is probably around AI art. But some of the same principles apply to anything that we publish. I mean, online, like blog posts and stuff like that, or replying to a Reddit comment. When you post a Reddit comment, are you giving OpenAI permission to train their models with that Reddit reply? Kind of murky, I will say, you know, that I think I'd use this example before. Derek Sivers, an author, he publishes a book every couple of years. And they're generally just compilations of his blog posts around certain topics. And it's a, it's a printed book and it collects all of those posts kind of around certain themes. And ultimately there's never, there's not anything to stop somebody from going out and finding all of those blog posts and reading them one by one. But that function of aggregating them, you know, is helpful and it's valuable and people go out and buy the book. The most helpful aspect to me of like, that chat experience right now is that somebody did the work of aggregating all that stuff for you into a single place that is helpful. There's probably like a celebrity level of like your Joe Rogan's where people will go out and do that and stand up that chatbot themselves and how kosher that is. Can you charge for that? Like, I don't know. That's a hard one. Right now you look at like the music industry and what those people are going through with somebody being able to spin up a Kanye song and most people not being able to know the difference. There's a lot of really sticky stuff around that right now. Now the ultimate protection to that, and we've talked about this a bit in kind of the chat bot and context discussions, is proprietary context that only lives behind that chatbot. So obviously I've got a ton of public context and rolling that all up into a single thing like that's pro. That's it is helpful because you don't have to look in all these different places if you're trying to find a specific answer. But I could still put additional context behind that. That doesn't live anywhere. It's not part of a blog post, it's not part of a YouTube video, anything like that. And the chat bot will not give you explicit access to that underlying context. But like it won't, it won't generate, you know, here's all of the context that doesn't live anywhere else, but it will let a user chat with it. So that's interesting if you want to go like one level of kind of abstraction. Further, many of the kind of emerging AI models right now are getting really good by being trained on the state of the art models, which right now is GPT4. So they will train the lesser AI models on GPT4's output and it improves the quality of the lesser models. So if you did put something proprietary behind a chat bot, like could you then train another bot on that bot with the proprietary information? Probably. But for most of us, like for what we do and the ways that we would use chatbots with our clients, I think we're, most of us are going to be small potatoes enough where that's probably not an issue. I mean, here we are talking about AI chatbots and accounting, like this is not the mainstream group and if anybody put a GPT, a client facing GPT experience out there in their firm right now, like you would be bleeding edge. If you had that GPT experience in the firm to help junior people or something like that, like you would be bleeding edge. So too early to know, like 100% what all that looks like and attribution for AI context and stuff like that. But that stuff's got to get more fleshed out than it is right now. Nathan, for the FAQ this week, wondering what your thoughts are on remote work trending down by corporate America and bigger firms as well. I don't. Like, I've never been like a thousand percent remote work guy from the beginning. I've. I literally went into an office every single day from the day I graduated college until just this year. And like, I'm still kind of getting over that. That being said, my accounting practice was fully remote from day one. We had people all around the world on that team and that there wasn't anything really local about that. So I've built both. At the end of the day, like it is nuanced, like there is not an absolute great answer. There is, in my opinion, a greater amount of latency in remote work and you lose the adjacency to your colleagues where you just pick stuff up because you're alongside them. And there's types of work where that matters and there's types of work where that really doesn't matter. So depending on the nature of the business and the stage at which that individual is at. Are you in a startup that is doing something new and novel and you are like, everyone is learning by leaps and bounds every single day as they are working on this problem together and developing this hard thing. That stuff is definitely suited for in person work or a fundamentally different type of remote work where like you are literally in an open forum all day long or something like that. So there are types of work where eliminating that a little bit of additional friction between how people learn and, you know, like those serendipitous water cooler conversations where you pick stuff up and just overhearing things in the office. There's environments where that really matters. There's environments where that doesn't really matter. A place where I think that's hard for almost everyone is when you're just new to a company unless you're coming into, I mean, unless you're coming into a job where you lick stamps all day or something like that. And the process is very cut and dried. I do think in my experience, like it is a little harder to bring those people along. It really forces you to invest in standardization and in systems. And so that's. That is a huge spectrum of how standardized firms are and how well documented their things are. You should have that no matter what. But the solution for that in the past has been, oh, Tina will just come and sit with you and train you for 20 minutes or you'll walk by and you'll see, you know, this other, this person's doing something in a different way than you. And that's how you'll figure it out. Like we have been. That's just how in person work always was. And so how you, how you replicate that in a remote environment, it's just different. So I think it really depends on the nature of the work. I do think this came up in the visionary versus integrator discussion. I do think most accounting firms, like, we're not generally doing earth shattering visionary types of stuff. We may be helping people in a unique way and with your unique voice, but fundamentally what we all do is pretty similar. And in the context of that discussion, it was like, do you really need a visionary to run an accounting firm or do you just need an integrator, Somebody that'll operate and build the systems and get it done? Because we're not like doing much in the way of creating new things from scratch besides just building a business, right? So I don't know, like, so for that mega innovative company that is like doing something novel and you know, building a space shuttle for the first time, I don't know that that's super analogous to accounting firms because what we do is pretty procedural. But it depends on where you're at. If you're in a big mature firm where there's very clear expectations of people like those folks I think don't really need to be in office. I what I miss, I do miss with remote work. Just that like the serendipitous human conversations like we, we've go gone so far the other way with locking down meetings and stuff like that that there is no bumping into people in digital hallways unless you manufacture this kind of in a different way. And that I do miss. And I think there's a lot of great things that come out of that that I haven't quite figured out how to replicate in a remote work environment. That being said, it's really easy to focus on the downsides of remote work and some of the things that it lacks. There are a ton of things about remote work that are just way better. So like two sides to that scale from rolling out of bed and going to work to the overhead, like the fact that you don't have to have somebody have a place for all these people to come and work, the fact that you can hire people from around the world, doesn't matter where they are. They can come do the work. That's all that matters. So upsides and downsides. And I think the right answer depends on what that team is doing. Nathan Sosa. No, we said that one. C.J. i'd love to hear your thoughts on the Canopy Tech Guru alliance since I have an in depth understanding of both companies. Yeah, so Canopy just kicked off this like, partnership thing and there's a bunch of companies on that list. There's Tech Guru, there's Ryan Lazanis's community future firm, Accelerate. There's Gary Boomer's consulting group. It's interesting. I get why people do it. Just kind of the cross pollination of audiences. And I completely understand the value of this that, like, for obvious reasons. But I tell you what, and I kind of spilled the beans a little bit on when we were talking about ads for this podcast around, my thoughts is anytime you do that stuff, I really do think you're creating this inherent conflict of interest. And sometimes that matters and sometimes it doesn't. But it makes it a little bit more sticky in certain situations. Inevitably, that's just how it is. And the counter argument could be like, how do you not have the same problem with sponsorships? I would say the difference with sponsorships is sponsorships are these encapsulated things where it's very obvious when it is a sponsorship. And I think there's greater isolation there and how it could impact the ways that you talk about things. So with a sponsorship, you know when you're getting an ad read with a partnership, I don't know that I know, you know, if there's like a rev share there or something like that, I don't know when I'm hearing things that are impacted by that partnership versus things that are not impacted by that partnership. And people could say, does the same thing apply for sponsorships? If you run somebody's name through the mud and that decreases the likelihood of them to come back and sponsor, like, are you gonna be less likely to do that? My answer to that is I generally don't crap on people. Like, I like. I highlight the good stuff and I don't find myself having that issue. If, and there's a lot of other people out there that do this, if you spend a lot of time talking about what other people are doing wrong and what you disagree with and how they're getting it wrong, then yes, it creates a bunch of potential conflict issues because you're talking bad about those folks. I generally don't talk bad about folks. I would rather drill down on what's the best of the best that's happening because I think that's more interesting than what people are getting wrong. And there's definitely, yes, this drama element of neener, neener, look at how they're not doing this thing right. And I get that that drives engagement. But for me and how I learn, I always learn not by how other people do it wrong. I learn by the best of the best and the cool new stuff that people are doing. So I've found that when I focus on what people are getting right, I don't run into conflicts. And that for me that, that hasn't really muddied the waters around sponsorship. But partnerships to me just feel a lot more ambiguous. You know, does that make you less likely to recommend carbon when carbon's the right solution for somebody? And so I get why people do it. You know, there's a lot of things out there where you have these preferred vendor lists and people are paying to be there and there's a, you know, a rev share and I think everybody gets how that stuff works, but I think it just at the end of the day gets in the way of transparency and doing the best by people. I don't know. I like, I also come from this, from a super privileged angle of, you know, the, of the paid accountant community that I have, which does not have any vendor tie ups financing about half of what I do. Like I've had that advantage from day one where if I didn't have that, I would probably be reliant upon partnerships to pay the bills. And so like the different business model there is like if those partnerships are your source of revenue, then what's the alternative? So I kind of have to temper that with the fact that I do feel like I'm in a very privileged position to not have to have those tie ups. But yeah, I don't know, I don't like that stuff. Obviously you don't see me or realize on that list or anything. And I know all the people involved and we talk on a pretty regular basis, but I just, I don't know, it just feels murky like this. There's, there's so many advisors and all like all that stuff in this space. And if you sit down and talk with someone, it's kind of like when people talk about, you know, if you could put politicians in like NASCAR uniforms that had the badges of all the people that have, you know, donated to what they do. If you sit down with an advisor today, like I want to know, like, who are they partnered with and Oftentimes you don't. And if I don't know that, then how do I know if they're being influenced by that stuff? Maybe they aren't at all. But the fact that partnership exists in the first place, like, to me, there's kind of an inherent level of conflict there that just complicates things. So I just don't do that because I just don't have to fuss with it. And I don't ever want there to be concerns about me saying things through a particular lens. If I went down the road of dunking on people more and pointing out the things people were getting wrong, that would get more problematic. I couldn't do that as much. But if I can set up camp in kind of the positive headspace and continue to learn from the people who are pioneering and doing cool new stuff and not worrying about all the people that are getting it wrong, then I don't find myself offending people and offending brands and. And I don't think that becomes an issue. But, yeah, I don't know. I get why people do the partnerships. It's just not for me. So that's been it. Q and A Wednesday. You got cues. Drop them in the comments. Thanks for coming and hanging and we'll be back tomorrow. Toodaloo.