1 00:00:00,000 --> 00:00:04,480 I saw me the other day and it made me think so. 2 00:00:04,480 --> 00:00:06,440 I can't take full credit for this, 3 00:00:06,440 --> 00:00:11,280 but in the same vein as Beauty and the Beast, 4 00:00:11,280 --> 00:00:15,440 the house staff is turned into household items, 5 00:00:15,440 --> 00:00:18,320 which made me continue to think 6 00:00:18,320 --> 00:00:22,880 that what household item I would be turned into. 7 00:00:22,880 --> 00:00:26,880 So probably have two. 8 00:00:26,880 --> 00:00:32,360 My first one would probably be just a bog standard coat rack 9 00:00:32,360 --> 00:00:35,800 because I'm a tall person. 10 00:00:35,800 --> 00:00:48,480 So probably a coat rack or an old stained Mr. Coffee machine. 11 00:00:48,480 --> 00:00:54,480 You haven't really cleaned it because it just makes a cup of coffee, 12 00:00:54,480 --> 00:00:58,760 like a pot of coffee every couple days. 13 00:00:58,760 --> 00:01:03,040 And the pot hasn't been cleaned a little while. 14 00:01:03,040 --> 00:01:11,280 So those are probably my two that I think about that I would be. 15 00:01:11,280 --> 00:01:13,600 I really like your reasoning for the first one. 16 00:01:13,600 --> 00:01:14,120 Yeah, thank you. 17 00:01:14,120 --> 00:01:14,880 It's great. 18 00:01:14,880 --> 00:01:17,320 It's great. 19 00:01:17,320 --> 00:01:19,440 Say, Ian, what about you? 20 00:01:19,440 --> 00:01:23,720 I think I like the coffee maker idea a lot. 21 00:01:23,720 --> 00:01:29,080 Maybe newer, but also always on the move. 22 00:01:29,080 --> 00:01:35,960 It's continue brewing to support the people in the household being alive. 23 00:01:35,960 --> 00:01:40,720 Here, this might be an even an espresso maker or something like that. 24 00:01:40,720 --> 00:01:44,920 Here, we're going to go in a little extra fancy, a little extra strong 25 00:01:44,920 --> 00:01:46,520 caffeination. 26 00:01:46,520 --> 00:01:47,320 Like that. 27 00:01:50,280 --> 00:01:53,320 I think mine isn't going to be in the house. 28 00:01:53,320 --> 00:01:57,200 It's in the garage because that's primarily where I live. 29 00:01:57,200 --> 00:02:01,600 And it would either be a tape measure or a level. 30 00:02:01,600 --> 00:02:05,680 Because I'm always running around the house doing house projects, 31 00:02:05,680 --> 00:02:08,160 doing a ton of remodeling and stuff. 32 00:02:08,160 --> 00:02:09,560 So that's where the tape measure comes in. 33 00:02:09,560 --> 00:02:11,680 It impacts every single room. 34 00:02:11,680 --> 00:02:15,120 And then the level because as I'm doing all these house projects 35 00:02:15,120 --> 00:02:18,720 and hammering, all my pictures keep getting slanted. 36 00:02:18,720 --> 00:02:22,360 And I hate pictures that are skewed. 37 00:02:22,360 --> 00:02:25,920 Yeah, it doesn't bother my family at all. 38 00:02:25,920 --> 00:02:30,160 But I've got the big four foot level and I put pictures all nice and lined up. 39 00:02:30,160 --> 00:02:33,640 And so yeah. 40 00:02:35,320 --> 00:02:38,400 I can see with the tape measure, you could be like whimsical and being like, 41 00:02:38,400 --> 00:02:44,480 let's see how far I can go out before I break that game that everybody and every kid plays. 42 00:02:44,480 --> 00:02:47,520 Right. Like, let's see how far I can go. 43 00:02:47,520 --> 00:02:50,320 And you're just like, that's the entertainment for the night. 44 00:02:50,320 --> 00:02:53,080 Like, it's going for the record again. 45 00:02:53,080 --> 00:02:54,760 Let's see. 46 00:02:54,760 --> 00:03:01,320 So yeah, I've got young kids and so I have a little baby tape measure as well. 47 00:03:01,320 --> 00:03:07,320 And so sometimes my son and I will just go measure things around the house together. 48 00:03:07,880 --> 00:03:10,480 And that's the entertainment for about five minutes. 49 00:03:10,480 --> 00:03:14,080 So my son has a. 50 00:03:14,760 --> 00:03:19,120 We borrowed one for our fence, a 100 foot tape measure. 51 00:03:19,120 --> 00:03:21,120 So we have. 52 00:03:21,840 --> 00:03:23,120 Run that around the house. 53 00:03:23,120 --> 00:03:26,000 So but Ariel or Kelsey. 54 00:03:26,640 --> 00:03:28,080 Yeah, Kelsey, what's yours? 55 00:03:28,640 --> 00:03:33,800 I know it's hard to think of the top of your head, but I think I would probably be one of those little music boxes, right? 56 00:03:33,800 --> 00:03:35,800 That a is slightly annoying. 57 00:03:35,800 --> 00:03:37,440 B always seems to be broken. 58 00:03:37,960 --> 00:03:42,640 But C is also like cute and like right on like, I love music boxes. 59 00:03:42,640 --> 00:03:47,240 I'm like, but I feel like there's like enough like little jakey things where I'm like, 60 00:03:47,240 --> 00:03:50,000 I'm danger prone enough to make that make sense. 61 00:03:51,120 --> 00:03:54,760 But also it's just a fun household object that if you had to be turned into something, I'm like, 62 00:03:54,760 --> 00:04:00,120 hey, that would be something fun to be that still kind of looks a little bit humanoid because I'm thinking of a level and going. 63 00:04:00,520 --> 00:04:04,640 So is your face like you just lay there and move your eyes or like. 64 00:04:04,640 --> 00:04:05,240 Yeah. 65 00:04:06,240 --> 00:04:06,560 Yeah. 66 00:04:06,560 --> 00:04:12,040 So I might have been a little bit more methodical there and been like, OK, that might give me some movement. 67 00:04:12,560 --> 00:04:16,920 So you're saying you're cute, a little annoying and a little broken. 68 00:04:16,920 --> 00:04:18,240 Yeah, pretty much. 69 00:04:18,240 --> 00:04:20,840 The cute thing is to bring a little, knowing little broken. 70 00:04:20,840 --> 00:04:21,840 That one's by conferred. 71 00:04:21,840 --> 00:04:22,360 Yeah. 72 00:04:22,360 --> 00:04:23,360 I love it. 73 00:04:24,120 --> 00:04:36,920 I think I also liked the coffee maker, but I think I think I want to be I would be one of those old like coffee grinders, you know, the ones you'd like to do by hand. 74 00:04:36,920 --> 00:04:37,920 Oh, yeah. 75 00:04:37,920 --> 00:04:44,920 But I'm just like the out of date, never used item in the corner who's like always complaining about. 76 00:04:44,920 --> 00:04:46,920 They knew coffee grinders. 77 00:04:46,920 --> 00:04:47,920 What are they doing? 78 00:04:47,920 --> 00:04:51,920 You just plug them in and you push a button to the corner. 79 00:04:52,920 --> 00:04:53,920 What are these? 80 00:04:53,920 --> 00:04:56,920 Oh, there's curing. 81 00:04:56,920 --> 00:04:57,920 What is that? 82 00:04:58,920 --> 00:05:00,920 Hey, Mike, back in my day. 83 00:05:00,920 --> 00:05:09,920 Yeah, my my grandpa has one mounted his kitchen cabinet and it's the the mason jar with the cast iron top and. 84 00:05:09,920 --> 00:05:10,920 Yeah. 85 00:05:10,920 --> 00:05:11,920 Yeah. 86 00:05:11,920 --> 00:05:15,920 I love how he just became the sassy older appliance. 87 00:05:15,920 --> 00:05:16,920 Yeah. 88 00:05:16,920 --> 00:05:17,920 Yeah. 89 00:05:17,920 --> 00:05:18,920 Out of date. 90 00:05:18,920 --> 00:05:20,920 But they don't make them like that. 91 00:05:20,920 --> 00:05:21,920 Right. 92 00:05:21,920 --> 00:05:23,920 Like it's like the plain next lesson. 93 00:05:23,920 --> 00:05:24,920 Right. 94 00:05:24,920 --> 00:05:26,920 They don't make them like me anymore. 95 00:05:26,920 --> 00:05:27,920 Yeah. 96 00:05:27,920 --> 00:05:35,920 So I have always said CIT's coffee grinder up in the front office here is, you know, an electric burr grinder. 97 00:05:35,920 --> 00:05:40,920 I have wanted to get rid of that thing because it grinds so stick and slow. 98 00:05:40,920 --> 00:05:42,920 You know, there's faster ones. 99 00:05:42,920 --> 00:05:46,920 And like you said, you know, the these new ones, you just push a button and let let it go. 100 00:05:46,920 --> 00:05:52,920 And when I I wanted to get that one with a nice one that, you know, can grind it in 30 seconds rather than two minutes. 101 00:05:52,920 --> 00:05:59,920 But maybe I'll go swap it out with whatever your burr grinder or old one hand crank. 102 00:06:00,920 --> 00:06:05,920 Because the only person to blame for the speed on that is yourself. 103 00:06:05,920 --> 00:06:06,920 Yeah. 104 00:06:06,920 --> 00:06:11,920 And I mean, Andrew was saying that he was doing pull ups every time he goes into his office. 105 00:06:11,920 --> 00:06:14,920 He should be ripped and he's going to be our coffee grinder. 106 00:06:14,920 --> 00:06:15,920 There you go. 107 00:06:15,920 --> 00:06:18,920 No, it's a pro. 108 00:06:18,920 --> 00:06:19,920 Progressive. 109 00:06:19,920 --> 00:06:24,920 If I start doing like two or three every time, then yeah, we'll start talking about it. 110 00:06:24,920 --> 00:06:26,920 This is halfway in between. 111 00:06:26,920 --> 00:06:28,920 We had one for a long time. 112 00:06:28,920 --> 00:06:32,920 It's it's like the burr and it's so heavy. 113 00:06:32,920 --> 00:06:36,920 It's a kitchen aid, but your jar goes right on top of it. 114 00:06:36,920 --> 00:06:38,920 It's awesome. 115 00:06:38,920 --> 00:06:40,920 But it is to put the picture. 116 00:06:40,920 --> 00:06:41,920 Yeah. 117 00:06:41,920 --> 00:06:44,920 In the description for everybody listening to be like, you gotta see that picture. 118 00:06:44,920 --> 00:06:45,920 Yeah. 119 00:06:45,920 --> 00:06:46,920 It was so awesome. 120 00:06:46,920 --> 00:06:57,920 But also if you lift that little flap while it's going, it's not unlike my but coffee everywhere. 121 00:06:57,920 --> 00:07:02,920 You get to wear it like and it doesn't like occur to you while you're holding it. 122 00:07:02,920 --> 00:07:04,920 You let it go for a few seconds. 123 00:07:04,920 --> 00:07:06,920 Like what's even happening? 124 00:07:06,920 --> 00:07:09,920 Oh, I did that. 125 00:07:09,920 --> 00:07:17,920 Well, you know, speaking of all these wonderful appliances and the beauty in the bees, such a such an amazing support cast. 126 00:07:17,920 --> 00:07:35,920 We've got a beautiful support group here today on our tech for business podcast. Kelsey and myself are joined by Nate, our director of cybersecurity and our quality assurance analyst and GRC specialist and Andrew, our customer strategy advisor. 127 00:07:35,920 --> 00:07:37,920 Boy, those titles. 128 00:07:37,920 --> 00:08:03,920 Today we're talking about AI tools and kind of best practices and this is an amazing team to join us today and I thought maybe it would be easiest to start with just kind of a little overview and maybe like compliance like what is required of us when we're using tools like this when we're integrating these things into our company. 129 00:08:03,920 --> 00:08:16,920 The, the one thing that I maybe want to mention and I've mentioned this on a prior podcast is just the language for artificial intelligence, machine learning, deep learning, everything like that. 130 00:08:16,920 --> 00:08:27,920 So if you're brand new to this, a lot of those terms have been used interchangeably quite a bit over the years, you know, and it seems like vendors like to use artificial intelligence. 131 00:08:27,920 --> 00:08:31,920 If you're a data scientist, you tend to use something more like machine learning. 132 00:08:31,920 --> 00:08:36,920 Or neural networks or neural networks, something along those lines. 133 00:08:36,920 --> 00:08:46,920 On a prior podcast, I mentioned there's a guy, Tan may I forgot his last name. It's like Bakshi. He has a YouTube channel called Tan may teachers. 134 00:08:46,920 --> 00:08:59,920 He did a presentation and I thought he illustrated it really, really well was the concept behind the technologies machine learning, you know, we can all agree that machine learning is what's actually being done. 135 00:08:59,920 --> 00:09:10,920 And it's usually taking algorithms to better predict or calculate things faster than a human can because computers are great at pattern recognition. 136 00:09:10,920 --> 00:09:19,920 Artificial intelligence is more of a user experience concept to where the computer feels like it has artificial intelligence. 137 00:09:19,920 --> 00:09:23,920 On the back end, it's still machine learning at the end of the day. 138 00:09:23,920 --> 00:09:34,920 And so if we mentioned machine learning, artificial intelligence, it's all the same thing. It's computers trying to better predict or help automate some of the work. 139 00:09:34,920 --> 00:09:42,920 Whatever the output is, right? There's a lot of different uses of AI and machine learning today. 140 00:09:42,920 --> 00:09:50,920 For sure. Thank you for that, you know, overview and kind of clarification as we're sort of losing using this language today. 141 00:09:50,920 --> 00:09:57,920 You know, in the marketing team, we have a lot of different tools. I know for cybersecurity, we have different things. 142 00:09:57,920 --> 00:10:09,920 So when it comes to companies using these, what are those like specific, you know, rules that they should be following? 143 00:10:09,920 --> 00:10:22,920 I'd speak for compliance first. Always, always be looking at what you would enter into it to ensure that it's not protected information to start out with. 144 00:10:22,920 --> 00:10:26,920 I think it depends on the type of the tool as well. 145 00:10:26,920 --> 00:10:33,920 Because there's so many, right? Like it's this big AI machine learning bubble. 146 00:10:33,920 --> 00:10:42,920 You know, I think we've all, it kind of got into this like geist with chat GPT, right? Everybody knows what that is. 147 00:10:42,920 --> 00:10:56,920 So that's a text pattern recognition, which is really just you're typing in something. It's guessing what the next word is based on your prompt. 148 00:10:56,920 --> 00:11:12,920 There's so many others that other companies have created and also chat GPT and open AI, the parent company have made where you can upload PDF files or Word documents. 149 00:11:12,920 --> 00:11:17,920 And I think that's where the biggest compliance is. 150 00:11:17,920 --> 00:11:34,920 Just as a work efficiency, you know, and you can correct me if I'm wrong, but just saying, hey, craft an email to, you know, have an introduction meeting about blank service. 151 00:11:34,920 --> 00:11:53,920 There's a whole compliance issue there. But if you're saying, hey, analyze this document or here's an Excel document with customer information, find patterns in it and start pulling, you know, where can we grow our revenue? 152 00:11:53,920 --> 00:12:22,920 That's where that tricky information is because we, I put it in our notes, but these tools are not full fledged box items that you would get with like when you go to Microsoft, they've had word for how many years, but, you know, it's not something that you go out and just, you can just, hey, yep, okay, it's a full fledged product. 153 00:12:22,920 --> 00:12:26,920 We're the beta testers. 154 00:12:26,920 --> 00:12:30,920 They're still learning. They still need to learn. 155 00:12:30,920 --> 00:12:44,920 And the information we provide them is helping the next prompts. So I think that's something to keep in mind when you're utilizing these items is that we're the beta testers. 156 00:12:44,920 --> 00:13:07,920 We're not necessarily going to get the best product as a result. If that is a prompt we're getting, if that's analysis we're getting, but also with providing that information is they're going to use that moving forward in what capacity we don't know. 157 00:13:07,920 --> 00:13:16,920 But that's where that compliance gets sticky is once it's out, it's out. You can't pull it back. 158 00:13:16,920 --> 00:13:29,920 So yeah, and I think that part of that is there's a scale of how it's been implemented thus far. So for example, yes, we see brand new tools. 159 00:13:29,920 --> 00:13:42,920 For example, Microsoft, they're just introducing loop, right? That's a AI component built into the office suite to help better, you know, create better PowerPoints, you know, help with different Excel formulas. 160 00:13:42,920 --> 00:13:51,920 That's more of a beta set of things. And then there's other things that are deeply integrated today that we just don't even realize that they're doing machine learning on the back end. 161 00:13:51,920 --> 00:14:01,920 So for example, you pull up your weather app on your phone and say, what's the weather, right? There's a lot of machine learning to best predict what the weather will actually be today. 162 00:14:01,920 --> 00:14:20,920 We don't look at the weather report and say, wow, this is artificial intelligence, right? Same thing with, you know, for example, Google and Apple on the, as you're typing, you know, on the even the latest version of Apple software, they're introducing more machine learning on the just how you type to better. 163 00:14:20,920 --> 00:14:32,920 Learn what you're going to say next. So that way their predictive text isn't holistic across everyone. It's going to be specific to you to give you the best results. 164 00:14:32,920 --> 00:14:42,920 We don't look at your keyboard and say, wow, this is a great machine learning, right? And so they might still be beta, you know, having you as the tester on the back end and trying to refine it. 165 00:14:42,920 --> 00:14:57,920 But there is a wide gamut of where it's implemented today. To Ann's point, I just wanted to circle back on this in terms of the compliance is data privacy is going to be absolutely critical to this. 166 00:14:57,920 --> 00:15:07,920 And, you know, Andrew, you had mentioned this a little bit as well. But, you know, a lot of these regulations already apply to AI today, right? 167 00:15:07,920 --> 00:15:13,920 Is you still have to protect the data. You still have to audit the data. You still have to manage that data. 168 00:15:13,920 --> 00:15:27,920 AI doesn't change that. It's just the tool that's being used. So, for example, let's go pick something like GDPR, right? There's a ton of regulations, and it's very enforceable as well, right? 169 00:15:27,920 --> 00:15:38,920 You could still use AI with that type of data. But the risk is, is it going to a third party? Is it something that you're developing internally and it never leaves your, you know, environment? 170 00:15:38,920 --> 00:15:46,920 But you're using that to manipulate the data and you can still adhere to all the same regulations over there. That's okay. 171 00:15:46,920 --> 00:16:01,920 But data privacy is critical when it comes to this because we are using so many third party tools, essentially. 172 00:16:01,920 --> 00:16:22,920 And I feel like it's, you know, it's hard to know what to use. And I think like, and let me kind of explain a little bit better myself is, they're interesting tools, and we all want to like try them out. 173 00:16:22,920 --> 00:16:44,920 But it's like, how can we step into that? And my first thought is start small, like look into the sources, but also like, if it's like chat GPT or barred, which is Google's, or a new one called clock, whatever that is, 174 00:16:44,920 --> 00:16:56,920 I feel like you like start small like craft me an email with like no customer information or anything like that like, hey, create an introduction email. 175 00:16:56,920 --> 00:17:08,920 I used it for some like personal like product description like I gave it some stuff for a product that I was looking to sell, and it like gave a product description. 176 00:17:08,920 --> 00:17:26,920 And I think utilizing it for those kind of things to understand the tool, because there are disclaimers on most of these tools, but not all of them. I saw it where it's like, 177 00:17:26,920 --> 00:17:38,920 I need to generate incorrect information. And when I say may it most likely will, because it's predicting it's not looking at. 178 00:17:38,920 --> 00:17:54,920 Okay, this is a good source. It's giving some information, and it's only as good as the information is, it's given. And so it made it very well may be wrong. 179 00:17:54,920 --> 00:18:05,920 I think the the speech kind of recognition behind it makes it feel like somebody is typing back to you. 180 00:18:05,920 --> 00:18:16,920 But it's it's it's an algorithm. And I think that's difficult to to understand sometimes when you're utilizing these tools. 181 00:18:16,920 --> 00:18:26,920 As an example of that. So I'm going to go pick on Tan make and he's again he's brilliant guys 19 and already traveled the world talking about AI. 182 00:18:26,920 --> 00:18:44,920 But one of the things that he did mention was, you know, there's an article where if you'd go put into, you know, chat, you put here one of those other tools saying, you know, how many jobs will I replace, right, you'll see articles saying 500,000 jobs based off of whatever it is 2030 or whatever 183 00:18:44,920 --> 00:19:00,920 the year is. If you go back to the original publication that said that they said 500,000 jobs will be partially impacted by AI in some degree, you know, it'll be incorporated into our workflows, not necessarily that the full job will be replaced 184 00:19:00,920 --> 00:19:18,920 but some tax, tasks of our jobs will have some type of impact. But because all the news reporters have gone in and said 500,000 jobs will be replaced and then that's perpetuates guess what now the AI thinks that 500,000 jobs will be replaced. 185 00:19:18,920 --> 00:19:31,920 And so it's perpetuating that giving you that false information even though that's not what the original study said. So, so absolutely, there's have to be concerns about that. 186 00:19:31,920 --> 00:19:47,920 The one thing this is going to be a little bit of a slight tangent but again, just touching on that compliance stuff again is as we're talking about some of the risks, you know, because we've been talking about data privacy and, you know, Andrew. 187 00:19:47,920 --> 00:20:02,920 You know, in putting other info in there that you know maybe starting small is the NIST agency has put out a AI risk management framework. So they really suspect in January. 188 00:20:02,920 --> 00:20:18,920 It's very entry level at the moment. But what they do is they basically say, you know, what are the main ways and considerations that we need to think about when implementing AI. 189 00:20:18,920 --> 00:20:31,920 And they talk about three different components is how can it potentially harm people itself. So maybe that's, I don't know, disinformation, right, you know, with the whole, by the way, your jobs are going to be replaced. 190 00:20:31,920 --> 00:20:43,920 How can harm the organization, you know, maybe there's false data and it's going to predict something that isn't true. And then we base off all of our decisions on that. 191 00:20:43,920 --> 00:20:55,920 And then, for example, how does it impact the ecosystem? I don't have a great example of that. But how does it maybe incorporate into other tools and then the tools potentially rely on that data to do another action or something. 192 00:20:55,920 --> 00:21:06,920 So taking a look at those three is going to be pretty important, at least, you know, again, how can harm people? How can it harm your organization? How can it harm the ecosystem? 193 00:21:06,920 --> 00:21:21,920 And then they do give three topics to consider on how to manage this risk. And again, they don't have a great detailed steps on everything at the moment, but they're starting to build that framework is identifying what the risks are. 194 00:21:21,920 --> 00:21:35,920 So you have to first do that, then how do you start measuring them and tracking them. And then the last one is how do you manage that risk and actually acting upon those as the risk levels and everything is changing. 195 00:21:35,920 --> 00:21:50,920 How do you stay on top of it? So that's what they're starting to say. And then I'm excited for the next revision when it comes out, because they tend to expand on those quite a bit more. 196 00:21:50,920 --> 00:22:08,920 And we'll make sure to put that link in the description. And I kind of want to ask you, do you think these, we're talking about these tools and we're talking about compliance and NIST and all these things. 197 00:22:08,920 --> 00:22:21,920 Do you think your conversations with customers are going to change or is it the same conversation when you're talking about data privacy and all of these things they should already have in place? 198 00:22:21,920 --> 00:22:40,920 It is interesting that you say that because we did just last week have someone ask, how are we going to address putting this in our policies? And the answer is, like addresses some of what Andrew said, what Nate said, this is not new. 199 00:22:40,920 --> 00:23:00,920 This is really a good battering different, but really being as educated as possible and putting some strict as needed guidelines on what you should be using being educated of an awareness of where these things already exist. 200 00:23:00,920 --> 00:23:18,920 It's not just because chat GPT is here and we can type something in it and there's something magical that comes out. We've had something like Grammarly, that's been something that's been asked a lot over the years. 201 00:23:18,920 --> 00:23:40,920 How do we stop people from doing this to predict how these, this should be said better? And it kind of frightens me because these are people that are either financial industry or healthcare industry and wanting, I want to know, but I don't want to know how they know that they're using this. 202 00:23:40,920 --> 00:23:53,920 Is this on a case file? Is this on a mortgage documentation? Is there a better way to say this and clearly guide where those things can live to Nate's point? 203 00:23:53,920 --> 00:24:17,920 It's not necessarily saying don't use them, but you're identifying what data you absolutely need to protect and where those should live and what, if there is any specific elements that you should be aware of within those existing applications or tools, that you can, 204 00:24:17,920 --> 00:24:41,920 I did, long story too long, the biggest thing is going to be education with unfortunately the policies devil into the repercussions to you can't you can't just go do a rain of mortgage on the on a AI site and hope that it's okay when you've now put all of your customer data in for a mortgage. 205 00:24:41,920 --> 00:25:04,920 Well, yeah, like you see it with I'm sure with in the education industry or an education field where you type in like I need a, you know, three paragraph essay on bail wolf chapter three, right, and they just, they just take it copy paste hey look I'm done, but they don't review what it was given. 206 00:25:04,920 --> 00:25:16,920 There's tools now that have been, you know, crafted for this for the education industry to be like, is it AI. So you can't just do that. 207 00:25:16,920 --> 00:25:45,920 They've always had like plagiarism tools, but there's actually a famous. I follow somebody on YouTube. He went over it, but they submitted a legal brief that cited a non existent case, because they put it into one of these chat generators or text predictors, and they didn't review it they said hey look my brief is done. 208 00:25:45,920 --> 00:25:57,920 And submitted it to the court and very possibly could ruin their career as a lawyer. And so, you know, take that as a warning. 209 00:25:57,920 --> 00:26:12,920 We, you know, it takes some examples for for most people to learn but again, it's always, it's like monitored freedom you could say is like. 210 00:26:12,920 --> 00:26:23,920 Okay, we can play with these things but make sure that we're getting back the appropriate, you know, information that we're looking for. 211 00:26:23,920 --> 00:26:31,920 I kind of go into it with a goal of this is what I want this tool to do. 212 00:26:31,920 --> 00:26:35,920 I'm going to try and get that but if I don't get I'm not going to. 213 00:26:35,920 --> 00:26:45,920 Like it's not like a end all be all like can't eat like okay I'm gonna have to use it no like if it's not right don't use it. 214 00:26:45,920 --> 00:26:47,920 It's a tool. 215 00:26:47,920 --> 00:26:59,920 Right we are talking about Nate having a level and a tape measure and you know if your tape measure or your level is off, you're not going to use it. 216 00:26:59,920 --> 00:27:08,920 So, I think just being very conscious of what is coming out of it. 217 00:27:08,920 --> 00:27:17,920 That kind of skeptical mind is always great to have especially as we get these new tools. 218 00:27:17,920 --> 00:27:25,920 You know security that's the biggest thing right is is is having that that skepticism of is this fishing. 219 00:27:25,920 --> 00:27:31,920 Is this really the CEO asking for six Apple. 220 00:27:31,920 --> 00:27:34,920 Apple store gift cards. 221 00:27:34,920 --> 00:27:48,920 So they can get out of Libya like it's right like it and it's we've always had to adapt so that's kind of just my thought on, you know, be have skepticism. 222 00:27:48,920 --> 00:27:51,920 Try and be. 223 00:27:51,920 --> 00:27:56,920 Embracing of this new technology but also like. 224 00:27:56,920 --> 00:27:57,920 Be careful. 225 00:27:57,920 --> 00:28:02,920 There's a lot to it. 226 00:28:02,920 --> 00:28:05,920 Yeah, the. 227 00:28:05,920 --> 00:28:18,920 I guess I had two thoughts here in terms of your you know gift card examples and having to do that right here at CIT we're even using tools that are using the machine learning artificial intelligence to better analyze those emails. 228 00:28:18,920 --> 00:28:34,920 So even if it does look like it's coming from a legitimate customer and you know maybe it's a typo scouting or you know which is essentially a domain that's similar to yours but someone's modified it just to try and intercept that conversation make you start talking to the wrong person. 229 00:28:34,920 --> 00:28:38,920 And then that's what's called a conversation hijack. 230 00:28:38,920 --> 00:28:45,920 Where then they start taking over and trying to set up something like a fake wire transfer gift cards something along those lines. 231 00:28:45,920 --> 00:28:54,920 We're using tools in place today in addition to your traditional spam filters to help identify those types of threads. 232 00:28:54,920 --> 00:28:59,920 But then what that does is it also gives the end users the capability of saying. 233 00:28:59,920 --> 00:29:10,920 Let me use the call on the tool and it will go analyze and say by the way this is unique to your location we only see one or two people talking to that and it pulls all these different metrics together. 234 00:29:10,920 --> 00:29:24,920 It puts it all into a nice package digestible thing for the end users to say we know that that is phishing or you know spam or solicitation or something along those lines. 235 00:29:24,920 --> 00:29:28,920 But it's artificial intelligence that's doing that on the back end. 236 00:29:28,920 --> 00:29:40,920 And then that kind of leads me to probably my closing thought is the tools are coming. You can't necessarily say I'm going to block all artificial intelligence today. 237 00:29:40,920 --> 00:29:47,920 You know and I know that on some of the policies you're getting asked that how do we block this all together. 238 00:29:47,920 --> 00:29:49,920 You can't really do that. 239 00:29:49,920 --> 00:29:58,920 All the manufacturers are integrating as core functions today. You know I talked about it as the keyboard or the weather app. 240 00:29:58,920 --> 00:30:02,920 It's getting further into you know the Microsoft ecosystem. 241 00:30:02,920 --> 00:30:15,920 You know we could extend that out. There's a lot of different solutions that Adobe Photoshop right there in implementing new things where you can say change the sky to sunset and it'll do it for you right. 242 00:30:15,920 --> 00:30:22,920 It's phenomenal stuff. You can't just say well now I have to block all those tools. 243 00:30:22,920 --> 00:30:26,920 Kind of I think the summary for me is again it's all about the data privacy. 244 00:30:26,920 --> 00:30:42,920 If you have some type of sensitive info know what's going into it know where it's going and then maybe there's not sensitive info but then it's the how do I take that output and monitor the use of it. 245 00:30:42,920 --> 00:30:49,920 So for example Andrew's legal example there right is if I'm just going to spit out a legal summary. 246 00:30:49,920 --> 00:30:56,920 I could get in trouble for that. There's tools that like to help develop scripts for you know DevOps. 247 00:30:56,920 --> 00:31:07,920 Don't blindly run those scripts validate them test them in a sandbox environment before you introduce that into production. 248 00:31:07,920 --> 00:31:17,920 So as we kind of close out here I want to give an and Andrew and made again if there's anything you want to if you were. 249 00:31:17,920 --> 00:31:19,920 Yeah, you're sitting in your round two. 250 00:31:19,920 --> 00:31:36,920 If you're sitting in front of a customer today and they were sort of lost on on what to do next and action step or how to implement or what to do you know what what would you tell them would be one of those first things that you would say to them. 251 00:31:36,920 --> 00:31:42,920 First thing I would say is what what are they trying to get out of the tool. 252 00:31:42,920 --> 00:31:46,920 I think they'll narrow down what tool it is. 253 00:31:46,920 --> 00:31:56,920 And then we can say okay just like anything else if it's a CRM or inventory management whatever that tool is. 254 00:31:56,920 --> 00:31:59,920 It has a specific purpose. 255 00:31:59,920 --> 00:32:10,920 So that's probably where I would start the conversation as somebody who may start these conversations with customers is. 256 00:32:10,920 --> 00:32:13,920 What do they want out of the tool. 257 00:32:13,920 --> 00:32:19,920 Trying to find that best tool do research on it and then help them through that process. 258 00:32:19,920 --> 00:32:34,920 And engaging Nate and and and the security team of okay they found this tool. Let's look at the back end just like we would for any tool of where is our data going. 259 00:32:34,920 --> 00:32:39,920 Where is it stored. Is this company using this data in other forms. 260 00:32:39,920 --> 00:32:44,920 So that's where I would start the conversation is. 261 00:32:44,920 --> 00:32:46,920 What do you want out of it. 262 00:32:46,920 --> 00:32:48,920 So. 263 00:32:48,920 --> 00:32:52,920 I see this as a. 264 00:32:52,920 --> 00:32:55,920 I said education before but. 265 00:32:55,920 --> 00:32:59,920 We had a lot of I would say not even five years ago. 266 00:32:59,920 --> 00:33:09,920 Customers that when you said taking things to the cloud they're like we don't want our cloud. No cloud what. No, I don't want any part of that. 267 00:33:09,920 --> 00:33:22,920 And then trying to start to explain what that really is educating yourself on your relationship with your customer on their relationship and understanding of their data and what they have. 268 00:33:22,920 --> 00:33:26,920 And then telling them more in education wise. 269 00:33:26,920 --> 00:33:28,920 I don't see this as much different than that. 270 00:33:28,920 --> 00:33:33,920 I really don't. Do we need this read and then to dovetail on Andrew. 271 00:33:33,920 --> 00:33:47,920 Do we need this resource. What are the what are the controls associated. Do we need to put data in that. Do we have an existing resource. Do we want to prohibit it for our use of certain applications. 272 00:33:47,920 --> 00:33:54,920 And just really being having that education and awareness around your your systems and your data. 273 00:33:54,920 --> 00:33:58,920 I will be the best first step in my mind. 274 00:33:58,920 --> 00:34:04,920 Yeah, like said round to I have one more. Yeah, the. 275 00:34:04,920 --> 00:34:10,920 The one thing that I'd ask business leaders to ask themselves is what are you looking to protect. Right. Is it. 276 00:34:10,920 --> 00:34:15,920 PII is it a system. Is it intellectual property. Is it. 277 00:34:15,920 --> 00:34:17,920 Maybe. 278 00:34:17,920 --> 00:34:23,920 You know board reports something along those lines that you don't want out. 279 00:34:23,920 --> 00:34:32,920 In this environment today from there. That's the first step of that high risk management framework is identify what you have today. 280 00:34:32,920 --> 00:34:39,920 Then at that point you can start protecting it with policies procedures technology something along those lines. 281 00:34:39,920 --> 00:34:44,920 But you have to know what you're trying to protect in the first place. 282 00:34:44,920 --> 00:34:53,920 I love it. This is such a good group. You know three very different teams and we got some great information. So thank you. 283 00:34:53,920 --> 00:34:59,920 Nate and Andrew for joining us today. If you enjoyed this podcast please like subscribe. 284 00:34:59,920 --> 00:35:03,920 If you have a question or you want us to come back to this topic. 285 00:35:03,920 --> 00:35:15,920 Please reach out to us at info at CIT dash net dot com or head out to our website CIT dash net dot com slash podcast and we'll be back next week with an all new episode.