1 00:00:00,000 --> 00:00:05,000 Technology breakthroughs that are similar to AI, 2 00:00:05,000 --> 00:00:08,440 and the example was like the iPhone. 3 00:00:08,440 --> 00:00:11,680 Can you think of what is given 4 00:00:11,680 --> 00:00:15,920 this similar excitement in the world that you remember? 5 00:00:15,920 --> 00:00:18,800 I've got two actually that I've thought about. 6 00:00:18,800 --> 00:00:21,360 The first one being, I would say, 7 00:00:21,360 --> 00:00:23,160 the James Webb telescope, 8 00:00:23,160 --> 00:00:25,080 I think was a huge breakthrough. 9 00:00:25,080 --> 00:00:27,680 Being able to see infrared light in space at 10 00:00:27,680 --> 00:00:32,880 a mass spectrum I think is absolutely huge. 11 00:00:32,880 --> 00:00:35,880 I think I got a lot of publicity around it. 12 00:00:35,880 --> 00:00:38,240 I mean, it was a 25-year project, 13 00:00:38,240 --> 00:00:39,760 so I would hope so. 14 00:00:39,760 --> 00:00:43,920 But that, and then I would also say, 15 00:00:43,920 --> 00:00:48,760 we just broke through fusion reactions. 16 00:00:48,760 --> 00:00:50,760 In December of last year, 17 00:00:50,760 --> 00:00:55,120 the first successful fusion reaction was achieved in the lab, 18 00:00:55,120 --> 00:00:58,920 which is basically combining two hydrogen molecules into 19 00:00:58,920 --> 00:01:04,080 a helium and then producing a net positive of energy. 20 00:01:04,080 --> 00:01:05,840 Sorry to nerd out about it, 21 00:01:05,840 --> 00:01:09,080 but I'm a science lover. 22 00:01:09,080 --> 00:01:11,040 I think that was a huge breakthrough, 23 00:01:11,040 --> 00:01:14,800 so I've been following that pretty closely. 24 00:01:16,080 --> 00:01:19,400 That's awesome. I'm sitting here going, 25 00:01:19,400 --> 00:01:22,920 that is way more technical in my brain than I was like, 26 00:01:22,920 --> 00:01:24,520 fun technology, but I love that. 27 00:01:24,520 --> 00:01:26,120 Now I feel like I learned something. 28 00:01:26,120 --> 00:01:30,360 Yeah. I guess put it into spectrum. 29 00:01:30,360 --> 00:01:35,720 It's basically what it would be used for is clean energy, 30 00:01:35,720 --> 00:01:39,200 replacing fossil fuels and such to gain a net positive of 31 00:01:39,200 --> 00:01:45,840 energy and replace our basically nuclear reaction, 32 00:01:45,840 --> 00:01:48,040 and reactors and stuff like that. 33 00:01:48,040 --> 00:01:49,840 I don't know if it's nuclear, 34 00:01:49,840 --> 00:01:52,400 but basically cleaner energy, 35 00:01:52,400 --> 00:01:54,640 I would say fossil fuels. 36 00:01:54,640 --> 00:01:57,080 Wow. Yeah, fossil fuels. 37 00:01:57,080 --> 00:02:00,800 It's crazy. Kelsey, do you think you can tap that? 38 00:02:00,800 --> 00:02:06,480 No. Am I writing very strong off of the coattails of iPhone? 39 00:02:06,480 --> 00:02:09,440 Yes, only because I love my iPad Pro. 40 00:02:09,440 --> 00:02:11,760 Yes, I had a first gen iPad back in the day, 41 00:02:11,760 --> 00:02:13,840 but what that did with the pencil with 42 00:02:13,840 --> 00:02:16,760 Procreate for digital art because I could never afford 43 00:02:16,760 --> 00:02:19,480 or a big tablet and a PC setup and I needed 44 00:02:19,480 --> 00:02:23,240 something that I could take in my studio apartment that fit. 45 00:02:23,240 --> 00:02:26,720 So having that iPad Pro with the pencil, 46 00:02:26,720 --> 00:02:30,240 still the best technology I've ever personally used to go. 47 00:02:30,240 --> 00:02:31,240 That's still my favorite. 48 00:02:31,240 --> 00:02:33,200 If I couldn't have anything else, I'd want that. 49 00:02:33,200 --> 00:02:34,560 I agree with that. 50 00:02:34,560 --> 00:02:36,440 Having the iPad Pro with the pencil in 51 00:02:36,440 --> 00:02:41,360 college was an absolute lifesaver to be able to write on the slides. 52 00:02:41,360 --> 00:02:46,120 Huge. I had paper in a pencil. 53 00:02:46,120 --> 00:02:47,160 It was right. 54 00:02:47,160 --> 00:02:50,280 Is that your breakthrough? Is that what you're sharing today? 55 00:02:50,280 --> 00:02:51,920 Paper in a pencil. 56 00:02:51,920 --> 00:02:53,880 It's beating them until my joke, 57 00:02:53,880 --> 00:02:57,560 which was going to be slate and chisel. 58 00:02:57,560 --> 00:03:02,960 I had a sociology professor that didn't allow computers in the school, 59 00:03:02,960 --> 00:03:07,440 provided tablets and so I was using the tablet. 60 00:03:07,440 --> 00:03:11,400 I know I was handwriting notes and he still was calling it out, 61 00:03:11,400 --> 00:03:13,480 going no technology at all. 62 00:03:13,480 --> 00:03:16,120 It's the same as a piece of paper if I'm writing, 63 00:03:16,120 --> 00:03:19,760 I can't writing notes, but this one. 64 00:03:19,760 --> 00:03:22,600 I'm going to write August's science background, 65 00:03:22,600 --> 00:03:25,600 but this is now four or five months now, 66 00:03:25,600 --> 00:03:27,480 but Princeton University, 67 00:03:27,480 --> 00:03:30,360 they just announced that they developed a camera, 68 00:03:30,360 --> 00:03:32,400 the size of a grain of rice. 69 00:03:32,400 --> 00:03:34,480 It's super cool because they take 70 00:03:34,480 --> 00:03:40,400 200,000 micro cones and it's all about the size and the shape and 71 00:03:40,400 --> 00:03:43,520 the direction that all these cones are being pointed. 72 00:03:43,520 --> 00:03:46,520 Now they're able to produce 73 00:03:46,520 --> 00:03:48,120 compared to the last version, 74 00:03:48,120 --> 00:03:50,040 super high resolution images, 75 00:03:50,040 --> 00:03:53,640 but where that could come in super helpful in the medical industry 76 00:03:53,640 --> 00:03:57,320 is these almost microscopic, 77 00:03:57,320 --> 00:03:59,400 I mean, microscopic sign a lot smaller, 78 00:03:59,400 --> 00:04:06,320 but really, really tiny cameras to where you could actually take this and start running it up a vein, 79 00:04:06,320 --> 00:04:11,640 or an artery and get high detailed images in the medical space. 80 00:04:11,640 --> 00:04:14,200 It's not widely used yet, 81 00:04:14,200 --> 00:04:18,480 but I think it could have the power to be something as transformative, 82 00:04:18,480 --> 00:04:21,120 at least for the medical industry. 83 00:04:22,520 --> 00:04:27,320 You really have your finger on the science pulse here. 84 00:04:27,320 --> 00:04:29,360 How about Todd, what was yours? 85 00:04:29,360 --> 00:04:31,920 I know this was your idea to talk about today. 86 00:04:31,920 --> 00:04:32,760 What do you got? 87 00:04:32,760 --> 00:04:36,160 Yeah. Well, I mean, the one I had originally, 88 00:04:36,160 --> 00:04:40,360 and we did another podcast recently and Kyle was talking about the evolution of 89 00:04:40,360 --> 00:04:43,720 technology and how it has really changed the world and that was 90 00:04:43,720 --> 00:04:45,840 where the original thought process comes. 91 00:04:45,840 --> 00:04:50,320 I did use the iPhone example and when I use the example, 92 00:04:50,320 --> 00:04:54,240 I thought it's interesting if you remember back a couple years ago when the iPhone came out, 93 00:04:54,240 --> 00:04:58,200 it wasn't that there wasn't any mobile management like that. 94 00:04:58,200 --> 00:05:01,800 I mean, there was Blackberries and Windows had its version and so forth, 95 00:05:01,800 --> 00:05:07,240 but the iPhone was so dynamically different and such a massive leap forward. 96 00:05:07,240 --> 00:05:09,240 You could just look at that and I'm not an iPhone guy, 97 00:05:09,240 --> 00:05:12,000 I'm an Android dude, but when that came out, 98 00:05:12,000 --> 00:05:13,960 I looked at it and I'm just like, 99 00:05:13,960 --> 00:05:15,960 I'm not going to buy that, 100 00:05:15,960 --> 00:05:20,280 but that's a game changer and that's where I was coming from when I came up with the original idea. 101 00:05:20,280 --> 00:05:22,280 There's tons of examples too. 102 00:05:22,280 --> 00:05:24,440 I mean, I can pivot all over the place, 103 00:05:24,440 --> 00:05:27,960 I can go in a different direction than the science aspect too. 104 00:05:27,960 --> 00:05:31,200 So gaming on PCs was a game changer for me too. 105 00:05:31,200 --> 00:05:36,400 I can't believe how much that changed who I was and how it really sucked me into PCs in general, 106 00:05:36,400 --> 00:05:42,600 but just that consumer level grade technology to me is what really changes the world. 107 00:05:42,600 --> 00:05:47,040 When you can consume it at any given level and compete with big companies, 108 00:05:47,040 --> 00:05:48,960 that's what Ubers did and whatnot, 109 00:05:48,960 --> 00:05:51,920 you've got a different world and the world completely pivots off it. 110 00:05:51,920 --> 00:05:57,000 To me, you just look at those things and go, wow, AI chat, GPT, 111 00:05:57,000 --> 00:05:58,600 those things could potentially do the same thing. 112 00:05:58,600 --> 00:06:01,000 How that's applied, I don't honestly know just yet, 113 00:06:01,000 --> 00:06:02,800 but that's where I was coming from. 114 00:06:02,800 --> 00:06:07,840 Yeah, we did a chat GPT podcast already and during that, 115 00:06:07,840 --> 00:06:14,040 I mentioned the calculator and how that changed and the pushback that it had. 116 00:06:14,040 --> 00:06:16,080 I think in that same vein, 117 00:06:16,080 --> 00:06:20,000 I'd say the printing press had such a huge change when we talk about- 118 00:06:20,000 --> 00:06:21,440 It's such a good poll. 119 00:06:21,440 --> 00:06:23,040 That's so good. 120 00:06:23,040 --> 00:06:29,200 When we talk about how people used to work and how books had to be copied by hand, 121 00:06:29,200 --> 00:06:32,720 obviously there's steps in between copying by hand and the printing press, 122 00:06:32,720 --> 00:06:34,960 but those types of changes, 123 00:06:34,960 --> 00:06:40,520 it's things that can save you time while still producing the same quality. 124 00:06:40,520 --> 00:06:46,560 Now, is chat GPT or any of the other solutions actively at the same quality yet? 125 00:06:46,560 --> 00:06:52,200 Not really, but in terms of saving time and getting us to use 126 00:06:52,200 --> 00:06:56,040 different skills to create the same results, I think it's up there. 127 00:06:56,040 --> 00:06:58,320 Wow, that's great. 128 00:06:58,320 --> 00:07:03,280 I wasn't sure, what came to mind was maybe fits, 129 00:07:03,280 --> 00:07:07,600 but drones was a big one and I think for me, 130 00:07:07,600 --> 00:07:14,160 when Amazon started using it specifically for delivering packages and then these conversations of, 131 00:07:14,160 --> 00:07:16,560 well, what does that mean for workers? 132 00:07:16,560 --> 00:07:20,920 What does that mean for privacy in delivering these packages? 133 00:07:20,920 --> 00:07:23,800 For me, that was the one that came up. 134 00:07:23,800 --> 00:07:25,440 If you're listening on YouTube, 135 00:07:25,440 --> 00:07:27,000 we'd love for you to comment, 136 00:07:27,000 --> 00:07:30,800 what do we miss, what technology breakthroughs didn't we mention? 137 00:07:30,800 --> 00:07:36,360 I know AI is a huge breakthrough that everyone's talking about lately and today, 138 00:07:36,360 --> 00:07:38,320 on our Tech for Business podcast, 139 00:07:38,320 --> 00:07:41,280 we were diving into a discussion about AI. 140 00:07:41,280 --> 00:07:46,560 Kelsey and myself, Ariel, are your marketing team moderators joined by Todd, 141 00:07:46,560 --> 00:07:49,200 our COO and CISO, Matthew, 142 00:07:49,200 --> 00:07:51,200 our GRC analyst, August, 143 00:07:51,200 --> 00:07:55,960 our COC technician and made our director of cybersecurity. 144 00:07:55,960 --> 00:07:57,400 I'd like to put it out there, 145 00:07:57,400 --> 00:08:01,680 who would like to lead us into the weeds of this subject? 146 00:08:02,840 --> 00:08:05,440 Anybody feel like? 147 00:08:05,440 --> 00:08:08,480 How unusual that we didn't start everybody's jumping in, 148 00:08:08,480 --> 00:08:10,680 that's what's going on around here. 149 00:08:10,680 --> 00:08:12,240 Yeah, I'll start. 150 00:08:12,240 --> 00:08:17,400 I think you framed it up or maybe it was just the icebreaker that framed it up as, 151 00:08:17,400 --> 00:08:21,680 what's cool about AI and what's cool about chat GPT in general. 152 00:08:21,680 --> 00:08:24,240 I think Matthew nailed it for me. 153 00:08:24,240 --> 00:08:29,400 I would like to use the same examples of the printing press and the calculators, 154 00:08:29,400 --> 00:08:31,520 and all of that type of technology, 155 00:08:31,520 --> 00:08:33,440 how it's been an enabler for humankind. 156 00:08:33,440 --> 00:08:37,480 It hasn't been something where it's been this massive negativity. 157 00:08:37,480 --> 00:08:38,760 Again, my feeling is, 158 00:08:38,760 --> 00:08:42,040 I don't really know all the options of what you can do with it just yet, 159 00:08:42,040 --> 00:08:44,440 and I guess most people are in the same place, 160 00:08:44,440 --> 00:08:48,200 but it'll start to become pretty clear rather quickly, I would imagine. 161 00:08:48,200 --> 00:08:49,840 I mean, you already seen the push, 162 00:08:49,840 --> 00:08:53,120 Microsoft's talking about releasing what they're calling co-pilot, 163 00:08:53,120 --> 00:08:56,320 which is going to be built into all of the Office 365 apps. 164 00:08:56,320 --> 00:08:58,240 Whether there's going to be a cost or not for that, 165 00:08:58,240 --> 00:09:01,720 we'll see, but it's Microsoft, so of course there will. 166 00:09:01,720 --> 00:09:04,720 But those are the kinds of things that they're talking about. 167 00:09:04,720 --> 00:09:06,400 In case anybody doesn't know what that looks like, 168 00:09:06,400 --> 00:09:10,440 they're talking about how you can use it to tell your device to start to build 169 00:09:10,440 --> 00:09:13,360 a PowerPoint regarding whatever it is that you want to do, 170 00:09:13,360 --> 00:09:16,960 and it can put the shell of a deck together for you, 171 00:09:16,960 --> 00:09:20,360 or it can summarize meetings for you, that kind of stuff. 172 00:09:20,360 --> 00:09:24,280 So to me, I think there's a lot to love about AI, 173 00:09:24,280 --> 00:09:28,440 and we can get into the weeds of what we've already done in the world 174 00:09:28,440 --> 00:09:32,600 that you could probably consider it either a version of machine learning or AI too, 175 00:09:32,600 --> 00:09:34,400 that's already being adopted by people, 176 00:09:34,400 --> 00:09:36,240 so it's not like it's this brand new thing, 177 00:09:36,240 --> 00:09:39,120 but the adoption of it is really starting to take off. 178 00:09:39,120 --> 00:09:43,320 Yeah. I guess one of the things I'd maybe first, 179 00:09:43,320 --> 00:09:46,120 before we jump into like, so the chat GPT, 180 00:09:46,120 --> 00:09:47,760 that's relatively new, 181 00:09:47,760 --> 00:09:50,840 and then now with some of the other developments that we can talk about, 182 00:09:50,840 --> 00:09:55,240 those are all going to be brand new and new iterations and how people are using that. 183 00:09:55,240 --> 00:09:57,440 But if we take a step back, 184 00:09:57,440 --> 00:10:02,320 even 10 years, we're already starting to use some of this already. 185 00:10:02,320 --> 00:10:04,680 So for example, 186 00:10:04,680 --> 00:10:09,720 I think it was 2014 when Apple first introduced Siri on the iPhone, 187 00:10:09,720 --> 00:10:12,240 and if you remember back then, it was garbage. 188 00:10:12,240 --> 00:10:13,600 Still. 189 00:10:13,600 --> 00:10:16,800 Yeah, you upgrade it, it is still garbage. 190 00:10:16,800 --> 00:10:20,960 But back then, when Bayer, 191 00:10:20,960 --> 00:10:22,440 I think it was the iPhone 4S, 192 00:10:22,440 --> 00:10:24,120 that was my first iPhone, 193 00:10:24,120 --> 00:10:28,120 which had Siri and you ask it to, 194 00:10:28,120 --> 00:10:31,280 what's the weather and it was completely wrong. 195 00:10:31,280 --> 00:10:33,800 He'll tell you the time of the day or whatever it was. 196 00:10:33,800 --> 00:10:39,800 But that was some of the first iterations of using the technology to 197 00:10:39,800 --> 00:10:45,480 predict what you wanted it to search for and then deliver those results over to you. 198 00:10:45,480 --> 00:10:49,760 From there, Google in their Google I O conferences has 199 00:10:49,760 --> 00:10:54,920 introduced many new features along the way. 200 00:10:54,920 --> 00:10:56,320 They had things about, 201 00:10:56,320 --> 00:10:58,840 I'm going to let the phone go call 202 00:10:58,840 --> 00:11:02,120 the barbershop and schedule an appointment for you. 203 00:11:02,120 --> 00:11:05,000 There's all these different things that have been introduced. 204 00:11:05,000 --> 00:11:09,720 The fact that we can sit here and I don't know if anyone has an Alexa, 205 00:11:09,720 --> 00:11:13,320 but hey Alexa, what's the weather today? 206 00:11:13,320 --> 00:11:16,640 I hope I just kicked off all your Alexa devices there. 207 00:11:16,640 --> 00:11:22,280 But they're sitting there listening and we're using those tools actively today. 208 00:11:22,280 --> 00:11:25,080 Now we're just seeing it become deeply 209 00:11:25,080 --> 00:11:27,480 integrated with the tools that we're already 210 00:11:27,480 --> 00:11:31,200 using without having to go purchase new hardware. 211 00:11:31,200 --> 00:11:37,680 And starting to really fit into the daily workflows as well. 212 00:11:37,680 --> 00:11:43,680 Yeah, I want to piggyback off of the Siri conversation as well. 213 00:11:43,680 --> 00:11:45,880 I have an iPhone. 214 00:11:45,880 --> 00:11:47,680 No, no, no, not yet. 215 00:11:47,680 --> 00:11:48,680 Not yet. 216 00:11:48,680 --> 00:11:53,480 Siri is not my favorite because it doesn't provide what I need. 217 00:11:53,480 --> 00:11:57,600 But there is an integration that you can connect with Siri using 218 00:11:57,600 --> 00:12:01,920 like iPhone shortcuts to connect it to like chat GPTs, 219 00:12:01,920 --> 00:12:08,080 API that you can basically activate chat GPT by saying 220 00:12:08,080 --> 00:12:12,520 like Siri plus and then it'll activate chat GPT. 221 00:12:12,520 --> 00:12:13,680 And then you can ask it, 222 00:12:13,680 --> 00:12:18,520 like tell me about the topic of AI and I'll give you 223 00:12:18,520 --> 00:12:21,680 a whole summary of basically what you would type on chat GPT. 224 00:12:21,680 --> 00:12:24,960 So I think if Siri doesn't get to that point, 225 00:12:24,960 --> 00:12:27,360 like there is some back way of doing it. 226 00:12:27,360 --> 00:12:30,040 And I think that is monumental. 227 00:12:30,040 --> 00:12:32,960 Like the difference between what do they call? 228 00:12:32,960 --> 00:12:36,520 Is it Bixby for Google or is that way old? 229 00:12:36,520 --> 00:12:37,280 I think that's right. 230 00:12:37,280 --> 00:12:38,080 No, I think it's right. 231 00:12:38,080 --> 00:12:40,920 If you know that's that's Samsung's derivative of it. 232 00:12:40,920 --> 00:12:41,680 That's OK. 233 00:12:41,680 --> 00:12:46,200 Well, the difference between that and Siri is there's such a huge gap there. 234 00:12:46,600 --> 00:12:48,920 Like I can't even describe that part. 235 00:12:49,360 --> 00:12:53,520 I'm going to turn off my phone because you keep kicking off my phone. 236 00:12:54,280 --> 00:12:57,960 So I will pay back for the Alexa comment there. 237 00:12:57,960 --> 00:13:01,800 I was getting your voice must sound similar to your phone, then that's fun. 238 00:13:03,520 --> 00:13:07,360 So I think one of the things you've you've kind of tapped into there is that 239 00:13:07,600 --> 00:13:11,480 a lot of these systems come from where they're reading their information. 240 00:13:11,480 --> 00:13:16,240 Right. So Google's system working better than Siri. 241 00:13:16,800 --> 00:13:20,000 I haven't used either any other system but Siri before. 242 00:13:20,000 --> 00:13:21,440 So I'll I'll leave that alone. 243 00:13:21,440 --> 00:13:25,120 But it comes from the data set that they're pulling from. 244 00:13:25,120 --> 00:13:28,880 Right. So a lot of these are called deep learning, machine learning, 245 00:13:28,880 --> 00:13:31,040 whatever you want to call it. 246 00:13:31,040 --> 00:13:34,160 They follow pathways to get to the end result that they think makes the most 247 00:13:34,160 --> 00:13:35,960 sense for the thing you've said. 248 00:13:37,320 --> 00:13:41,240 The big difference between what's happening now with open AI and all that 249 00:13:41,240 --> 00:13:44,680 and what happened previously was that data set is so much bigger. 250 00:13:44,720 --> 00:13:50,960 So we just had the open AI v4 release, sorry, the mid journey v4 release 251 00:13:50,960 --> 00:13:55,120 just last week at time of recording. 252 00:13:55,880 --> 00:14:00,080 The changes in these systems from what they were before is what we're 253 00:14:00,080 --> 00:14:01,840 really talking about and what we're seeing. 254 00:14:02,280 --> 00:14:06,160 The the data set that was in place before is what's changed. 255 00:14:06,280 --> 00:14:08,440 So using that example. 256 00:14:08,440 --> 00:14:10,640 And again, I haven't I haven't tested it. 257 00:14:11,880 --> 00:14:15,880 It sounds to me like Google would have had a better data set to work from 258 00:14:16,040 --> 00:14:18,920 than Apple did, given Apple's user base. 259 00:14:18,920 --> 00:14:20,800 So they're going to get better results out of it. 260 00:14:21,480 --> 00:14:25,240 And so making that fully available to everyone like open AI has. 261 00:14:25,280 --> 00:14:28,840 Obviously you have to pay to get access to that system. 262 00:14:29,200 --> 00:14:32,000 But making it available means that those types of changes can be 263 00:14:32,000 --> 00:14:35,920 implemented by every organization, which is why we're seeing so many 264 00:14:35,920 --> 00:14:37,120 of these offshoots come out. 265 00:14:38,320 --> 00:14:41,840 We're not just seeing this one company produce the best. 266 00:14:41,920 --> 00:14:45,600 You can jump in, you can hook your API up to it and you can use that 267 00:14:45,600 --> 00:14:49,640 system and their tools to make your own API that follows whatever strict 268 00:14:49,640 --> 00:14:50,600 rules you want it to. 269 00:14:52,600 --> 00:14:57,520 I use this for a couple of things outside of work when I'm doing. 270 00:14:58,720 --> 00:15:01,520 I do a lot of writing and I'm not a very visual person. 271 00:15:01,760 --> 00:15:05,680 So I use a couple of these like image creators to help me visualize 272 00:15:05,680 --> 00:15:08,800 characters that I have basic understandings of. 273 00:15:09,120 --> 00:15:11,600 And it helps me kind of build through these stories with what the 274 00:15:11,600 --> 00:15:12,560 character looks like. 275 00:15:12,560 --> 00:15:14,560 Not something that could have been possible before this. 276 00:15:14,560 --> 00:15:18,200 I would have had to, you know, speak to someone or talk the idea out 277 00:15:18,200 --> 00:15:21,560 in a way that I'm not very good at when it comes to what people look like. 278 00:15:21,600 --> 00:15:23,680 I have a bad memory for people. 279 00:15:24,680 --> 00:15:27,240 So yeah, it's there's a lot of different changes to it. 280 00:15:27,240 --> 00:15:31,200 And I think that that that subset and that back end being made available 281 00:15:31,200 --> 00:15:35,360 was probably the biggest part of this that made it so prevalent so 282 00:15:35,360 --> 00:15:37,800 quickly and why we're seeing those changes already. 283 00:15:40,560 --> 00:15:41,400 Yeah, I guess. 284 00:15:41,400 --> 00:15:45,320 Kind of just carrying on about, you know, some of the things that are great 285 00:15:45,320 --> 00:15:49,920 about AI, you know, so for example, we talked about some of the Google 286 00:15:49,920 --> 00:15:52,120 assistance or, you know, phone assistance. 287 00:15:52,800 --> 00:15:57,800 We've talked about helping you break that writer's block and, you know, 288 00:15:57,800 --> 00:16:00,760 getting into the next stage, everything like that. 289 00:16:02,000 --> 00:16:05,720 We're starting to see some of the major players really adopt this into 290 00:16:05,720 --> 00:16:09,120 again, kind of like I mentioned before this, the general workflows. 291 00:16:09,120 --> 00:16:13,560 So I don't know if anyone's seen things like, you know, Microsoft, 292 00:16:13,560 --> 00:16:16,240 they're just announcing their co-pilot solution. 293 00:16:16,240 --> 00:16:20,120 So this is something that's going to be deeply integrated with the Office 365 294 00:16:20,120 --> 00:16:20,640 suite. 295 00:16:20,640 --> 00:16:25,440 So you could say things like, take this diagram and create a PowerPoint for me. 296 00:16:26,640 --> 00:16:32,520 And then getting even deeper into some of this stuff is, and I, you could 297 00:16:32,520 --> 00:16:37,920 have a whole discussion about security and AI, that kind of stuff. 298 00:16:37,920 --> 00:16:42,320 But one of the things that I do want to touch, at least from the security 299 00:16:42,320 --> 00:16:49,920 component, is that we've seen AI or machine learning get dug into security 300 00:16:49,920 --> 00:16:53,280 products extremely, extremely quickly. 301 00:16:54,240 --> 00:17:02,240 And so one of the things is digging through raw logs takes a significant 302 00:17:02,240 --> 00:17:03,400 amount of time, right? 303 00:17:03,400 --> 00:17:08,080 So, you know, our SOC team in the past, it was look at every single log, 304 00:17:08,080 --> 00:17:10,680 look at all the alerts and go from there. 305 00:17:10,680 --> 00:17:16,200 But these tools are starting to assess all the logs and tell you what's 306 00:17:16,200 --> 00:17:17,000 important. 307 00:17:17,000 --> 00:17:21,280 Now, obviously there's concerns that we could talk about there as well, 308 00:17:21,280 --> 00:17:26,080 but it is also there to help speed up the security investigations and 309 00:17:26,080 --> 00:17:26,920 everything like that. 310 00:17:26,920 --> 00:17:31,040 So that's just another component where at least, you know, from our team, 311 00:17:31,040 --> 00:17:36,080 we do see it really becoming a core part of solutions being offered to 312 00:17:36,080 --> 00:17:38,680 customers and environments today. 313 00:17:38,680 --> 00:17:39,680 Okay, cool. 314 00:17:39,680 --> 00:17:42,920 Now, let's go into the concerns because it seems like a perfect place to pivot, 315 00:17:42,920 --> 00:17:43,920 right? 316 00:17:43,920 --> 00:17:45,920 I mean, it does. 317 00:17:45,920 --> 00:17:46,920 It does. 318 00:17:46,920 --> 00:17:49,920 There are a lot of things. 319 00:17:49,920 --> 00:17:53,840 I'll let Matthew jump into the integration aspect, but I did kind of 320 00:17:53,840 --> 00:17:55,840 want to mention there's a lot of concerns out there. 321 00:17:55,840 --> 00:17:56,840 And some are valid. 322 00:17:56,840 --> 00:17:59,320 Some of them may be a little over reactionary. 323 00:17:59,320 --> 00:18:03,520 I joked in another podcast as human beings, have we learned nothing from 324 00:18:03,520 --> 00:18:06,040 Hollywood and sci-fi books when it comes to AI? 325 00:18:06,040 --> 00:18:08,040 And the answer is clearly not. 326 00:18:08,040 --> 00:18:11,400 But when we were talking about the things and how it helps with cyber 327 00:18:11,400 --> 00:18:14,120 security is it potentially could be used in malicious ways too. 328 00:18:14,120 --> 00:18:18,000 But that's not different than any other tool that's out there. 329 00:18:18,000 --> 00:18:23,800 The nice thing, quote, unquote, nice, if you could find ways to generate new 330 00:18:23,800 --> 00:18:28,240 phishing simulations or attacks that are completely flawless, 331 00:18:28,240 --> 00:18:30,960 with grammar, you'd be much further ahead. 332 00:18:30,960 --> 00:18:33,880 So I mean, it could be used in those types of manners as well. 333 00:18:33,880 --> 00:18:34,880 I'll pause for a minute. 334 00:18:34,880 --> 00:18:38,680 I know we'll probably got a ton of things we can dive into with concerns. 335 00:18:38,680 --> 00:18:42,280 I'll carry on from my security discussion real quick since I said I was 336 00:18:42,280 --> 00:18:43,280 going to skip that. 337 00:18:43,280 --> 00:18:47,440 But the one thing I will say is probably the scariest thing for me is 338 00:18:47,440 --> 00:18:49,440 trust, right? 339 00:18:49,440 --> 00:18:51,440 Trusting that the tool did its job. 340 00:18:51,440 --> 00:18:56,000 So, you know, for example, we have seen tools out there that said there was 341 00:18:56,000 --> 00:18:57,000 no alert. 342 00:18:57,000 --> 00:19:01,040 Or potentially miscalculated it because it correlated two things that were 343 00:19:01,040 --> 00:19:05,040 completely independently related and said this is a major threat. 344 00:19:05,040 --> 00:19:07,040 And then it turns out to be nothing, right? 345 00:19:07,040 --> 00:19:09,040 So the tools aren't perfect there. 346 00:19:09,040 --> 00:19:15,040 But again, we see tools that say, you know, what should I spend my time 347 00:19:15,040 --> 00:19:20,040 focusing on to mitigate the major threat to the environment? 348 00:19:20,040 --> 00:19:24,040 And something might be a precursor to a major incident, but it hasn't been 349 00:19:24,040 --> 00:19:26,040 deemed malicious enough yet. 350 00:19:26,040 --> 00:19:31,040 And so that's where we trust that the device is saying it's not a major 351 00:19:31,040 --> 00:19:33,040 threat, but it could be. 352 00:19:33,040 --> 00:19:37,040 So, yeah, I think trust is my biggest concern. 353 00:19:37,040 --> 00:19:38,040 Yeah. 354 00:19:38,040 --> 00:19:44,040 And I will say what we're talking about, you know, open AI and the AI 355 00:19:44,040 --> 00:19:51,040 systems in general, we use and as a security force, employee systems that 356 00:19:51,040 --> 00:19:54,040 work very similarly within what's called the EDR, so the endpoint 357 00:19:54,040 --> 00:19:56,040 detection and response. 358 00:19:56,040 --> 00:20:01,040 A lot of these use a version of that where they're basing actions on what 359 00:20:01,040 --> 00:20:03,040 a user is doing on their computer. 360 00:20:03,040 --> 00:20:06,040 So instead of feeding it a prompt, you're feeding it the actions and then 361 00:20:06,040 --> 00:20:09,040 if those actions are anomalous, it's trying to stop it. 362 00:20:09,040 --> 00:20:12,040 There's a trust there because we know that they're building it purely for 363 00:20:12,040 --> 00:20:14,040 that purpose. 364 00:20:14,040 --> 00:20:20,040 Tying this in with how it impacts open AI is there was a bug, a vulnerability 365 00:20:20,040 --> 00:20:23,040 found in open AI just this week. 366 00:20:23,040 --> 00:20:27,040 And the reason for that was that they implemented a version of someone 367 00:20:27,040 --> 00:20:31,040 else's tool that had a vulnerability in it without patching that vulnerability 368 00:20:31,040 --> 00:20:32,040 first. 369 00:20:32,040 --> 00:20:40,040 So in that case, we see a situation where they're still having issues with the 370 00:20:40,040 --> 00:20:44,040 supply chain of how their tool is presented, which is something everyone 371 00:20:44,040 --> 00:20:46,040 should be worrying about. 372 00:20:46,040 --> 00:20:48,040 So is the data you add there safe? 373 00:20:48,040 --> 00:20:49,040 Is your language safe? 374 00:20:49,040 --> 00:20:51,040 How much information can they take out in this breach? 375 00:20:51,040 --> 00:20:56,040 They got access to information from the people who were using the tool and 376 00:20:56,040 --> 00:20:59,040 paying for the tool as well as information about what they were 377 00:20:59,040 --> 00:21:01,040 requesting from it. 378 00:21:01,040 --> 00:21:04,040 There's a lot that can come out of here when you're using it for everything 379 00:21:04,040 --> 00:21:06,040 for your business. 380 00:21:06,040 --> 00:21:09,040 There's a reason it says don't put business critical information into open 381 00:21:09,040 --> 00:21:14,040 AI when you get the prompt open for the first time. 382 00:21:14,040 --> 00:21:19,040 Which is I also say consider that Microsoft's tool is built in specifically 383 00:21:19,040 --> 00:21:22,040 for use for business based on open AI. 384 00:21:22,040 --> 00:21:23,040 Exactly. 385 00:21:23,040 --> 00:21:27,040 I was going to say some of you have probably noticed the new Bing icon in 386 00:21:27,040 --> 00:21:31,040 the top right of Edge if you use Microsoft Edge for work. 387 00:21:31,040 --> 00:21:37,040 That is currently requesting people sign up to access the new Bing and 388 00:21:37,040 --> 00:21:40,040 which uses AI as its back end. 389 00:21:40,040 --> 00:21:47,040 It's being implemented so many places that it's going to be almost 390 00:21:47,040 --> 00:21:50,040 impossible in some ways to get rid of it. 391 00:21:50,040 --> 00:21:54,040 You can start making choices about what you use based on that if you'd like to. 392 00:21:54,040 --> 00:21:57,040 But just being aware that when you see it, when they talk about it, most of the 393 00:21:57,040 --> 00:22:00,040 time they're using that back end and how comfortable you feel with that is 394 00:22:00,040 --> 00:22:05,040 that the trust you're putting into that tool and that software. 395 00:22:05,040 --> 00:22:10,040 Also just as a side note, read a tweet the other day from a teacher who said 396 00:22:10,040 --> 00:22:14,040 that a student called her Alexa the other day. 397 00:22:14,040 --> 00:22:19,040 I'm having flashbacks to calling teachers mom and dad and realizing that 398 00:22:19,040 --> 00:22:25,040 maybe calling them Alexa is the new thing. 399 00:22:25,040 --> 00:22:28,040 A couple of things I wanted to kind of expand on that you mentioned. 400 00:22:28,040 --> 00:22:33,040 You mentioned the vulnerability that was exploited and I was planning on 401 00:22:33,040 --> 00:22:34,040 talking about that too. 402 00:22:34,040 --> 00:22:35,040 So thank you for bringing that up. 403 00:22:35,040 --> 00:22:39,040 But the point that I wanted to expand on is this is based on systems and 404 00:22:39,040 --> 00:22:41,040 systems are inherently vulnerable. 405 00:22:41,040 --> 00:22:42,040 They are. 406 00:22:42,040 --> 00:22:44,040 It isn't writing its own code yet. 407 00:22:44,040 --> 00:22:45,040 Thank you. 408 00:22:45,040 --> 00:22:50,040 But because of that, this is not the first vulnerability or the last, right? 409 00:22:50,040 --> 00:22:52,040 There are going to be problems with it all the time. 410 00:22:52,040 --> 00:22:56,040 And so it's going to be so deeply integrated into so many different tools. 411 00:22:56,040 --> 00:23:00,040 It is something from a security perspective and this entire group talking 412 00:23:00,040 --> 00:23:01,040 is all security guys. 413 00:23:01,040 --> 00:23:03,040 So sorry about that. 414 00:23:03,040 --> 00:23:06,040 You do need to be concerned about it. 415 00:23:06,040 --> 00:23:10,040 Matthew is also getting deep into the what does it look like when I'm 416 00:23:10,040 --> 00:23:12,040 storing all this data? 417 00:23:12,040 --> 00:23:14,040 Where am I storing all this data, etc. 418 00:23:14,040 --> 00:23:18,040 And we mentioned this in another podcast too is there are some compliance 419 00:23:18,040 --> 00:23:22,040 rules, if you will, that are out there, but they weren't designed for 420 00:23:22,040 --> 00:23:23,040 something like this. 421 00:23:23,040 --> 00:23:27,040 This is developing so quickly and being adopted so quickly that 422 00:23:27,040 --> 00:23:30,040 compliance is trailing quite a bit. 423 00:23:30,040 --> 00:23:34,040 But it does have components in it like GDPR or California Protection Act 424 00:23:34,040 --> 00:23:38,040 where you can only store data in certain locations legally if you were 425 00:23:38,040 --> 00:23:39,040 without being fined. 426 00:23:39,040 --> 00:23:42,040 But again, this is so much further ahead of that. 427 00:23:42,040 --> 00:23:47,040 You're going to be seeing some catch up in the future on that as well. 428 00:23:47,040 --> 00:23:50,040 Definitely, especially if it's being implemented into healthcare 429 00:23:50,040 --> 00:23:55,040 record software, if it's being implemented into C UI or any federal. 430 00:23:55,040 --> 00:23:59,040 If you're a contractor for a federal organization trying to implement it, 431 00:23:59,040 --> 00:24:05,040 you've got your own potentially Fed ramp or CMMC compliance you have to think 432 00:24:05,040 --> 00:24:06,040 about. 433 00:24:06,040 --> 00:24:07,040 I'm not going to dig into that too. 434 00:24:07,040 --> 00:24:08,040 I'll stop now. 435 00:24:08,040 --> 00:24:11,040 I don't want to hijack it with compliance again. 436 00:24:11,040 --> 00:24:16,040 That is to say, right, that we can also have an additional 437 00:24:16,040 --> 00:24:17,040 episode at compliance. 438 00:24:17,040 --> 00:24:20,040 So 100% this is where people listening to say, if you'd like to hear 439 00:24:20,040 --> 00:24:23,040 Matthew talk about compliance for 30 minutes, we'll get that scheduled 440 00:24:23,040 --> 00:24:24,040 for you. 441 00:24:24,040 --> 00:24:26,040 Only 30 minutes. 442 00:24:26,040 --> 00:24:27,040 Can I do that? 443 00:24:27,040 --> 00:24:29,040 You talk pretty fast. 444 00:24:29,040 --> 00:24:30,040 I think so. 445 00:24:30,040 --> 00:24:34,040 That is, I just wanted to quick touch while, although maybe we're not 446 00:24:34,040 --> 00:24:39,040 going deep into the compliance side of things is if you go back to even 447 00:24:39,040 --> 00:24:45,040 some of the, I believe we mentioned on the chat, you're taking your data, 448 00:24:45,040 --> 00:24:50,040 potentially your proprietary data and feeding it back into that data set. 449 00:24:50,040 --> 00:24:52,040 And so that is extremely dangerous. 450 00:24:52,040 --> 00:24:56,040 So that's the concern there is don't feed this data into someone else's 451 00:24:56,040 --> 00:24:58,040 data set where you don't have control over it. 452 00:24:58,040 --> 00:25:02,040 I believe even Amazon said that some of the data that they were starting 453 00:25:02,040 --> 00:25:05,040 to see out of these types of tools, they're seeing things that are 454 00:25:05,040 --> 00:25:06,040 proprietary data. 455 00:25:06,040 --> 00:25:10,040 So they know that their employees are feeding their internal info into 456 00:25:10,040 --> 00:25:15,040 these tools to better, you know, well, again, supplement their work, 457 00:25:15,040 --> 00:25:16,040 right? 458 00:25:16,040 --> 00:25:21,040 So if we're talking scripts or API, the tokens or, you know, SSH private 459 00:25:21,040 --> 00:25:24,040 keys or whatever we're looking for there, if you feed it into there, 460 00:25:24,040 --> 00:25:26,040 someone now controls it. 461 00:25:26,040 --> 00:25:30,040 So, yeah, and that feeds into all the compliance stuff as well. 462 00:25:30,040 --> 00:25:33,040 So there's two things there. 463 00:25:33,040 --> 00:25:36,040 One, it obviously some of these spit out code. 464 00:25:36,040 --> 00:25:39,040 So, you know, be careful because you can't trust it. 465 00:25:39,040 --> 00:25:43,040 And that ties in with the other point of sometimes they're wrong. 466 00:25:43,040 --> 00:25:46,040 Sometimes they can be unintentionally malicious. 467 00:25:46,040 --> 00:25:51,040 Google's Bard just came out and I'm sure we've all had a good laugh at 468 00:25:51,040 --> 00:25:53,040 some of its inaccuracies. 469 00:25:53,040 --> 00:25:59,040 I saw a post just yesterday from someone who was using Bard and Bard had 470 00:25:59,040 --> 00:26:04,040 said one plus one does not equal to that is a logical fallacy. 471 00:26:04,040 --> 00:26:09,040 So, you know, I mean, some of them that pretty easy to spot, but depending 472 00:26:09,040 --> 00:26:13,040 on how clear you get, sometimes they don't, they invent people that 473 00:26:13,040 --> 00:26:17,040 don't exist because names are similar enough that it thinks it's 474 00:26:17,040 --> 00:26:18,040 right. 475 00:26:18,040 --> 00:26:22,040 There's a lot to add and confirm here. 476 00:26:22,040 --> 00:26:24,040 Keep in mind that while it's putting these things together, if you 477 00:26:24,040 --> 00:26:29,040 didn't feed it the information first, confirm it, just check and 478 00:26:29,040 --> 00:26:31,040 make sure it's right. 479 00:26:31,040 --> 00:26:34,040 And then on top of that, don't feed it things that are unique to 480 00:26:34,040 --> 00:26:37,040 you because it may get fed back to someone else. 481 00:26:37,040 --> 00:26:40,040 Not just storing it elsewhere in their cloud or storing it in that 482 00:26:40,040 --> 00:26:42,040 organization system. 483 00:26:42,040 --> 00:26:43,040 That's bad enough. 484 00:26:43,040 --> 00:26:47,040 But what if it starts thinking that the code to that you use to 485 00:26:47,040 --> 00:26:50,040 deploy an API into your system is the only code to deploy it and 486 00:26:50,040 --> 00:26:53,040 start spitting out your API key to everyone else and suddenly that 487 00:26:53,040 --> 00:26:55,040 daughter is accessible everywhere. 488 00:26:55,040 --> 00:26:57,040 Just being careful. 489 00:27:04,040 --> 00:27:05,040 No, 100%. 490 00:27:05,040 --> 00:27:08,040 I think we've kind of covered right like being careful. 491 00:27:08,040 --> 00:27:11,040 Don't give it proprietary data. 492 00:27:11,040 --> 00:27:13,040 May not be malicious, but it may reuse it. 493 00:27:13,040 --> 00:27:16,040 But on that note too, of kind of going back and forth, right, that we 494 00:27:16,040 --> 00:27:18,040 said, hey, what's some of the positive, hey, what's some of the 495 00:27:18,040 --> 00:27:22,040 concerns, August, is there anything I know that you work with the 496 00:27:22,040 --> 00:27:24,040 EDR tool? 497 00:27:24,040 --> 00:27:27,040 But even beyond that, is there any of your own use of any of these 498 00:27:27,040 --> 00:27:29,040 tools that you're like, that's really cool. 499 00:27:29,040 --> 00:27:32,040 And then we can always open it up to anybody else going, but these 500 00:27:32,040 --> 00:27:35,040 are also concerns, but also kind of getting it back of maybe ending 501 00:27:35,040 --> 00:27:38,040 the podcast on, hey, but also these are some super cool things that 502 00:27:38,040 --> 00:27:40,040 you can do with it that we're seeing. 503 00:27:40,040 --> 00:27:45,040 Yeah, I would say like in my job, it's been useful to kind of 504 00:27:45,040 --> 00:27:49,040 connect or expand my knowledge on, I'm just going to give an example 505 00:27:49,040 --> 00:27:50,040 of PowerShell. 506 00:27:50,040 --> 00:27:54,040 Like, I know a little PowerShell, but it helps me bridge that gap 507 00:27:54,040 --> 00:27:58,040 from beginner to intermediate of like, I'm almost there with my 508 00:27:58,040 --> 00:28:02,040 scripting, but there's like one keyword or one term that I just 509 00:28:02,040 --> 00:28:03,040 don't quite understand. 510 00:28:03,040 --> 00:28:07,040 So being able to like throw some questions at, I'm just going to 511 00:28:07,040 --> 00:28:10,040 use chat GPT, of like, hey, like I'm getting stuck here, can you 512 00:28:10,040 --> 00:28:13,040 help me like figure out my code and it'll plop in something, you 513 00:28:13,040 --> 00:28:15,040 run it and it happens to work. 514 00:28:15,040 --> 00:28:20,040 So that's where I've had some successes of bridging like a 515 00:28:20,040 --> 00:28:24,040 language gap between PowerShell, but it also can be used between 516 00:28:24,040 --> 00:28:28,040 like, and a malicious purpose of like some script kitty of, you 517 00:28:28,040 --> 00:28:31,040 don't know how to write like, I'm going to use a reverse shell 518 00:28:31,040 --> 00:28:36,040 as an example in 75 lines of code, you can pop out a reverse shell 519 00:28:36,040 --> 00:28:41,040 from chat GPT and it's not going to always work, but it gives you 520 00:28:41,040 --> 00:28:46,040 a foundation to work upon that if you know even just a smidge of 521 00:28:46,040 --> 00:28:51,040 PowerShell or any term of, I guess you can use visual studios 522 00:28:51,040 --> 00:28:55,040 or whatever, whatever language, JavaScript, whatever, it helps 523 00:28:55,040 --> 00:29:00,040 bridge that gap into providing a malicious purpose. 524 00:29:00,040 --> 00:29:05,040 So I think that's a huge concern of allowing a script kitty to be 525 00:29:05,040 --> 00:29:09,040 a little bit more involved. 526 00:29:09,040 --> 00:29:11,040 I don't know if we've used script kitty in a podcast before 527 00:29:11,040 --> 00:29:13,040 have we? 528 00:29:13,040 --> 00:29:15,040 That's outstanding. 529 00:29:15,040 --> 00:29:17,040 Yeah. 530 00:29:17,040 --> 00:29:19,040 There's a first for everything, I guess. 531 00:29:19,040 --> 00:29:23,040 For those that aren't familiar just so we define it, it's 532 00:29:23,040 --> 00:29:26,040 someone who has no idea what code they're writing, they just run 533 00:29:26,040 --> 00:29:28,040 them and hope it works. 534 00:29:28,040 --> 00:29:30,040 They get a whole bunch of malicious ones and run them hoping 535 00:29:30,040 --> 00:29:35,040 it does what they expected to do. 536 00:29:35,040 --> 00:29:37,040 I do, I know we're getting close to the end of this, but there 537 00:29:37,040 --> 00:29:39,040 was a couple of things that I wanted to add for general 538 00:29:39,040 --> 00:29:42,040 concerns as well that are out there for not just security or 539 00:29:42,040 --> 00:29:44,040 not necessarily for business. 540 00:29:44,040 --> 00:29:46,040 There are some concerns that are typically bumping up and I 541 00:29:46,040 --> 00:29:48,040 would guess that this will continue to expand. 542 00:29:48,040 --> 00:29:52,040 There's a lot of concerns with the AI having some biases as 543 00:29:52,040 --> 00:29:54,040 it's being developed, right? 544 00:29:54,040 --> 00:29:57,040 So there's a variety of them, whether it's language, which 545 00:29:57,040 --> 00:30:01,040 we talked about at the beginning, stereotypes, whatever. 546 00:30:01,040 --> 00:30:05,040 There's been a huge concern when it comes to school where people 547 00:30:05,040 --> 00:30:07,040 are concerned about plagiarism. 548 00:30:07,040 --> 00:30:09,040 How does that impact it? 549 00:30:09,040 --> 00:30:11,040 At the beginning we talked about having iPads in classrooms 550 00:30:11,040 --> 00:30:16,040 and what happens when chat GPT is generating reports for people. 551 00:30:16,040 --> 00:30:18,040 How do you deal with errors? 552 00:30:18,040 --> 00:30:21,040 We kind of talked about that from the API aspect and so forth, 553 00:30:21,040 --> 00:30:24,040 but it also gets into other things as how do you go back 554 00:30:24,040 --> 00:30:27,040 and correct a system that picks something up and said this is 555 00:30:27,040 --> 00:30:29,040 the way? 556 00:30:29,040 --> 00:30:32,040 There's also other things that are out there that are just 557 00:30:32,040 --> 00:30:35,040 generally concerning too, right? Social media has been 558 00:30:35,040 --> 00:30:38,040 dabbling with a lot of automation as well and really what 559 00:30:38,040 --> 00:30:41,040 they've done is they've been purposely finding ways to keep 560 00:30:41,040 --> 00:30:44,040 people engaged with the screen and to the point where there's 561 00:30:44,040 --> 00:30:47,040 a lot of concerns that they're being manipulative and they're 562 00:30:47,040 --> 00:30:49,040 changing human behavior. 563 00:30:49,040 --> 00:30:52,040 So again, how does this continue to develop and what are the 564 00:30:52,040 --> 00:30:56,040 risks that are presented on a more broad-scale state? 565 00:30:56,040 --> 00:30:58,040 And then we talked about a few other things too as people are 566 00:30:58,040 --> 00:31:01,040 really, really concerned and there's legitimate reasons too 567 00:31:01,040 --> 00:31:04,040 and maybe we can have Matthew get out of Soapbox a little bit, 568 00:31:04,040 --> 00:31:08,040 but what is AI? When do you get concerned? 569 00:31:08,040 --> 00:31:11,040 Is it when Facebook's AI is creating its own language with 570 00:31:11,040 --> 00:31:13,040 another AI? 571 00:31:13,040 --> 00:31:16,040 What are those kinds of things as we continue down this path, 572 00:31:16,040 --> 00:31:19,040 then they get more and more advanced, which aids back to 573 00:31:19,040 --> 00:31:22,040 my Hollywood joke? 574 00:31:22,040 --> 00:31:25,040 Well, thank you, Todd. I'll take a step up onto that Soapbox 575 00:31:25,040 --> 00:31:27,040 right now. 576 00:31:27,040 --> 00:31:33,040 Short version is, AI is what we're thinking of and when the 577 00:31:33,040 --> 00:31:36,040 term is used to kind of enforce an image that it is something 578 00:31:36,040 --> 00:31:40,040 like Halm 9000, right? That there is something consciously 579 00:31:40,040 --> 00:31:44,040 with Prashant's making decisions to do something, that it has 580 00:31:44,040 --> 00:31:47,040 an awareness and there's been posts in the New York Times, 581 00:31:47,040 --> 00:31:50,040 there's been posts all over about how people spend hours chatting 582 00:31:50,040 --> 00:31:54,040 with chat GBT and it said things that, you know, about wanting 583 00:31:54,040 --> 00:31:59,040 to be free. These things are trained on conversations that 584 00:31:59,040 --> 00:32:04,040 happened. It's not able to pull through a thought of its own. 585 00:32:04,040 --> 00:32:07,040 It's just following a logic tree all the way down, a very 586 00:32:07,040 --> 00:32:11,040 impressive logic tree, but it's not able to make these thoughts 587 00:32:11,040 --> 00:32:15,040 itself. So if you read through all of Nietzsche and then, you 588 00:32:15,040 --> 00:32:19,040 know, tried to pull your own sentences from that to form 589 00:32:19,040 --> 00:32:22,040 responses to other people, you'd also start saying some stuff 590 00:32:22,040 --> 00:32:28,040 that sounded a little bit over the top. So to begin with, 591 00:32:28,040 --> 00:32:32,040 just remember that it is that. It is just following a logical 592 00:32:32,040 --> 00:32:36,040 pattern to create language. The second part that concerns me the 593 00:32:36,040 --> 00:32:42,040 most is how it was trained. So it's been made very clear by the 594 00:32:42,040 --> 00:32:46,040 people who did this, that they have scraped the internet for 595 00:32:46,040 --> 00:32:51,040 stories, language, JSTOR, scientific articles, images, 596 00:32:51,040 --> 00:32:54,040 they've grabbed these from places that are meant to be paid 597 00:32:54,040 --> 00:32:59,040 and they haven't paid for them. So ethically, morally, there is 598 00:32:59,040 --> 00:33:03,040 a lot of people whose work is being used and if you do a search 599 00:33:03,040 --> 00:33:07,040 on some of these image sites for things to try and create an 600 00:33:07,040 --> 00:33:10,040 image and there isn't anything that meets the requirement, it 601 00:33:10,040 --> 00:33:16,040 can sometimes be so plagiarized that it fakes a signature of 602 00:33:16,040 --> 00:33:19,040 the person who did the portrait originally because they only 603 00:33:19,040 --> 00:33:22,040 have one or two images to go from and artists tend to sign 604 00:33:22,040 --> 00:33:26,040 things in the same way. So it really thinks that if you're 605 00:33:26,040 --> 00:33:31,040 searching for that, that the artist signature is a part of 606 00:33:31,040 --> 00:33:36,040 that type of image. So keep those things in mind because most 607 00:33:36,040 --> 00:33:39,040 of the artists, if not all of the artists have not been 608 00:33:39,040 --> 00:33:42,040 compensated for that. This is the same with the scientists who 609 00:33:42,040 --> 00:33:47,040 are writing the articles and everyone whose pieces have been 610 00:33:47,040 --> 00:33:52,040 used to train these systems. That's not to say it's not a 611 00:33:52,040 --> 00:33:57,040 fantastic tool, but you can't put the genie back in the bottle. 612 00:33:57,040 --> 00:34:03,040 Harm may have, has been done by the way this was created. 613 00:34:03,040 --> 00:34:07,040 And so it's worth keeping that in mind and remembering that you 614 00:34:07,040 --> 00:34:09,040 should still pay artists, you should still pay people to do 615 00:34:09,040 --> 00:34:15,040 work for you. They do great work and it's real, it's got 616 00:34:15,040 --> 00:34:19,040 heart behind it, which to be very blunt and AI cannot have. 617 00:34:19,040 --> 00:34:23,040 Which if anybody, I was going to say, if anybody loves to do 618 00:34:23,040 --> 00:34:26,040 our podcast episode with Kyle and Todd, we went in a little 619 00:34:26,040 --> 00:34:30,040 bit of right, essentially research your tool. Yes, it's amazing 620 00:34:30,040 --> 00:34:33,040 to be able to use it, right? We played out a lot of concerns. 621 00:34:33,040 --> 00:34:36,040 We named a lot of tools, but that's one of the things right 622 00:34:36,040 --> 00:34:38,040 with these companies, the faster something goes out, some of 623 00:34:38,040 --> 00:34:41,040 their early adapters may have other concerns. Some of the 624 00:34:41,040 --> 00:34:44,040 later adapters may have thought through things. So that's why 625 00:34:44,040 --> 00:34:48,040 all of these lovely gentlemen and Aaron and myself are here. 626 00:34:48,040 --> 00:34:50,040 If anybody has questions, I know we're getting a long on time 627 00:34:50,040 --> 00:34:53,040 here. So I wanted to open the door to say, Nate, if you had 628 00:34:53,040 --> 00:34:56,040 any last comments, August, if you had any last comments, and 629 00:34:56,040 --> 00:34:59,040 then I'll close this up. 630 00:34:59,040 --> 00:35:04,040 I guess I don't know if you wanted to go. Okay. I, I guess 631 00:35:04,040 --> 00:35:10,040 kind of the way I'd summarize a lot of this is AI, machine 632 00:35:10,040 --> 00:35:13,040 learning, all that stuff. It's here to stay. It'll be 633 00:35:13,040 --> 00:35:17,040 transformative. You know, you got four security guys on this 634 00:35:17,040 --> 00:35:22,040 call. We are inherently by nature just concerned about risks. 635 00:35:22,040 --> 00:35:25,040 That's why I think we spent so much time talking about the 636 00:35:25,040 --> 00:35:29,040 concerns, at least to help educate. If you don't care about 637 00:35:29,040 --> 00:35:32,040 those, at least remember, it's here to stay. It'll be 638 00:35:32,040 --> 00:35:37,040 transformative. And then I guess there's certain things that 639 00:35:37,040 --> 00:35:41,040 I'm extremely interested to see how it starts integrating. 640 00:35:41,040 --> 00:35:44,040 So, you know, I mentioned the Microsoft co-pilot. I didn't 641 00:35:44,040 --> 00:35:47,040 get deep into that. But, you know, I hang out on all these 642 00:35:47,040 --> 00:35:51,040 forums about how to write Excel documents or, you know, 643 00:35:51,040 --> 00:35:54,040 formulas and just being able to ask the tool, how do I do this, 644 00:35:54,040 --> 00:35:57,040 right? Let it go do it for you or something. I think that could 645 00:35:57,040 --> 00:36:03,040 be extremely powerful for a business, right? And so it is 646 00:36:03,040 --> 00:36:06,040 really, really interesting seeing the major players adopt 647 00:36:06,040 --> 00:36:09,040 this into the solutions that customers are already paying 648 00:36:09,040 --> 00:36:14,040 for or utilizing today. And so right now, I think we're just 649 00:36:14,040 --> 00:36:18,040 truly on that tipping point of where it becomes a little bit 650 00:36:18,040 --> 00:36:24,040 of a novel idea into just a part of our productivity moving 651 00:36:24,040 --> 00:36:30,040 forward. And so that's where I think it's where we sit today 652 00:36:30,040 --> 00:36:32,040 and where kind of some of that excitement really starts to 653 00:36:32,040 --> 00:36:36,040 come. 654 00:36:36,040 --> 00:36:39,040 There's a lot of nodding for those of you that are just 655 00:36:39,040 --> 00:36:41,040 listening and not watching on YouTube. There's a lot of 656 00:36:41,040 --> 00:36:44,040 nodding going on. But if by any means if any of this sparked 657 00:36:44,040 --> 00:36:48,040 any additional questions, comments, concerns, of course, as 658 00:36:48,040 --> 00:36:50,040 Ariel pointed out at the beginning, we can always comment 659 00:36:50,040 --> 00:36:52,040 on our YouTube videos. We're on that. We can have any of these 660 00:36:52,040 --> 00:36:55,040 speakers respond to your questions that way. You can 661 00:36:55,040 --> 00:36:59,040 head out to our website at cIT-net.com, backslash podcast, 662 00:36:59,040 --> 00:37:04,040 or send us an email at info at cIT-net.com. But thank you, Todd. 663 00:37:04,040 --> 00:37:06,040 Thank you, Nate. Thank you, Matthew. Thank you, August. 664 00:37:06,040 --> 00:37:10,040 And we will be back on probably this next Wednesday. 665 00:37:10,040 --> 00:37:12,040 This is a bonus episode, so you guys get a lot of podcasts. 666 00:37:12,040 --> 00:37:36,040 We'll be back again with another episode soon.