Transcript
which is a big topic of conversation and also I think the latest buzzword that may have replaced the cloud. And AI has a lot of, there are a lot of AI products now and it's hard to determine what is actually AI, what is safe to use, things like that. So today I brought in a couple of IT leaders that are here to discuss a little bit about AI. And I also have some examples. I asked the Spiceworks community what kind of AI tools they use. So we're going to go over those as well. But first and foremost, Brian, do you want to go ahead and introduce yourself to the community? I know you've been on a couple of times before, but just in case there are people here that haven't met you before. Hi, I'm Brian Walters. I'm the IT Director at Pinnacle Structures. I'm in Cabot, Arkansas, which is about 18 miles northwest of Little Rock, northeast Little Rock, sorry. And just, I get to enjoy doing all kinds of IT things here. And I would say probably in the last six months, AI has just really, really taken over my world. It really has. Yeah. Yeah, absolutely. Cynthia, you want to jump in here? Yeah. Hello everyone. My name is Cynthia. I'm CTO of an MSP called WDCS. We are a healthcare MSP. I've been on quite a few of these talks and I'm happy to be here today with y'all. Yeah. Well, thank you for joining. I really appreciate it. And you have been on quite a few of them and I'm always glad to have you back. So before we jump into actual use cases of AI and things that you might use AI for, I kind of want to talk about the topic of AI in general. So first and foremost, how should IT leaders be thinking about AI in their long-term technology strategy? And is it something that you're currently thinking about? I think that that is definitely the way that IT is headed among many other things in the world. I mean, the term AI is now slapped on everything. It's even used for glorified if-then logic. So you have to really be specific when you're talking about AI and AT and what the use is going to be, what your idea is. I mean, true AI for IT is predictive for anomaly detection and capacity forecasting. It's generative for scripting assistance and config templating. And then also assistive where it can summarize tickets or documentation, run scripts for you. So I think when you're talking about AI and IT, it's very important to specify how you're going to be using that in IT. Again, it's the new buzzword. Like you said, it's taken over the cloud. Yeah. And for us, it's taken over Google. Instead of Google it, we're going to AI it. Right. If I had a dollar for every time I heard AI powered in a product demo, I could afford my own AI engineer. Yeah, absolutely. It does seem to be everywhere. What about you, Brian, as far as including AI in a long-term strategy? I'm really including it really as my co-worker, not to borrow from Microsoft's co-pilot, but I mean, it's definitely my co-worker for sure. I've been in several binds, especially on development, and I'll share a little bit later how well AI has really worked on development projects for sure. My co-worker that I've hired on last year, he wants to show everybody how to one of our biggest tools in the IT world right now. And I think most people, when they hear AI, I don't want to be part of that. So we've still got some time to use it well for ourselves before we give it out to other people. Yeah. Yeah. So whenever you talk about people that are apprehensive about using AI, I'm assuming that has to do with data security or things like that, are there any other reasons that people might be a little hesitant to use AI tools? I think it mainly is just because just probably what they hear in the news and stuff, oh, AI is going to take over the world, it's going to be dangerous. But for us, it's been a helpful tool. There is a couple of users we've actually got to use it. They use it on their teams meetings or to summarize a call, and that's just been a game changer. That's been the biggest use outside of IT in our company is to use it for summarizing meetings. Yeah. Now, Cynthia, are you guys apprehensive about adopting AI tools? It seems like maybe you're a little bit more on the, you know, it can be used for good, but also there are security issues that come with it. Absolutely. And one of the biggest things when we're talking to vendors that are offering AI is how is that large language model or LLM used or small language model? But the LLM in general, how is that used? Where is that data going? Where is that stored? So that's always a concern. Are they willing to sign an NDA? Do they have a business associates agreement with us kind of thing? Again, we're in healthcare, so we're very aware of the security risks and really have to take precautions for that. So I think anybody getting into any kind of AI space, it's very important to know where the data from that LLM is going and how it's being used outside of your current, you know, outside of your employer. Yeah, absolutely. And as somebody that does work in the healthcare space and that does have to, you know, work within HIPAA and all of these guidelines, how are you keeping your users from or your customers or anything like that from using AI in a way that will potentially cause some sort of, you know, damage? As far as our customers using AI, I mean, look, they can do what they want with their systems. Our job is to make sure that we're keeping their system secure. So there are certain, you know, we take precautions like we would for any other site. We block a lot of things. If the customer states, yeah, we don't want our employees on chat GPT or on Grok or, you know, using any of these computer generated graphic sites, then, you know, we handle that like we would any other site. If it was Facebook or TikTok at work, we would block those sites. Unfortunately, with a lot of the search providers, it's integrated now. So it's almost unavoidable to block everything. But, you know, you go about doing what you do and don't let non-administrators run scripts and things like that. So it's just, it's very similar to how we would, you know, lock down anything else for any other customer. Mm-hmm. Yeah. How do you kind of balance the hype versus reality when it comes to AI? Because there's, like y'all have both mentioned, there's so many products out there that are saying that they're AI powered. And I'm sure you even have executives that are also wanting to fall into that shiny new toy trap where they're like, it's AI. AI is the big new thing. So how do you kind of balance that hype versus reality? That is a daily challenge. As far as AI and technology for us as an MSP, I mean, technically, any script you write is going to be considered an AI. Now, we probably generated the script ourselves or maybe got a script from GitHub and edited it some. But technically, I mean, anything that's automated, you could consider an AI, so to speak. But when people want to get into AI, for instance, I have a client now that wants to use AI as their auto attendant. We know how they want to use it. So it's very easy. We, again, we're not AI developers. So we would hire somebody. But you have to know how you're going to use the AI. Being that we know they want to use it as an auto attendant, we just need to know what level do you want to use it as an auto attendant? Do you want it to have an API and integrate into your software to make appointments? What is it that you want? And I think that's the greatest challenge is somebody saying, oh, we can get AI to do that. To do what? What is it that you want AI to do? Because you have to be really specific. Do you want it to answer phones? Do you want it to type emails? Do you want it to, for us, you know, respond to tickets or do password resets? Is it a chat bot that you need? Is it something behind the scenes? So I think that's the biggest challenge because people don't understand. They just say, oh, we'll get AI to do that. What do you want AI to do? How do you want it to integrate into your system? And that's the biggest challenge is that's when people start thinking, oh, I don't really know. I just know that AI can do it. Go ahead. And we're kind of the opposite in that no one's really asking for it here. But we're just we're using it. We're trying to we're trying to bring it in to make our lives better in the IT world and then eventually get to that point saying, hey, look, look what this can do, you know. And I'm more excited about AI on the actual learning aspects and how well it learns because we have a design program that is pretty much been the same thing since Windows 3.1. They just build upon it over and over and over again. And we would like AI to learn that program because there's so many components to that program. But it's that's that's been a challenge for us is for it to learn legacy software. You know, I've had moments where I've needed to write some C-sharp code that I didn't know how to do and it did really well. But but on our end, the biggest challenge is letting it learn legacy software because we use so much of it here. Yeah. Yeah, that's tough. And that actually leads into one of the next questions I had was what are the biggest challenges that you faced when implementing AI solutions, whether that is technical or with users or things like that? What are some of the biggest challenges you faced? Well, like I mentioned a while ago for us, it's it's our secret weapon right now. So we really don't want a lot of people using it because we can be the hero, be the IT hero sometimes. But there are going to be times where it's going to make their lives easier once we can once we can implement it in our daily lives here. So Dan and I, my coworker, we're working hard every day to see what else can we do? Where can we bring AI in and then finally release it to some users? Like I mentioned, we've given Copilot Pro to a couple of users so they can summarize their calls. But we've actually given some of our salespeople some tools to better write emails and things like that. So we're we're kind of still in the baby steps of giving it to our end users. But we would love to give them more power. Hmm. Yeah, absolutely. Yeah. And again, you know, we're we're using AI in our everyday processes where AI is is we have it in our ticketing system where it's summarizing tickets for us. We don't allow AI to to do to do scripts that aren't they it has to be it has to be overseen at all times. So it's not allowed to just say, oh, it's this problem. We'll run the script. We have to give approval. But so we have we do have AI for script generation, for script review. All of our documentation and SOP now are are going through AI. We have it log our security or review our security logs every 12 hours. And then we do now have a new end user support chatbot for things like password resets and answering fax or escalating tickets. Mm hmm. Yeah, that stuff takes up a lot of time. That is it's nice to minimize the time that's used on repetitive tasks, which I think is a big use case for AI. I think Derek had a really great comment here. The biggest challenge is that AI is making users lazy, where users were, for the most part, forced to understand how something works. AI is robust and providing information. However, I find that users take the information and know that the output solution works, but they don't understand how the solution works. So I think that because users don't really understand what is going on behind the scenes, they're definitely more prone to putting sensitive information in there or, you know, not really understanding how it interacts. information in there or, you know, not really understanding how it interacts with their day to day. So, um, and then Chris said, we have someone that's going to move at the end of the year is looking for a new job. He told me that he got a full on phone interview from AI. Oh, wow. Definitely a moment when you start to question authenticity. Absolutely. Absolutely. I was saying we have a customer who's, we're having a phone, actually just an AI receptionist, uh, done for them on the phone. And it's still in the, it's still being tweaked as we go, but it's, it's live. And I can tell you, there are many people that are calling that have zero idea that it's an AI chat voice. They have no idea because her name's Haley and she introduces herself and she answers everybody's questions. And if she doesn't know the answer to a question, she, you know, tells them, we'll have to get back to you. I'm going to put you through to the mailbox for you to leave a message. I mean, people just have zero idea that who they're talking to and making an appointment with is not a human being. It is scary. Yeah. You feel like, uh, the use of AI, uh, can make it folks feel a little bit less secure in their job, or do you feel like, um, it is something that they see as an asset that helps them do their better? Um, is there some sort of blocker there? Yeah, I'll jump in here. Um, you know, um, Dan, my coworker, he's, he's all in on AI, but he, sometimes he'll come back and go, oh, we're going to have a job in a year or so, you know? And I tell him, oh yeah, you know, we definitely will. Um, but I can, I can see where that, you know, that becomes a thought in an IT person, they're going to replace us, you know, and, um, you know, in our company, I don't see that, that ever happening. They're always going to need us, you know, they're always going to need us for the little things or the big things, but, um, it's definitely, it's definitely come up more than once about, is this going to take over our jobs? You know? Yeah. I think both of those, those, those concerns are true. I think, yes, it's definitely going to make things easier. And yes, it is going to, uh, cause people to maybe repurpose their titles and find other things to do or find another position altogether, because some of these low level, um, activities that are performed daily can now be done with AI. I mean, it can be done now. I can imagine what it's going to be three, six months, 12 months from now, you know, because it's just growing in capabilities exponentially. Um, so I think a lot of lower level positions will be filled by AI, but at the same time, you know, that is, if you, if, if you're, you know, a one or two man show three man, small it company, I mean, that also can be very, very, very helpful as it is handling those small fires so that you can concentrate on the large fires or, you know, concentrate on other things. And the best advice I can give to somebody that's in a small shop is to use AI every day and know how to talk to it. That's, that's the biggest thing to do, because a lot of people who use AI don't realize that you continue that you can continue the conversation. You know, it knows where you left off. It, you can, it can give you some advice and say, Hey, go back to step three. That didn't work for me. You know, a lot of times it'll zero in, in there and there you got your So I think a lot of small shops is just learning how to talk to it. Especially for a small shop, I think it's important not to focus on one AI. So a lot of people will focus on chat GPT it's affordable. So I understand. And like Brian said, it does, you know, hold the memory of your previous chats. However, it's not infallible. So it's not a, it's not an AI trained in what you do. It's a, it's a AI that's using a language model from every bits and pieces from everywhere. So it's got its limitations and it does make mistakes. So I think it's important when you're using AI in that manner, where you're asking maybe for a script, no matter how specific you are, it is important to take that output and compare it to an output from maybe grok. And I'm just giving examples here, but it's very important not to take what AI says as the final word, because mistakes are made all the time. So it's, it's, it's vital to have more than one AI that you're using so that you can compare the two. Ask one, is this correct? Ask another, is this correct? Look at it with your own eyes so that you are also learning what's happening so that you can recognize when mistakes are made. Cause it's a fantastic learning tool. And I understand what Chris was saying. It makes people lazy 100%, but in it, I think it's important that we go everything, go through everything with a fine tooth comb so that if we're asking something, we don't know we are, we are still learning from it. So we're actually using it not only to enhance our processes and our positions, but our knowledge, our own personal knowledge. Yeah. Yeah, absolutely. Has AI changed the way that y'all run IT? It sounds like Brian, you have like really, really benefited from it. Cynthia, I believe that you've benefited from it as well. But what are some things that may have made a huge impact? So if someone was going to implement AI, they're looking at an AI tool. What are some of those things that have made a big impact? For me, it's been development, development tools. You know, and like Cynthia said, there's different, there's different ways you can use it. We use chat, we can use co-pilot. I traditionally use co-pilot, but co-pilot for GitHub is actually a different beast. It really is. When I had a C sharp program I was struggling with and finally got it, just talked to it, finally got it to narrow down and actually ended up writing the code for me. And I knew it could do that, but when you get to that first time it does it for you, you're just like, wow. And it actually worked, you know? And there were times I worked co-pilot and AI was not working for me. So that's when I would go out to the forums like Spiceworks and talk about it. But for that one project, I will say the AI did resolve it for me. And that's just been the biggest thing for us, you know, because, you know, they were about to go out and hire a developer short-term to fix this problem and I fixed it for free, you know? So that was a great moment for us with AI. So it's really helped the development side of our job. Yeah. And like I said, we're using it for low-level tickets and we're using a chatbot assistant for also low-level tickets that are when people can go in on their portal as a customer and ask questions where we used to have somebody manning that all the time. So we still have somebody manning it, but now they're tier two. So if they can't get it resolved on the chatbot, which is our tier one, it just gets transferred over to tier two and they get a human. So that's definitely freed up a lot of time for other things and so that they're able to concentrate on the bigger projects. Yeah. I was kind of looking into all of the different use cases for AI because I feel like a big one is coding. I feel like having that review is really helpful, but there are a lot of other ways that people are using AI too that I find really interesting. One of them is predictive monitoring. I don't know if y'all use AI in anything like that, but that just seems like a cool use case. I'm not sure if that's something that y'all have looked into before. I'd like to. Go ahead. Oh, so we're actually, our RMM is getting into predictive modeling. So that's something that we're very excited about. It's a nice way to get a forecast and look into what could be happening with the system that you currently have. So it's one thing to be flagged when you have errors or things that need to be brought to your attention, but it's a fantastic thing to be able to look at those errors and see if those aren't corrected, this is what will happen in the future because this is going to break down, this is going to break down, this is going to break down. So that predictive modeling is amazing for us for that purpose. I know people are using predictive modeling for many different aspects of life in general, but that's our, at this point, that is our experience with it and what we're looking forward to implementing. Yeah, I think it's very, very helpful to be able to get ahead of issues before they start to become bigger. And then Brian, you said that you wanted to look into using some predictive modeling or things like that. Yeah, we have a lot of, well, on simplest terms, one of the biggest things that we support here are cameras, IP cameras. And we get alerts if they're down, so we got to go out there and reset the switch, BOE on the switch. We haven't really implemented anything like that to maybe go automatically AI, go, oh, yeah, let's do this, let's reset that switch for it. So that's an area that I would like to see us really build upon, would be the predictive problems, take care of the problems so we don't have to right away. Before somebody goes, oh, there's five cameras down, well, AI has got to back up for us. I would love that. Yeah, yeah, absolutely. Are there any other uses that y'all want to look into in the future that maybe you don't have right now? I would love for AI to be plug and play. That is, that's something, and I know that eventually that's going to come down the pike where it's going to go in and, and that's, again, I'm talking about it in an office setting where it can go in and look at what it is you do, what you've done, how you've done it, and just, you know, soak it all up. Yeah, absolutely. That would be very nice to have. Yeah, because just because it's AI, it doesn't mean that it's smart. So yeah. Oh, I, I had a an issue once where I put a transcript into a tool, and then it came up with points or like quotes from the transcript. And I was like, these are awesome. And then I went to go into the transcript to see where they were, and they were not there. They just made up quotes. And I was like, okay, so something to remember is you have to double check some of these. Yeah, I mean, I use, I use it to summarize all the time in Teams meetings. And I mean, I've got to go back and edit every one of them. Just save some really goofy things, you know. A lot of times, too, if I bring in people into my office for the meeting, it says, well, Brian said this. Well, I didn't say that. I actually said that. So I've got to edit that too, you know. Yeah. Yeah, there's, it's definitely, I mean, it is a learning model. It is constantly learning. So hopefully it'll get there at some point. But yeah, definitely important to double check. And, and very important. And we say this all the time in IT, garbage in, garbage out. So it's really important to know, to know, to understand that what you put into your LLM matters. What kind of questions you ask matters. And I'm not talking about how you phrase your question. I'm talking about the question itself matters. Because it's going to keep that and use that for future things. So it's, so it's very important. Garbage in, garbage out. Yeah. Yeah. Eric said there too, documentation would love for it to be more involved with our documentation. Oh, hey, watch me do this and then write it up for me, you know. Yeah. Like Cynthia said, it needs to be plug and play. So we're just ready to go, you know. Yeah. The great thing about the, the documentation capabilities are, there are so many softwares out there that allow you to, you know, with AI that allow you to document things that will watch you do something and come up with everything for you. And there's so many to choose from. I've seen some of them there, some of them are really good. And it's very, very important, obviously, especially in IT to document everything. So I can see at some point where something like that will be running all the time. Or it's recording everything, logging everything for you, giving you your daily write-ups. I mean, I can see that, that coming soon. Yeah. Where it's, it's not really going to be an option to turn it off. It'll just be on and running and smile. You're on, you're on candid camera kind of thing. Yeah. Yeah. Which, you know, I know raises some red flags for some people too, is constantly running and constantly looking over your shoulder and things like that. But I feel like we are so watched all the time. I, yeah. I, yeah, I. It's just one more thing. But, and I, and at work, you know, you're not really, you don't really have privacy at work anyway. No. I don't wanna leave my house, but I think at work, it's, there's no, there's no privacy. Yeah, yeah, for sure. I know that at any point my IT team will be able to be like, oh man, she spent her morning playing GeoGuessr on her laptop and, you know. Oh, well. Yeah, so Ryan actually had a really cool question. When dealing with non-IT executives that have the mindset of add AI to every department workflow, what are some strategies or questions we can ask to help focus their thinking or expectations? I'll jump in real quick, and we don't have anybody asking for it yet. So, you know, if they just have their email, they're happy. So we haven't got to that point yet here. I think some of the strategies are to, I think it's our job to make people understand that AI, again, it's not plug and play. You can't just slap it in and it starts doing things for you. I think people need to understand that this is something that has to be trained and you have to know what you want. What do you expect from AI? Again, people say it all the time, just put it in AI. What does that mean? What is it that you want AI to do? What are the functions that you want AI to perform? I don't think people think about it in that manner because they believe that it is just plug and play, just add AI. But that, you know, that is such a blanket and nonsensical statement that it's difficult to get people to understand. So I think getting them to focus on the fact that this is a learning module, it has to learn how to function. And we do have to feed it data in order to do that. So it's not like we can just plug it in and immediately it starts running. It takes a lot of time. And I know people don't understand that because AI is everywhere and they see it just happening overnight. They think you can add AI into your daily workflow and immediately it's there and it's working, but it's not that way. And it's very difficult to get non-technical people to understand that this has to be trained. It is a program that has to be written. And maybe we're not physically typing in that information like we would if we were writing, you know, a script, but it does have to be created and it takes time and it takes a lot of direction from the person that's requesting this. Again, what are your expectations for AI? And then explain to them how that works. Yeah, I think that Derek said it well, when speaking about and to AI, getting specificity is key. Kind of what you were saying, Cynthia. What is the final outcome that someone is looking for? Because I feel like a lot of people might just think like you're saying, Cynthia, you just put it in and you can, suddenly your job is done for you. And it's definitely not always the case. I'll jump on that real quick. We've been trying to, and I won't be embarrassed to say that we're still on a Windows 2012 domain controller and we've been trying to get it over to the latest version and we've tried three times and it's not worked. And we've used AI and it said, oh yeah, you need to do this, you need to do this, you need to do this. But we've reached a point sometimes with AI where we're just, we're stuck, we're getting in a loop. And sometimes it doesn't have the final outcome. We still have to rely on our, you know, well, I say spice heads, but you know, the community out there to get past these stuck points. You know, AI has been great for us, but it hasn't quite given us the final outcome on several things. Yeah, yeah, for sure. It definitely will take a lot of work. I think Nate also said something really helpful too, preparing a cost benefit analysis would also be helpful in deciding if it makes business sense. Yes, but in order to do that, you need to know what is desired from the client. And that can be very difficult to get from the employer to find what, like Derek said, what is the final outcome you're looking for? So pulling that information out of somebody and making them understand is difficult, but you have to have that information before you can even do the cost benefit. You don't know if you're looking at a $2,000 LM or if you're looking at a $20,000 LM until you have all the information. It's difficult. Yeah, yeah, for sure. What are some like quick wins per se that someone could get with AI that might not have anything in place right now? You know, I know it's not plug and play. I know it's a little bit more difficult to set up than you would think, but are there any ways that somebody could use AI right now that? Let's write those professional emails. Let's you write your email and now slap it into an AI and get it rewritten so that you sound like, you know, that you sound professional. I think that's a big thing. Yeah, not too professional because you don't want to sound like a bot either. But I think helping with writing documentation, helping writing emails, I think that's a wonderful way to start using AI. Yeah, I agree. It works really well. Sometimes we have to send out really important emails. I do use that, but sometimes it's overexcited about, oh my God, I'm looking forward to working with your team and the things I wouldn't normally say. So yeah, you still have to go back in there and just kind of pull it back just a little bit. But I do want to stress when you're writing emails and you're doing documentation, you also have to be aware of the LM that you're using because like I said, you want to be careful what you're putting into AI because depending on what AI you're using, where is that information going? So that is very important when using an AI. You really, and again, I know you're looking for something easy and that is an easy thing to do. You can slap your email into AI and get it to come out very polished and grammatically correct. But it's also important to understand you don't want to be putting sensitive data in there if you don't know how that LM is being used. Yeah. Yeah, absolutely. Derek also said right now everyone should be using it to learn since it's not going away. The more you interact with different models helps user go from knowing about AI to the understanding of AI. So, and if you're me who likes to name their AIs, I'm always big on make sure you thank your AI because when they get to the point where they think they know better than us, they're going to remember who thanked them for the information. And I, yep, every single time I'm like, can you please do this? Thank you. I still call mine Cortana, so you know. No, everybody's computer just went, yes. I hope not, I'm not Cortana anymore, so. Yeah, so how do you implement new forms of AI without overwhelming your team since it is such a big topic of conversation right now? For us, it's kind of like, it's like any other upgrade. It requires education, that's all. So anything we were adding or upgrading to, we want to learn everything about it so that we know how to use it. So that's kind of the way we handle it. It's just another upgrade. Okay, yeah. Yeah, and like I said, we're just kind of using it for IT right now, but I'm just looking for ways where we can give it to people for, okay, this will make your job easier, but like I said, my biggest challenge is getting past that legacy, getting past the legacy software. Yeah, yeah, absolutely. And like Derek said, using AI to learn about AI. Everybody should be using it to learn about anything, but I tell people all the time when they say, I don't know how to use AI, that I can give you a five minute crash course and then you can use the AI to learn about AI. Yeah, yeah. Well, it's killing two birds with one stone because you're learning how it interacts that way. Yeah, yeah. Well, I'm going to go ahead and share my screen real quick because I kind of put together a presentation, not really a presentation, it's just a few slides with some examples from Spiceworks. Let me try to share this. Oh, can't do that. Nevermind. Google Chrome doesn't have permissions. I feel like I've shared my screen before. Well, I'll go ahead and just kind of talk through them and then I can post it in the lab too after this in the summary. But there were four responses that I kind of wanted to talk a little bit about. The first is IntelliCode auto-completions in VS Code. I had no idea this existed, but in Visual Studio Code, there is something called IntelliCode. And basically what it says from their website is, from the Microsoft website is, IntelliCode whole line auto-completions predict the next chunk of your code based on your current code so far and presents it as a gray text inline prediction, which is, I think that's pretty cool. I use that for emails sometimes and it's interesting that it's in code too. So it kind of auto-completes, so you can kind of fill in the code with things instead of having to write every single thing yourself. So obviously that is something that you're gonna wanna check on and make sure that it is correct before you just blindly accept that. We've all used predictive text and it's gone awry. Yes, that is true. I think Derek also, co-pilot chat is available with licensed Microsoft users. This is a great start at little or no cost. However, I'm a big believer that a governance plan needs to come first. So employees know what is allowed to be ingested and what is not. Definitely really important. I think the thing that I'm getting the most from this is you have to be as specific as possible, whether it's with AI itself or with users, you have to be specific about what goes into AI and what comes out. So, but yeah, co-pilot is great. If you're already using Microsoft, I think that is a great addition. Somebody also mentioned a new platform that I hadn't heard of. It's called Clod. I don't know if y'all have heard of Clod. Okay. Do either of you use it at all? Not yet. Okay. I looked it up and it seemed like it was a pretty helpful platform, but basically you use it with your, you can use it in your command line. You can use it in different tools. You can use it for documentation. This specific user also uses it to summarize log files and it does data analysis. So I think those are all pretty cool pieces that you can use to kind of summarize your files and things like that. Also, sorry, my cat is up here and I'm trying to make sure he is not distracting anyone. But yeah, so it seems like a really cool tool to be able to summarize everything and check on your code. I'll go ahead and drop it in here as well. Yeah. Yeah, I'm like trying to get him to stop doing whatever he's doing. And I guess this is one of the benefits of working from home or not so much benefits. But the next one is something that I think that y'all talked about a little bit and it's kind of summarizing tickets and having a virtual service agent that can answer those smaller questions. And this person specifically using it in Jira service management, but I'm sure that there are other ticketing systems and things like that, that have kind of those AI capabilities. But again, it's something that this user kind of noted that you should double check it before taking what it says as fact, because they can't get things wrong. And then the last one, have y'all used Notebook LM, the Google Notebook app? So this one is basically like a research assistant. So you can put in all of your documents So you can put in all of your documents, and then it'll kind of summarize things, and then it'll kind of... but also connect documents together. So you can, if you have like specific findings, they can, that connect between your documents, it can kind of help you get those insights and answers and things like that. So that's another one I can definitely drop in here as well. Let me grab that link. I think documentation is one of those topics that has a lot of, a lot of potential for AI, because we don't always have time to go through every single document to try to find the answer that you need. And I think what's cool is that this one will actually show you the source of the answers rather than just telling you the answer, and you kind of just have to take that as fact. It'll show you the source of it. So you can kind of see the text. But yeah, those were the ones that I found on Spiceworks that I think would be really cool for people to be able to use on a day-to-day basis. But do you all have any specific other recommendations for AI tools? I have to give a big shout out again to GitHub Copilot. I paid the extra $20 for that, and it's been very valuable. I would highly recommend that, especially if you're a small shop and you need to do some development. I can't wait to see where it goes, because it's really been helpful for me. Yeah, I'm a fan of GitHub in the IT realm, for sure. I think there's just different LMs, different AIs, depending on what it is that you need. If I'm going to do something, you know, IT-related, GitHub. If I'm going to do an email, probably ChatGPT. If I'm going to, you know, fix my car, I'm probably going to use Grok. I don't really use Microsoft so much. I really don't use it at all, personally. But, I mean, so there's an AI LM for everything that you can do, but some are more geared towards certain things, like I said. So it's not a one-size-fits-all solution. Yeah. Yeah. I have not seen this GitHub Copilot, and that's someone that is trying to learn C Sharp. I might look into that. I do want to mention, we have been, you know, we have a couple of subscriptions for our employees for IT courses. And I have seen a lot out there now that are very AI-generated, yet they're basically like learning in a classroom live, where instead of a video or something, it actually will help you create a lab. You can ask it questions. And I think that that is an excellent, excellent use of doing AI for learning, because it is interactive. And it might not be able to answer every real-world scenario, but it's so far beyond watching a video to learn something that you have to constantly pause, follow along with. It's really, I think, changed the education landscape exponentially. Yeah. Yeah. Absolutely. Derek said, especially with Copilot Chat right now, it's already integrated into the Microsoft Multiverse, and it can reference to SharePoint, Outlook, OneDrive, things like that, which I think is really cool. So something to consider is to look at the tools that you already have, and what AI kind of integrates with that, because that will also help determine what AI you might use, what platform you might use. And I've used Adobe's AI a couple times, and it's actually been quite impressive. I haven't got really deep into it, but if you're using Adobe products, it's pretty good. Yeah. Okay. Good to know. I've never used it. I see Jesse here mentioned using it for a legal matter, and I, too, used it for a legal matter a couple of months ago, where I literally uploaded all kinds of court documents and was blown away at the response. I mean, you know, court documents are going to be answered differently based on what state you're in, and it took all of that into account and gave me the documentation even to respond, not just the verbiage, but the documentation to use. And, of course, I did cross-check it with an attorney, and they were like, wow, you don't need us. It's too bad AI can't show up in court with you. So, yeah, it was pretty incredible. Wow. It is wild how many things that you can use AI for. Tyler mentioned, and it's something we did touch on, but something important to be aware of is data protection. Chat, QPT, Copilot, and Claude have the ability to turn off data sharing in training, and training for Copilot is on by default, including chat, which is now included in all 365. Meanwhile, Gemini, Grok, and DeepSeek do collect and use your data for training. Yeah, so I did touch on that, where you do want to be mindful of what you're putting in and how data is being used. I think that's very important, and people forget that, you know, especially when you're doing it online. One, you're doing it online, so we're always telling people to be careful of the data that you're posting. But now, even more so, where is that data going, and how is it going to be used afterwards? So, like Tyler is saying, they're using it for data training, but it's not far-fetched to think, if I put in my social security number, and somebody asks a question about social security, is it going to pull up my data? Now, normally, that doesn't happen. It's blocked from doing that. There are certain things that, you know, AI is not going to do for you, especially when you're on these public language models, but it could happen easily. Yeah, yeah. And what if the data itself that's being stored were to be breached? Well, everybody's just pouring all of their life information into AI. You know, you have to think about that. Yeah, absolutely, and especially in different industries that have different privacy requirements, like you mentioned HIPAA and things like that. Important to know what you can and can't put, and what you shouldn't put into AI. And we got a question, why don't you use Microsoft Copilot? That's mostly because we're not a 365 house, so that's why. Yeah, yeah, it's one of those things that I think that you would mainly use it if you were, you have that platform, and it's all of the different 365 tools, but if you don't have a bunch of 365 tools, it's... Right, all of our 365 tools are, they're, you know, they're kind of piecemeal because we're not a 365 house. So sure, you know, we use Office, I use Outlook, we use Zoom, we use Teams, but they're not all integrated into everything that we do. So that's not our, we use those when we have to. They're not our go-tos. Yeah. All right. Well, thank y'all so much for jumping on. We're right at time, so I'll go ahead and close it out, but really, I really appreciate y'all being on. Thank you for all of the conversations and chat as well, and then I'll go ahead and get a wrap-up posted. I'm hoping to be able to get that done today with all of this information and links to all the different resources, and feel free, anybody here is welcome to comment on that and add their own resources that I might not have covered as well. And then we will be back on July 8th. We're going to be talking with a technical recruiter at Ninja One to discuss how to interview for technical roles and how you can find your next team member. But I think that's it for me today. Thank y'all again for joining, and we'll catch y'all in the next one. All right. Thank you. Take care. Bye, y'all.
