Impact of AI on MSPs

[00:00:06] Welcome to MSP 1337. I'm your host, Chris Johnson, a show dedicated to cybersecurity challenges, solutions, a journey together, not alone.

[00:00:23] Welcome everybody to another episode of MSP 1337. I'm joined this week by Jimmy Hatzel of Hatz AI. Jimmy, welcome to the show.

[00:00:33] Hey, thanks for having me. Glad to be on. Excited to hopefully talk some AI today.

[00:00:38] Well, I mean, I can't imagine bringing someone who has .ai in the company name. So I'm going to say let's talk about future of MSP in an AI world.

[00:00:50] Obviously, your company, the products that you have obviously facilitate some really interesting things in the enablement for an MSP.

[00:01:02] You know, we were joking about it earlier. You know, I remember when I used to do ConnectWise tips and tricks and how to get people to do, you know, open tickets, whether they did it through the web portal or they did it through, you know, submitting an email.

[00:01:16] We would always say if you call, there's a bill or there's a fee attached to it. So please open a ticket. Then let's, you know, figure out what's going on.

[00:01:26] And I think one of the products that you were just describing to me kind of solves that. Like, I don't care if they call anymore because AI is taking that call and creating a ticket on my behalf.

[00:01:35] So, you know, along those lines, what are you, what do you envision are going to be sort of like the big, I mean, we've seen in the security space and we can go down rabbit holes, I think on that.

[00:01:48] But like from an MSP perspective, where do you see the real, sort of the new frontier of AI and how it intersects with what MSPs do?

[00:02:01] Yeah, great, great question. You know, I think that there's a lot of like misconceptions. I think the key thing to keep in mind is every huge technology trend.

[00:02:14] Historically, there's always been a group of people that could help other people figure it out. And in my experience, you know, the most recent ones, right, movement to cloud, cybersecurity, those have been MSPs, right?

[00:02:29] Sure.

[00:02:29] So, like, I don't, like, I think people are nervous of like AGI sort of like taking everything over and what's the point of even working anymore. And that may very well happen. But, you know, we got to almost pretend that it's not.

[00:02:45] And I think it's a long way off. I can't tell you exactly when, right? You know, you can ask Sam Altman that and form your own opinion.

[00:02:54] But I think that there are some new realities that are quickly forming on sort of what's possible and with AI, whether that be inside your MSP or with your customers.

[00:03:11] And I think it's happening faster than a lot of people think. So, like, there's, there's like people will go immediately to like, oh, well, you know, AI will go and automatically solve every ticket and we won't need level one anymore.

[00:03:26] And like that kind of stuff. And, and like, maybe, but like, you know, we still have plenty of time before that's going to happen.

[00:03:33] And I think like you can look at like past sort of industrial revolutions on like, what's going to happen.

[00:03:42] So if you look at like the invention of the steam engine, right, this is a common comparison.

[00:03:49] All right. All of a sudden you didn't need to make all the ingredients for a product you're selling.

[00:03:55] You didn't, you know, you had a trade across regions, you had cheap, reliable transportation that sort of came quickly.

[00:04:04] And I'll get to the comparison of like where it falls with MSP and sort of like, you'll get on a train.

[00:04:11] Yeah, yeah, yeah, yeah. Two, two, two things sort of happened, right?

[00:04:15] You had like, you had companies that like, either like just laid off and their, their team and became like small, like sort of mom and pop businesses.

[00:04:27] And, and, you know, some of them survived, some of them didn't.

[00:04:31] And then you had companies that began to scale really quickly.

[00:04:35] So like Guinness, Guinness is a classic example of that, right?

[00:04:39] They were, they were, you know, a local brewery at the time.

[00:04:43] And instead of sort of like, okay, now we don't need to, you know, source everything.

[00:04:49] We can get ingredients from different places.

[00:04:50] We can sell across the entire world.

[00:04:53] You know, they decided to, to do that.

[00:04:56] And what they did was they kept instead of like saying, Hey, I can get, you know, 3X output out of one person or 4X output out of one person.

[00:05:05] And let me, let me drop down three or four, you know, you know, 75 or 60% of my staff.

[00:05:14] I'm going to increase what I do.

[00:05:17] And so I think what, what you're going to see is an increase in the amount of work or the amount of customers that an MSP can support with the same number of people.

[00:05:31] And it's, it's starting to happen a little bit and it's already happened with technology we have, but we're, we're sort of like getting to this like agentic world.

[00:05:41] And I think it's going to be the big talk of 2025 where you sort of have all these day-to-day functions that you do, whether it's, you know, communicate with a customer, create tickets, right?

[00:05:53] These are things that we've already sort of solved for, or like, you know, reset someone's password or, you know, create an account for someone or something like that.

[00:06:02] And previously you had to really spend a lot of time mapping out exactly how that happened if you wanted it to be automated.

[00:06:10] But in this like agentic world, you can really have AI do a lot of that with a human as a co-pilot.

[00:06:18] So it, yeah. So it's just like, you know, I, I think like that's sort of where things are going for MSPs.

[00:06:26] Cause it's here in the consumer space, right? Like we did, it wasn't like an Amazon made a big announcement about incorporating the chat bot into your, your app.

[00:06:36] Like it's suddenly there, suddenly you can ask questions directly to the chat bot, like say about the product that you're looking at.

[00:06:42] Like my favorite question is like when I'm looking at new laptops and I'm curious, like Kate pricing is where I want it.

[00:06:47] And I'll ask questions like, how much does this laptop weigh?

[00:06:50] And you're like, Oh yeah, there's definitely a compromise.

[00:06:53] I'm saving $600, but I'm not saving my back.

[00:06:56] Cause it's going to weigh six pounds instead of three or whatever.

[00:06:59] But like that was not a, an AI interaction two, three years ago.

[00:07:05] That was absolutely like you either found it in the, in the forums through a search or you maybe broke down and decided to chat with customer service.

[00:07:15] Most likely you went out of your way to look up that laptop on the manufacturer's website.

[00:07:20] If you really wanted to know what the weight was.

[00:07:22] So just even on that, the amount of time it's saved, you know, saves us in the consumer space.

[00:07:28] So I, I can't imagine, you know, you started talking about tickets.

[00:07:31] Like there was no direct cost associated for me.

[00:07:36] If I don't care about my time, right.

[00:07:38] In the consumer world space.

[00:07:39] Like it's just me, you know, figuring out my black Friday habits.

[00:07:43] Right.

[00:07:43] But if you look at it through the lens of a MSP, we're talking about actual dollars being saved.

[00:07:48] If you can lower the, the minutia around users will always need someone to reset their password at some point in their 30, 60, 90, 12 month life cycle of, I forgot.

[00:07:59] I deleted it accidentally in my password manager.

[00:08:02] I can't find it in my password manager.

[00:08:03] And let's just continues to go on.

[00:08:06] Yeah.

[00:08:07] And I think you make a good point because AI is really different in past mega trends like cybersecurity or movement to cloud.

[00:08:17] And that it's happening bottom up at this, like before it's happening top down.

[00:08:21] So like all of a sudden the average user or maybe, you know, many, many users say like 30 to 40% is, is typically what the percentage of American workers who use AI in a given week.

[00:08:34] Like Gen AI, like chat GPT or something like that.

[00:08:38] So they're like experiencing this and using it sort of before companies have time to put a strategy in place, put guardrails in place, that sort of stuff, which wasn't the case.

[00:08:52] Like nobody, like, it was, it was a lot smaller percentage of users who were, you know, setting up lots of cybersecurity measures on their computer.

[00:09:04] You know, right.

[00:09:06] Like, like, so it's like a, it's a unique situation and, and it, it, it, it is forcing people to adopt quicker than they like because the guardrails around it and the value that users are getting out of it are happening sort of like before companies can say like, here's what you're allowed to do.

[00:09:23] And here's what you're not allowed to do.

[00:09:24] So, and, and people sort of do it and they don't necessarily tell their boss either.

[00:09:29] So like that, like,

[00:09:30] No, this is the epitome of shadow IT with the repercussions potentially being felt immediately and it's painful or we have no idea what those repercussions are yet to be because that data leakage, if you will, hasn't actually been consumed again to, to know that it was just sitting out there.

[00:09:52] I just have to ask the right questions because someone was not realizing that there is no real walls being built around this sort of like free generative AI tools that are, you know, so prevalent now.

[00:10:05] Yeah.

[00:10:05] So it's, it's like two parts.

[00:10:07] It's a, or three parts, I should say.

[00:10:09] It's, you know, user training data, right?

[00:10:12] If someone's using free software and uploading a bunch of company documents, the next model's trained on it.

[00:10:17] So proprietary secrets, API keys, things like that.

[00:10:20] Like you, you, you run into a lot of problems with that.

[00:10:23] And then there's access.

[00:10:25] So not things not set it being set up properly.

[00:10:28] Sure.

[00:10:29] All of a sudden AI has access to a bunch of files.

[00:10:31] It wasn't supposed to, it may be in a walled garden.

[00:10:34] Right.

[00:10:34] But if, if user access control wasn't provisioned properly you know, maybe the user who had never thought to look up the weight of the laptop is now getting it every time.

[00:10:44] Right.

[00:10:45] And, and the weight of the laptop could be the CEO's salary or, you know, like some company legal documents or something.

[00:10:51] Right.

[00:10:51] So let me ask you a question.

[00:10:52] And I don't know that, you know, hats AI does this, but you know, I've had it brought up to me on more than one occasion.

[00:10:57] Now, wouldn't it be slick if I could use AI and have a conversation with it around like, Hey, go tell me what's not configured properly and fill in the blank firewall config.

[00:11:09] And Hey, go check that against all of my clients.

[00:11:13] And because you have been feeding it enough about your organization, assuming you keep it in the right container walled garden, that it would be able to then give you those reports.

[00:11:23] Because I think one of the biggest areas that collectively we all fail at is to maintain proper configurations, even if they were properly configured out of the gate.

[00:11:33] Maintaining that state of security is, is very hard.

[00:11:37] And as you pointed out, and more often than not, it wasn't set up properly to begin with.

[00:11:42] So just even being able to go back and address those types of, you know, accelerated, rolled it out.

[00:11:48] We'll go back to this after we get them up and running.

[00:11:51] And that never happens.

[00:11:52] Are you seeing that becoming a.

[00:11:55] Yeah.

[00:11:56] Yeah.

[00:11:56] So like, like we see people doing that in more of like a human in the loop type of way right now.

[00:12:02] And like AI has the last mile problem.

[00:12:05] So like the foundational models are capable of incredible things.

[00:12:10] But they, they don't necessarily have a good way to interact with, you know, your Fortinet firewall configurations or whatever it is.

[00:12:19] They don't talk to assets very well.

[00:12:21] Yeah.

[00:12:22] Yeah.

[00:12:22] Yeah.

[00:12:22] And it's not, it's not an AI problem.

[00:12:24] It's just that last mile delivery, which is sort of happening like in the background.

[00:12:28] And it isn't necessarily what everyone sees because the big news is like, you know, GPT, whatever came out.

[00:12:34] So like if this, if the, if all new development stopped on like AI models today, it would still be another five years of, of innovation of people figuring out the last mile problems for, you know, all these edge use cases.

[00:12:49] So like, like that, the examples of that is exactly what we're like focusing on in 2025, right?

[00:12:56] We sort of like, like we launched this year, right?

[00:12:59] We, you know, brought on the first, you know, beta testers.

[00:13:03] And then, you know, we sort of went public towards in the back half of this year.

[00:13:07] And like 2025 is all about like last mile delivery.

[00:13:11] How do we, you know, allow AI to safely connect into sort of like all of these edge use cases and edge systems.

[00:13:18] And, you know, it, it shouldn't have to be like an API request for everything, right?

[00:13:23] There should be easier ways to do it.

[00:13:25] Yeah.

[00:13:25] API scare me still because we don't seem to be evolving the API from a security standpoint very fast.

[00:13:34] Yeah.

[00:13:35] Yeah.

[00:13:35] Well, and it's like, you know, if, if it's not set up properly, like you set up an API with admin privileges, like to go look in, you know, name your system.

[00:13:45] And then it's shared amongst the whole team, right?

[00:13:48] Or it's not set up right.

[00:13:49] And it's leaking the key somewhere.

[00:13:50] Like there's big repercussions that can come from that.

[00:13:55] So like, like AI's problem is its utility or it's problem for MSPs because people find immediate value a lot of times from, from using it.

[00:14:06] Whether that be, you know, something as simple as summarize my email or summarize my calls to like, you know, something much more complicated, which, you know, name your, name your use case, put together this, you know, onboarding document.

[00:14:20] Right.

[00:14:20] Like, like look through these logs for me, that type of thing.

[00:14:25] Which gets into more of like, it's not that it's intelligent.

[00:14:28] It's that it has been given the right formulas to analyze that, which it has access to.

[00:14:36] It's just looking at it like a, it's a, it's just a file to it, right?

[00:14:40] Like it's unstructured data that it's looking for patterns based on what you, what patterns you've asked it to, to go and get.

[00:14:46] But, you know, I've, I recently started using, it's called HiDoc and HiDoc has a chat GPT built in at the hardware level.

[00:14:55] And when you go and do the, you know, transcribe it into something meaningful, aside from it being an audio file, it actually identifies who the different speakers are.

[00:15:05] So you can have two or five or 10, it'll identify the different speakers.

[00:15:09] And, and then you've seen like with, you know, Zoom AI and some of these others, Otter AI, you know, depending on your flavor, it gives you some decent summaries of a meeting and they'll give you some decent, maybe some decent action items.

[00:15:22] I started using this thing and I'm like, oh my word, this is insane.

[00:15:25] Like it's leaps and bounds ahead of, you know, what these other tools are doing.

[00:15:30] And it's doesn't even have a monthly subscription fee attached to it.

[00:15:35] Yeah. And I just like, like the utility that people get causes them to sort of forget about security very quickly.

[00:15:45] And so like, we're, we're going to see huge amounts of problems as people start leaving jobs, right.

[00:15:54] Or whatever, right.

[00:15:56] And they've been using AI in their personal account and have every single meeting they've ever been in recorded and transcribed right on their personal email address.

[00:16:04] Like, because the company never got, you know, their, their plan together to roll out something that's approved and the person just set it up and paid the monthly fee themselves or, you know.

[00:16:14] And we have laws for this too, right?

[00:16:16] Like depending on what state you're in, one party or two party, depending on California's two party is, you know, say using Zoom AI without recording the meeting.

[00:16:26] Does that count as a one or two party conversation?

[00:16:30] Does it count when it comes to like the equivalent of using the telephone?

[00:16:34] Cause it's not a telephone.

[00:16:36] You know, we're going to see a lot of laws coming, I think in the near future too.

[00:16:40] Yeah. You know, I'm, I'm no lawyer.

[00:16:42] I think a lot of times there's little pop-ups, right.

[00:16:45] This meeting is being recorded or something like that, or AI assistant has entered, right.

[00:16:49] Which people may turn off because they think the announcement's annoying.

[00:16:52] Um, but I, you know, like, like, I think AI is one of those things where the innovation happens so fast that the legal system won't be able to catch up with it.

[00:17:03] So we'll, we'll be, we'll be, uh, litigating problems from a year or two before when we're iterating like a hundred times faster.

[00:17:14] So like, I, I think people are just going to do it and it's going to be a mess.

[00:17:18] And, you know, like, like, I think it's an opportunity for MSPs though, right.

[00:17:23] To have that privacy and security conversation.

[00:17:26] I mean, it's not necessarily bad.

[00:17:27] It doesn't make it bad.

[00:17:28] It's just, I think to your point, going back to the beginning of like, you know, most organizations and specifically MSPs don't have policies in place yet that say what is appropriate for how I interact with AI.

[00:17:40] And I think to some extent, the, the challenges that say, you know, you know, hats AI has is that when you think about your products in the context of what they're for can easily put employees of an MSP into a mindset that says, because we use.

[00:17:58] Hats AI or, or fill in the blank that's been designed for the MSP as a, as a product to consume to help with, you know, from a tools and services standpoint, that that automatically means the things like maybe it's Grammarly or ChatGPT.

[00:18:10] Or fill in the blank are automatically okay to use.

[00:18:14] And yet, because they're not being provided by the organization, they're not containerized.

[00:18:18] They don't have the security protocols in place to, to protect both the employee and the organization from, and the clients from, from threat actors or just having data out there that shouldn't be.

[00:18:30] Are you seeing any challenges that way?

[00:18:33] Like, I just feel like it's super easy almost to consume a product as an organization and get behind that.

[00:18:41] And then because that's the, the pioneer into our space, it's suddenly become Pandora's box without even, you know, taking a minute to look.

[00:18:50] I think we're even like a few steps behind that.

[00:18:53] So like the conversations I hear from our partners is, I went into a subsidiary of a well-known company.

[00:19:01] The president said that no one's using AI in their organization.

[00:19:06] They don't want to look at any AI products because they're not sure that there'd be value in it.

[00:19:10] The MSP runs a Cisco umbrella report for, you know, desktop traffic to free, free AI sites.

[00:19:21] Yeah.

[00:19:21] And 25% of all users had been on it in the past week.

[00:19:26] Sure.

[00:19:27] And it's like, yeah, it's like, like there's like, there's that, that bottom up and top down sort of like happening.

[00:19:33] Like, I think the, the bottom up is going to happen anyway.

[00:19:36] And I don't, I don't know.

[00:19:38] Right.

[00:19:38] It could be that like an approved tool brings that like, you know, 25% to 40% use.

[00:19:46] Sure.

[00:19:47] But, but I think it's more like, like, like business leaders are just unaware of how people are using it because there's an incentive misalignment to tell your boss that this task that took you two hours takes you 20 minutes now.

[00:20:05] And right.

[00:20:06] Like you don't have, or you don't even do it anymore because you automated, you automated a script using chat TPT and you never touch it again.

[00:20:12] And you, you, you bring up an interesting point.

[00:20:17] CompTIA released the, the state of the industry 2025.

[00:20:22] 2025.

[00:20:23] It had a lot in there about AI.

[00:20:26] And one of the things I thought was really interesting is, is to your point that 25%.

[00:20:30] I think it's much higher than that.

[00:20:32] I think there's still, it goes back to, if we go back a few years to when those two verbal commands that we can say to, you know, the IOT devices in our home, whether you're an Apple or an Amazon subscriber.

[00:20:49] And that's AI.

[00:20:52] I mean, it is like, it may not be the same, it's not the same type of AI per se, but you're, you're communicating with a voice that is in fact, basically converting what you would ask it verbally into, you know, a written prompt that you would do in chat GPT or, or any of these other ones.

[00:21:13] And if you look at how long we've been doing that without anybody really talking about it, in fact, more or less promoting it.

[00:21:21] In fact, the only thing you would see is in certain environments are like, you can't bring a phone in here largely because it either has a camera or it is smart enough to do those other things.

[00:21:29] We've ignored it.

[00:21:30] We've largely ignored for the last five years, anything AI related until suddenly we saw it being incorporated into products that were not expected.

[00:21:39] And the affordability of bringing the AI to my desktop or my laptop, or even a Raspberry Pi has revolutionized the ideas that we can come up with.

[00:21:51] We haven't gotten to a super intelligent being with AI.

[00:21:56] I don't think we ever will, because today we've got to keep in mind that people are still the ones responsible for building the AI functionality.

[00:22:05] And until somebody finds someone that has that super intelligence to build it, I don't think it will, you know, exist, but I might be naive.

[00:22:14] No, it's like the craziest thing because you have someone who, something that has glimpses of super intelligence in lots of different areas, right?

[00:22:24] You ask, you ask AI to...

[00:22:27] In our, in our mind, our perception of what it's doing, we believe it to have that level of capability.

[00:22:33] Yeah, yeah.

[00:22:35] Well, and in some cases it can, like, you know, there's been tests of like foundational models taking like NCAT exams or LSAT exams and, you know, performing in like top percentiles.

[00:22:47] And now sure, those were, they could definitely be optimized, but there's no doubt that these AI models could do amazing things in very specific tasks and then just fail miserably counting the number of R's in the word strawberry, right?

[00:23:01] So it's like, it's just not intuitive to, to us at all as, as people.

[00:23:07] But like, you know, like some really complex, like coding tasks and, you know, debugging tasks, you know, AI can, can, you know, have glimpses of like a senior crazy engineer, right?

[00:23:22] But then also like, just make things up and just be totally wrong about something too.

[00:23:27] And justify being wrong.

[00:23:28] Like, no, the reason I came to this conclusion is because that's how it should be.

[00:23:34] Yeah.

[00:23:35] Yeah.

[00:23:36] Yeah.

[00:23:36] So that's, that's, that's interesting.

[00:23:38] So, I mean, circling back, largely AI is happening in the MSP space, in the business world.

[00:23:46] And to your point, we're seeing a lot of the bottom up to how AI is being adopted.

[00:23:53] Are you seeing as a result of, of things like that you're doing, where it's an opportunity to engage leadership that's potentially buying your products to say, Hey, you know, we should probably do an evaluation of, you know, AI, because this is a, you're not just buying, this isn't a car.

[00:24:11] Like this doesn't come with the owner's manual and the glove box.

[00:24:15] And it looks like a car has a steering wheel, has pedals.

[00:24:18] I mean, even, even cars that are electric cars, they had to do things to them to make it at least have the illusion of operating just like the gas powered one they used to have.

[00:24:28] Yeah.

[00:24:28] So like it boils down what we've seen to two hours of training, 10 hours of usage.

[00:24:35] That's what it takes to get someone where they're thinking of their own use cases.

[00:24:38] They understand AI and they, you know, can use it as a co-pilot in their, in their daily life.

[00:24:44] And now for really technical people, they, that two hours of training might've been like Googling and experimenting and figuring out like, like, you know, you've shared with me some of the ways that, that you and people on your team have used AI.

[00:24:57] Like you probably never sat down for a certification or maybe did a CompTIA certification.

[00:25:02] I don't know.

[00:25:03] I bet those are coming.

[00:25:08] If you're an engineer and an MSP, we didn't call it AI, but you've been doing that for a long time, right?

[00:25:15] Whether you were doing it in RMM.

[00:25:17] Yeah.

[00:25:17] But like you jump in your RMM tool and based on what you found through Google or whatever the search tool was, you figured out a way to create automation that while it's not necessarily a generative AI, you're, you're removing the human element.

[00:25:33] And I think that's a large part of what generative AI is making us do faster because it's able to not just take a command and execute it.

[00:25:43] It can learn from the command success or failure or find the missing pieces to what would make your command do more than it actually does.

[00:25:51] And you don't have to be a programmer or write code to pull that out.

[00:25:55] And we've had examples of this for the last decade, right?

[00:25:59] Like, you know, no code and object-based coding that, you know, you drag things onto a screen and all of a sudden, boom, it's done.

[00:26:06] I remember building my first auto attendant.

[00:26:08] I think it was Jive Communications.

[00:26:10] And I'm literally in a flash-based environment that would literally light up red if it was going to be an invalid connection for my auto attendant.

[00:26:19] Like, can you imagine, you know, 20 years ago, you call a hospital and you're like, option three, you're like, busy signal.

[00:26:25] Like, apparently that was not the right option to click, right?

[00:26:27] Like, we've evolved from that even with automation and what would be now considered to probably be more in that AI category.

[00:26:37] Any, anything that you want to highlight or throw out there for those listening about what the future of AI impact on the MSP space is before we wrap it up?

[00:26:48] Yeah, I think just, you know, preparing for, like, you got to have some fundamental AI skills, some competencies in your organization.

[00:26:57] Like, because things are happening faster than people realize.

[00:27:01] And I'm saying that and I live in it, right?

[00:27:03] And, you know, I'm working on things and three weeks later they're outdated or there's some new way to do it, right?

[00:27:09] Which is crazy.

[00:27:11] But, like, we're nearing the, what I have been calling the agentic future.

[00:27:18] And I'm not the person alone to do that.

[00:27:20] But where all of these sort of, like, last mile things have been figured out and AI can act as an orchestrator to do them for you.

[00:27:29] But there's a bit of, like, making sure you can set up all those last mile deliveries.

[00:27:33] Got it.

[00:27:34] So, like, I think that that is going to be where a lot of MSPs will adopt quickly and run fast.

[00:27:41] And you're going to see a massive gap between those who do and those who don't in efficiency in their organization.

[00:27:48] Every MSP is busy.

[00:27:50] I've never talked to someone who works at, runs, owns an MSP who just, you know, I got plenty of time.

[00:27:56] You know, like.

[00:27:58] Yeah, it's funny you say that because with Trustmark, you know, we run into that quite a bit of, or just adopting frameworks in general.

[00:28:05] You know, I don't have time for that.

[00:28:07] You know, we had this project, or we this, or that.

[00:28:10] I mean, quite honestly, with what you're saying, those that embrace and adapt to how AI can, I mean, remove a lot of those barriers to getting things done.

[00:28:22] And especially those, you know, more remedial tasks.

[00:28:25] I mean, in theory, they should start freeing up their time to focus on things that are more business efficiency models that are not tied to whether or not I did or didn't open or close a ticket.

[00:28:35] It's whether or not we've expanded, you know, to a new vertical or we've gotten some of those things that we keep saying we're going to get done are now, in fact, finally at least visible on somebody's radar.

[00:28:48] Yeah, and well, I also want to caveat that too, because I think a lot of people will say, okay, well, I'll start adopting when that's here.

[00:28:56] And they discount the efficiencies that are possible today and very common today.

[00:29:01] And just the amount of benefit an organization can get just by, you know, getting everyone on the team their 10 and 2, right?

[00:29:11] Their 10 hours of usage and two hours of training.

[00:29:13] So, like, that's an easy way to get started.

[00:29:16] And that's, you know, what we recommend for every MSP before offering stuff to their customers.

[00:29:21] And then for all of their customers, how do they get started with AI?

[00:29:26] I like it.

[00:29:26] Well, this has been an episode of MSP 1337.

[00:29:30] For those listening, thanks and have a great week.