WEBVTT 1 00:00:00.719 --> 00:00:04.480 Oh, artificial intelligence. It is here to stay. And in this conversation, we're 2 00:00:04.480 --> 00:00:08.094 talking about how to use it thoughtfully and mindfully as a 3 00:00:08.094 --> 00:00:11.934 human being interacting with the robots of the Internet. And 4 00:00:11.934 --> 00:00:15.615 today's guest is Avery Schwartz to bring this conversation home. I'm excited 5 00:00:15.615 --> 00:00:17.154 about it. Let's dive in. 6 00:00:23.250 --> 00:00:26.689 You're listening to the Mindful Marketing Podcast. I'm Andrea 7 00:00:26.689 --> 00:00:27.189 Jones. 8 00:00:30.355 --> 00:00:34.195 I've recorded over 300 podcast episodes. Yeah. It's a 9 00:00:34.195 --> 00:00:38.020 lot of podcast episodes, and I've tried a lot of different virtual recording 10 00:00:38.020 --> 00:00:41.640 studios. But my favorite has been Riverside. 11 00:00:42.019 --> 00:00:45.815 Riverside makes their virtual recording studio look so profesh. 12 00:00:46.035 --> 00:00:49.795 My guests love it. Plus, I also low key love 13 00:00:49.795 --> 00:00:53.495 recording YouTube videos in here as well because it's so easy to use. 14 00:00:53.820 --> 00:00:57.660 My team also loves Riverside because it spits out separate audio, video 15 00:00:57.660 --> 00:01:00.960 tracks, making editing easy breezy, lemon squeezy. 16 00:01:01.285 --> 00:01:04.905 And if you want a little magic, they've got this tool called MagicClips, 17 00:01:05.604 --> 00:01:09.205 which uses AI to take your video and turn it into 18 00:01:09.205 --> 00:01:13.000 perfect social media sized videos. I'm talking vertical videos for TikTok 19 00:01:13.000 --> 00:01:16.840 and Instagram, Facebook Reels, all the places you can post these videos with 20 00:01:16.840 --> 00:01:20.685 the captions included, and you don't have to hunt and search for 21 00:01:20.685 --> 00:01:24.064 that perfect clip. So if you wanna try this out for yourself, 22 00:01:24.845 --> 00:01:27.969 click the link that goes with this video. Or if you're listening to the audio 23 00:01:27.969 --> 00:01:31.429 on the podcast, it's in the show notes. Okay? Click that link. 24 00:01:31.570 --> 00:01:35.195 Use the 15% off coupon code. It's Drea, d r 25 00:01:35.195 --> 00:01:38.875 e a, and try Riverside for yourself. Thank you, 26 00:01:38.875 --> 00:01:42.715 Riverside. Avery, welcome to the 27 00:01:42.715 --> 00:01:46.250 show. Oh, thank you so much for having me. I'm really excited to be here. 28 00:01:46.250 --> 00:01:49.290 Yeah. I'm excited to talk all things AI, but I wanna go back to the 29 00:01:49.290 --> 00:01:52.795 beginning for a sec, and talk about you because you 30 00:01:52.795 --> 00:01:56.415 have a really strong history in just bringing 31 00:01:56.475 --> 00:02:00.290 things digital. You've got a best selling book to talk about all things digital and 32 00:02:00.290 --> 00:02:04.130 small business. You, are the tech correspondent for 33 00:02:04.130 --> 00:02:07.845 CTV and all things business. But how did this start? Like, 34 00:02:07.845 --> 00:02:11.685 what's the origin story? Okay. Well, I I am a 35 00:02:11.685 --> 00:02:15.460 person of a particular age. Like, I'm an eighties kid. So that means 36 00:02:15.460 --> 00:02:18.280 that when I was growing up, like, I could not have said to my parents, 37 00:02:18.500 --> 00:02:21.460 mom and dad, when I grow up, I wanna be a web developer. They would 38 00:02:21.460 --> 00:02:25.265 have said, that is not a thing. So I took a 39 00:02:25.265 --> 00:02:29.025 really winding route to get to where I am now. I my first stop was 40 00:02:29.025 --> 00:02:32.840 actually in fine arts school. I have a a fine arts degree. I 41 00:02:32.840 --> 00:02:36.600 have a and then I went to design school after that, so I made my 42 00:02:36.600 --> 00:02:40.394 way into the web through originally through 43 00:02:40.394 --> 00:02:42.955 the root of of web design. So I was a web designer. I had a 44 00:02:42.955 --> 00:02:46.394 web design studio, and then I realized that as as much as I love 45 00:02:46.394 --> 00:02:50.010 designing, my favorite part was actually just working with clients, was 46 00:02:50.010 --> 00:02:53.530 talking with them, figuring out, what do they need. I had a 47 00:02:53.530 --> 00:02:56.985 really deep empathy and still do have really deep empathy for 48 00:02:56.985 --> 00:03:00.745 how, you know, there there's this narrative out there that tech is supposed 49 00:03:00.745 --> 00:03:04.590 to be easy, that tech is supposed to make things easier for people. And I 50 00:03:04.590 --> 00:03:07.970 think in general, overall, it does, but 51 00:03:08.350 --> 00:03:11.630 that doesn't recognize that there is a bit of a struggle, that there's a there's 52 00:03:11.630 --> 00:03:14.655 an adoption cycle. There is a bumpy road. 53 00:03:15.115 --> 00:03:18.635 And my favorite part of of working with my clients was always just kind of 54 00:03:18.635 --> 00:03:21.970 helping them get set up. And then I realized, wait a minute. Like, this is 55 00:03:21.970 --> 00:03:25.430 something I could do. So that's when I I pivoted to being a tech educator. 56 00:03:25.810 --> 00:03:29.590 I own a tech education company where we run practical workshops 57 00:03:29.665 --> 00:03:33.504 for small businesses, charities, and nice people. Like you said, I wrote a 58 00:03:33.504 --> 00:03:37.344 book. I go on TV, but it but it's all it's all in service of 59 00:03:37.344 --> 00:03:40.640 this mission of trying to help nontechnical people 60 00:03:41.019 --> 00:03:44.379 bring technology into their lives, in a way that is 61 00:03:44.379 --> 00:03:48.045 thoughtful, in a way that can help them reach their goals. And 62 00:03:48.045 --> 00:03:51.645 my role is just to be an educator and an interpreter. Yeah. 63 00:03:51.645 --> 00:03:54.865 Ugh. I love everything you do. I 64 00:03:55.400 --> 00:03:58.840 was I feel like one of the first people to buy See You on the 65 00:03:58.840 --> 00:04:02.600 Internet, the book, because I'm just a fan. And I was like, I 66 00:04:02.600 --> 00:04:06.075 want this. Plus, it's, like, gorgeous, and it makes sense with your 67 00:04:06.075 --> 00:04:09.915 design background how pretty everything looks too. So I love that. I did I 68 00:04:09.915 --> 00:04:13.510 did work with the publisher, and my publisher they they sent a a first 69 00:04:13.510 --> 00:04:16.630 draft of the cover of the book, and it was really like, you know those, 70 00:04:16.630 --> 00:04:19.510 like, tech books where it's it's got all these weird, like, 71 00:04:20.294 --> 00:04:24.135 like like, lines and lots of, you know, like like, things looking like 72 00:04:24.135 --> 00:04:27.740 they're flying around, and it's very technical. And I was like, this is 73 00:04:27.900 --> 00:04:31.419 awful. I'm like, no offense, but this is this is not my vibe at all. 74 00:04:31.419 --> 00:04:34.699 I was like, we want it to feel friendly. We want to feel approachable as 75 00:04:34.699 --> 00:04:38.095 like, can we put an emoji on the cover? Can we have more 76 00:04:38.095 --> 00:04:41.855 color? Like, let's use a fun font, like, strip out all of 77 00:04:41.855 --> 00:04:45.615 this, like and and it's always blue. Right? It's like dark navy blue and, 78 00:04:45.615 --> 00:04:46.115 like, 79 00:04:50.850 --> 00:04:54.505 Yeah Yeah. Why the 80 00:04:54.505 --> 00:04:57.645 blue? It's always blue. You're right. It's always blue. Yeah. 81 00:04:58.345 --> 00:05:01.725 Oh, my gosh. I love that. Okay. So switching gears over to AI. 82 00:05:02.180 --> 00:05:05.560 I think that, AI, specifically generative 83 00:05:05.700 --> 00:05:09.460 AI, has shaken us up a lot. Yeah. It has. Yeah. 84 00:05:09.460 --> 00:05:13.135 It really has. And I feel like listeners of 85 00:05:13.135 --> 00:05:16.975 this podcast, we've been talking about the, you know, chat 86 00:05:16.975 --> 00:05:20.740 GPTs of the world for a while. So I wanna just dive right 87 00:05:20.740 --> 00:05:24.580 into the conversation with AI and, like, why does it feel like 88 00:05:24.580 --> 00:05:28.314 sometimes AI is making stuff up? It's like, did you 89 00:05:28.314 --> 00:05:31.854 even check the Internet on this? Don't you have access to this information? 90 00:05:32.235 --> 00:05:35.940 Why did that happen? Okay. So to understand 91 00:05:36.080 --> 00:05:39.600 why generative AI systems like Chat GPT, why they sometimes make things up, 92 00:05:40.800 --> 00:05:44.615 the technical word for it is a hallucination, which I think is hilarious. 93 00:05:44.675 --> 00:05:48.115 It sounds like Chatcheapiti just, like, went on, like, an acid 94 00:05:48.115 --> 00:05:51.715 trip, that it's that it's over in the corner. It's on drugs. It's 95 00:05:51.715 --> 00:05:55.539 hallucinating. But it goes back to to really 96 00:05:55.539 --> 00:05:58.759 understanding what is a a 97 00:05:59.139 --> 00:06:02.419 transformer style generative AI model. And I know I just used a whole bunch of 98 00:06:02.419 --> 00:06:06.075 jargon, but tools like chat GPT are different from 99 00:06:06.075 --> 00:06:09.915 other types of AI. They're different from even other types of machine learning 100 00:06:09.915 --> 00:06:13.750 AI, which has been around us for a a number of years now. But the 101 00:06:13.750 --> 00:06:17.509 way that a tool like ChatGPT works, you've probably heard that it taps 102 00:06:17.509 --> 00:06:21.145 into something called a large language model. And a large 103 00:06:21.145 --> 00:06:24.905 language model is kind of like its knowledge base. It's all of 104 00:06:24.905 --> 00:06:28.425 the information that ChatGPT and other generative AI 105 00:06:28.425 --> 00:06:32.220 systems, have access to. But the thing about a large language model 106 00:06:32.220 --> 00:06:35.980 is that, you know, we've heard that they have basically read every page of the 107 00:06:35.980 --> 00:06:39.775 Internet. It's read all of the publicly available books. It has, you know, 108 00:06:39.775 --> 00:06:43.535 scraped information. It's pulled in all kinds of data to get into that large 109 00:06:43.535 --> 00:06:47.375 language model. But the large language model is 110 00:06:47.375 --> 00:06:51.000 not a search engine. It doesn't behave the way something like Google 111 00:06:51.000 --> 00:06:54.840 does. So the way that Google behaves is that it goes through the 112 00:06:54.840 --> 00:06:58.515 entire Internet with its robots and indexes the 113 00:06:58.515 --> 00:07:02.035 Internet and looks at and it indexes and looks at every single 114 00:07:02.035 --> 00:07:05.720 word and also the connection between web pages, you know, 115 00:07:05.800 --> 00:07:09.160 what's linked to other sites. The way that a large 116 00:07:09.160 --> 00:07:12.840 language model works is that it's looking through all of that original 117 00:07:12.840 --> 00:07:16.275 information, but it's not really looking so much at the words 118 00:07:16.275 --> 00:07:19.715 themselves. It's looking at the connection between the 119 00:07:19.715 --> 00:07:23.520 words. So how often certain words 120 00:07:23.580 --> 00:07:27.020 tend to go together. There are, there are patterns in our 121 00:07:27.020 --> 00:07:30.675 language. There are words that are more, like, based on 122 00:07:30.675 --> 00:07:34.134 on probability that are more likely to appear together. 123 00:07:34.514 --> 00:07:38.055 And so when you interact with a chatbot like chat GPT, 124 00:07:38.890 --> 00:07:42.670 it taps into its large language model, which in this case is called GPT. 125 00:07:43.850 --> 00:07:47.625 And what it's doing is it's it it doesn't 126 00:07:47.625 --> 00:07:50.765 actually know any of the words. It doesn't know what any of the words mean. 127 00:07:51.065 --> 00:07:54.560 It only knows the relationship between the words. So 128 00:07:54.560 --> 00:07:58.000 it's using its understanding of probability, and it's 129 00:07:58.000 --> 00:08:01.599 using its prediction skills to try 130 00:08:01.599 --> 00:08:05.425 and give you the next word in the sentence, the next sentence in 131 00:08:05.425 --> 00:08:08.965 the paragraph, etcetera, that it thinks is probabilistically 132 00:08:10.310 --> 00:08:13.910 most likely to be what you're looking for. So some 133 00:08:13.910 --> 00:08:16.970 people say that Chat GPT is autocomplete on steroids, 134 00:08:17.465 --> 00:08:20.745 and it kind of is. Like, it it it is literally just kind of trying 135 00:08:20.745 --> 00:08:24.425 to autocomplete itself. And so in that way, it is 136 00:08:24.425 --> 00:08:28.000 not it it it it's not self conscious. It's not aware 137 00:08:28.139 --> 00:08:31.900 of the information that it's giving you. So if you ask it for something, 138 00:08:32.220 --> 00:08:35.095 like like, if you're asking it to help you write a caption for a social 139 00:08:35.095 --> 00:08:38.615 media post, that is a situation where there isn't 140 00:08:38.615 --> 00:08:42.375 one right answer. There isn't one factual true 141 00:08:42.375 --> 00:08:46.080 answer. There's a lot of different, ways that you could go with that 142 00:08:46.080 --> 00:08:49.440 caption. Some of them might not be the best quality. You might not like them. 143 00:08:49.440 --> 00:08:53.075 They might not be on brand, but it it it's not looking for a particular 144 00:08:53.075 --> 00:08:56.835 fact. It's when you ask ChatChiBT or another tool like 145 00:08:56.835 --> 00:09:00.610 it to go deep on a particular like like 146 00:09:00.610 --> 00:09:04.370 a vertical, like a particular depth of its knowledge, where 147 00:09:04.370 --> 00:09:08.185 it tries to get into facts, that is where it's more likely to 148 00:09:08.185 --> 00:09:11.945 hallucinate because it it doesn't have although here's the 149 00:09:11.945 --> 00:09:15.560 big caveat. It will happen less and 150 00:09:15.560 --> 00:09:18.779 less as the large language models get more and more sophisticated. 151 00:09:19.160 --> 00:09:22.459 Okay. And it will happen less and less as 152 00:09:22.685 --> 00:09:26.225 generative AI models have something called retrieval 153 00:09:26.365 --> 00:09:30.205 augmented generation, which means they can augment what they generate. So they can change what 154 00:09:30.205 --> 00:09:33.710 they generate by retrieving information from somewhere 155 00:09:33.710 --> 00:09:37.330 else. So, for example, if you've used, a tool like Perplexity, 156 00:09:37.870 --> 00:09:41.635 which I really, really like, Perplexity is using a large 157 00:09:41.635 --> 00:09:45.315 language model, but then it's also searching the live web and 158 00:09:45.315 --> 00:09:48.675 pulling information in that way and kind of smashing those two things 159 00:09:48.675 --> 00:09:52.339 together to give you more accurate answers. And it also gives 160 00:09:52.339 --> 00:09:55.860 you citations. It'll say, we found this on this website, and then you have a 161 00:09:55.860 --> 00:09:59.585 link, and then you can click and you can go and find more. So, 162 00:09:59.585 --> 00:10:03.265 like, once you understand that that's actually how the large language models 163 00:10:03.265 --> 00:10:06.960 work, like, they're not Google, they're not a search engine, then 164 00:10:06.960 --> 00:10:10.720 you were like, oh, okay. So it's actually the thing that makes 165 00:10:10.720 --> 00:10:14.400 it really good is also the thing that makes it 166 00:10:14.400 --> 00:10:18.065 sometimes tell lies. Yeah. Which is so weird. Right? 167 00:10:18.065 --> 00:10:21.904 Yeah. So weird. You know, I, I 168 00:10:21.904 --> 00:10:25.640 actually I use Chat gbt quite a bit, especially for coming up with 169 00:10:25.640 --> 00:10:29.480 ideas. And I have to literally tell it 170 00:10:29.480 --> 00:10:32.860 words not to use because it just latches on to certain words, 171 00:10:33.125 --> 00:10:36.565 like unlock or unleash or, 172 00:10:37.125 --> 00:10:40.245 there there's, like, a few words where I'm, like, can we please stop using this 173 00:10:40.245 --> 00:10:43.959 word? I'm never gonna use it again now because you've ruined it for me. 174 00:10:43.959 --> 00:10:47.720 And I feel like that makes sense because it's just trying to 175 00:10:47.720 --> 00:10:50.915 predict what I wanted to say. Very interesting. 176 00:10:51.535 --> 00:10:54.915 Absolutely. So when we think about this as a marketer, 177 00:10:55.750 --> 00:10:59.509 they're they're fall, people fall into 2 camps. There's a people who are, like, never 178 00:10:59.509 --> 00:11:02.709 in a 1000000 years, and then there's people who are, like, I use it every 179 00:11:02.709 --> 00:11:06.365 day. And I feel like I'm somewhere in the middle, but then I feel like 180 00:11:06.365 --> 00:11:10.045 I have to always explain. Like, I use it every day, but I don't 181 00:11:10.045 --> 00:11:13.670 do this because it feels like I'm cheating. Like, I'm not cheating, but I 182 00:11:13.670 --> 00:11:17.029 am using this as a shortcut. Why does it feel like we're cheating when we're 183 00:11:17.029 --> 00:11:20.795 using these tools? Yeah. I mean, there's a real ethical 184 00:11:21.015 --> 00:11:24.774 dilemma here and and quite a paradox of I I 185 00:11:24.774 --> 00:11:28.400 think the the more you use generative AI and the 186 00:11:28.400 --> 00:11:31.780 more time you put into it, the more time you put into training a system 187 00:11:31.920 --> 00:11:35.520 to get to know you, to be able to write like you, you you give 188 00:11:35.520 --> 00:11:39.005 it that feedback, like, hey. In my case, it's the word 189 00:11:39.065 --> 00:11:42.825 demystify, the word wield. It always wants to 190 00:11:42.825 --> 00:11:46.230 wield things. And I'm like, no. We don't we don't use that verb in my 191 00:11:46.230 --> 00:11:49.910 life. So it's like teaching it which words to use, what 192 00:11:49.910 --> 00:11:53.425 tone, etcetera. Because generative AI is a machine learning 193 00:11:53.425 --> 00:11:57.125 system and machine learning systems get better. They literally 194 00:11:57.185 --> 00:12:00.920 learn and they learn through trial and error, through feedback, and through 195 00:12:00.920 --> 00:12:04.519 lots of practice. So the more you train it, the better it gets. So 196 00:12:04.519 --> 00:12:08.315 anybody that's that's invested a bit of that time, it it is 197 00:12:08.315 --> 00:12:11.995 undeniable that this is a tool that can be helpful. This is a 198 00:12:11.995 --> 00:12:15.750 tool that can supercharge your productivity. But then you have that moment 199 00:12:15.750 --> 00:12:18.230 of, like, you know, if you're if you're someone that, you know, has any sense 200 00:12:18.230 --> 00:12:21.830 of integrity, you've heard that, okay, maybe it was trained on on 201 00:12:21.830 --> 00:12:25.505 copyrighted data. There's a bit of a gray area there. You're thinking, 202 00:12:25.565 --> 00:12:28.925 okay. I'm not really creating this. Can I say that it's something that I actually 203 00:12:28.925 --> 00:12:32.704 made? Is that ethical? There's also some environmental 204 00:12:33.084 --> 00:12:36.579 concerns with generative AI. The actual 205 00:12:36.959 --> 00:12:40.500 computing power required to run these systems 206 00:12:40.885 --> 00:12:44.725 is immense. Wow. And when you're using computing power, that means you 207 00:12:44.725 --> 00:12:48.560 might be using more fossil fuels, you might be using more electricity, you know, 208 00:12:48.560 --> 00:12:51.540 etcetera, etcetera. And so there's an environmental impact there as well. 209 00:12:52.880 --> 00:12:56.205 So it's it's kind of like like, what is a thoughtful marketer to do? 210 00:12:57.165 --> 00:13:00.925 You know, you you want to potentially do better work. You 211 00:13:00.925 --> 00:13:04.540 want to do better work in less time, or with that extra 212 00:13:04.540 --> 00:13:07.360 time, you could just do more work, I guess. So 213 00:13:09.500 --> 00:13:12.995 is is it cheating to use Chat GPT? This is this is one of those 214 00:13:12.995 --> 00:13:16.595 conversations that, I think is best had with, like, a glass of wine because it 215 00:13:16.595 --> 00:13:20.250 it almost goes a little bit more into a, like, a theoretical conversation. But I 216 00:13:20.250 --> 00:13:23.050 was talking with someone recently, and they said that they felt like cheating. And I 217 00:13:23.050 --> 00:13:26.650 said, okay, let's unpack that a little bit. Is it 218 00:13:26.650 --> 00:13:30.425 cheating? Use a calculator. And we talked about that a little bit. 219 00:13:30.425 --> 00:13:33.964 And I said, I I have a daughter who's in middle school. And right now 220 00:13:34.264 --> 00:13:37.225 she is she's at the point now where she's allowed to use a calculator in 221 00:13:37.225 --> 00:13:41.069 math class. But for a number of grades, she wasn't allowed to use a calculator 222 00:13:41.449 --> 00:13:45.149 because it was very important that she knew how to manually do 223 00:13:45.475 --> 00:13:48.835 long division because she was understanding the 224 00:13:48.835 --> 00:13:52.360 mathematical concepts of addition, subtraction, multiplication, and division. 225 00:13:52.920 --> 00:13:55.420 So in that case, it would be very bad for her to use a calculator, 226 00:13:55.640 --> 00:13:59.480 but I use a calculator all the time because it frees 227 00:13:59.480 --> 00:14:02.875 up my time and my mental energy to do something that's, you know, 228 00:14:02.875 --> 00:14:06.175 more intensive. Right? Yeah. So 229 00:14:06.875 --> 00:14:10.175 the thing about ChatChiPT though is that if you 230 00:14:10.235 --> 00:14:14.010 completely outsource everything to it, if you 231 00:14:14.010 --> 00:14:17.370 if you if you've trained it really well so that it can write just like 232 00:14:17.370 --> 00:14:21.130 you and you say, write me 5 blog posts, and then you just kind 233 00:14:21.130 --> 00:14:24.654 of, I don't know, you you go and do something else and you come back 234 00:14:24.654 --> 00:14:27.535 and you're like, yeah. Yeah. That looks good. Good. Good. You copy and paste. You 235 00:14:27.535 --> 00:14:30.254 publish. You put it right up on the web. You let it do everything for 236 00:14:30.254 --> 00:14:33.650 you. I think that's when it feels like 237 00:14:33.650 --> 00:14:37.330 cheating because there there's 2 things. There's 238 00:14:37.570 --> 00:14:41.025 are you potentially, cheating by getting to the final 239 00:14:41.985 --> 00:14:45.345 product without you being the one that's going there? So it's like, is that ethical 240 00:14:45.345 --> 00:14:48.625 to then turn around and, you know, sell it to your clients and to and 241 00:14:48.625 --> 00:14:52.420 to bill people for for those hours? But what I think you're really cheating 242 00:14:52.420 --> 00:14:56.260 is you're probably cheating yourself. Because, you 243 00:14:56.260 --> 00:15:00.100 know, I'm a writer. Like, I, I actually do like to write. I wrote 244 00:15:00.100 --> 00:15:03.715 a book, and I and I wrote the book myself. I didn't write it with 245 00:15:03.715 --> 00:15:07.555 artificial intelligence. When you're 246 00:15:07.555 --> 00:15:11.060 going through the process of writing, you're going through the process of 247 00:15:11.060 --> 00:15:14.759 thinking. And it's hard. Writing 248 00:15:14.820 --> 00:15:18.255 is hard. And it's hard because you're 249 00:15:18.255 --> 00:15:21.774 doing more than just stringing together words in a 250 00:15:21.774 --> 00:15:25.490 sentence. You're working out your thoughts. You know, you're figuring out, 251 00:15:25.490 --> 00:15:29.089 okay. Does this make sense here? Is this a logical argument? Do I need 252 00:15:29.089 --> 00:15:32.769 more examples over here? If I'm going to counteract my argument, is 253 00:15:32.769 --> 00:15:36.195 this flowing? Does do I need to to have something here that 254 00:15:36.195 --> 00:15:39.795 will, wrap this up? And, you know, there's different types of writing. There's 255 00:15:39.795 --> 00:15:43.440 writing an email to my sister-in-law. Who cares? ChatChapiti 256 00:15:43.660 --> 00:15:47.100 can do that. I don't care. But writing, you know, a 257 00:15:47.100 --> 00:15:50.620 detailed blog post about, you know, thoughts in the industry or 258 00:15:50.620 --> 00:15:54.045 something. That's the point where 259 00:15:54.425 --> 00:15:58.105 I might use AI tools to help me. They might help me 260 00:15:58.105 --> 00:16:01.700 to edit a little bit. They might help me to provide a, you know, 261 00:16:01.700 --> 00:16:05.540 a counterbalance to my argument. But I I personally will not 262 00:16:05.540 --> 00:16:09.380 let them just do everything Yeah. Because it's cheating me out 263 00:16:09.380 --> 00:16:13.065 of the thought process. Yeah. Yeah. And I think part of it 264 00:16:13.065 --> 00:16:16.905 too is this idea of, creating original work 265 00:16:16.905 --> 00:16:20.550 as well. I love that I can use chat 266 00:16:20.550 --> 00:16:24.070 gbt to help me with things like, you know, how can I wrap this up, 267 00:16:24.070 --> 00:16:27.750 or or what's a good title for this, or what's a hook 268 00:16:27.750 --> 00:16:31.125 for this? But when I think about contributing 269 00:16:32.464 --> 00:16:35.125 interesting stories, adding flavor, 270 00:16:36.020 --> 00:16:39.380 adding perspective. Those are things that I don't 271 00:16:39.380 --> 00:16:42.920 think ChatGPT can do nor should do. 272 00:16:43.515 --> 00:16:46.575 And I've even been playing around with Gemini recently as well. 273 00:16:47.195 --> 00:16:50.820 Same thing. I don't I don't want those tools to do those things 274 00:16:50.820 --> 00:16:54.420 for me personally because it does feel like cheating. But I do wanted to 275 00:16:55.060 --> 00:16:58.180 one of the things I did recently was, like, here are my top performing podcast 276 00:16:58.180 --> 00:17:01.954 episodes. What are 10 more things I can talk about on the on the podcast? 277 00:17:01.954 --> 00:17:05.155 And it gave me 10 ideas that were terrible and one that was great. And 278 00:17:05.155 --> 00:17:08.680 I was like, okay. Good. So I feel like there's some there's 279 00:17:08.680 --> 00:17:12.359 some the calculator shortcut there where I can go relieve my 280 00:17:12.359 --> 00:17:16.054 brain of thinking about this for 1 minute so that I can focus on the 281 00:17:16.054 --> 00:17:19.755 things that kind of add more context there, which I love. I love. 282 00:17:20.375 --> 00:17:24.039 Yeah. There's, there there's a sense of of getting 283 00:17:24.039 --> 00:17:27.799 to know what the tools are really good at and getting to know what they're 284 00:17:27.799 --> 00:17:31.559 not good at at all and then picking and choosing. Like like, treat 285 00:17:31.559 --> 00:17:34.695 them like a coworker. You'd be like, hey. I know what I'm good at, so 286 00:17:34.695 --> 00:17:37.095 I'm gonna focus on that part. I know what you're good at, so I'm gonna 287 00:17:37.095 --> 00:17:40.855 let you, you know, run with this part. I I used Chetche Bt 288 00:17:40.855 --> 00:17:44.540 recently to do some, data analysis. I I took an Excel 289 00:17:44.540 --> 00:17:47.760 spreadsheet that had something like 3,000 290 00:17:48.060 --> 00:17:51.894 rows of feedback from a program that we had run-in my company. 291 00:17:52.274 --> 00:17:56.115 And I took out all of the the names and email addresses of people because 292 00:17:56.115 --> 00:17:59.780 I don't wanna put confidential information into chat gpt. And I 293 00:17:59.780 --> 00:18:02.900 put it into chat gpt, and I said, read through this all of this, 294 00:18:03.220 --> 00:18:06.925 feedback and do a sentiment analysis. And it 295 00:18:06.925 --> 00:18:10.225 did it it did a really good job kind of pulling out key ideas. 296 00:18:10.365 --> 00:18:14.045 Now that was me playing to its strengths, and then I played 297 00:18:14.045 --> 00:18:17.750 to my strengths by then taking that and then 298 00:18:17.750 --> 00:18:21.350 crafting it into a proposal for a new program that I wanna 299 00:18:21.350 --> 00:18:25.075 run. So I wouldn't let ChatcheBT write the the proposal, but I would 300 00:18:25.075 --> 00:18:28.595 let Chachi Bt do the sentiment analysis for me because that would have oh my 301 00:18:28.595 --> 00:18:32.275 gosh. That would have taken me probably 6 or 7 hours Yeah. To read 302 00:18:32.275 --> 00:18:35.910 through all of that, to copy and paste, to to highlight, 303 00:18:36.530 --> 00:18:40.130 you know, commonalities and common threads. Like, it did it in, you know, 3 304 00:18:40.130 --> 00:18:43.745 minutes or something. Yeah. So. Brilliant. Ugh. I 305 00:18:43.745 --> 00:18:47.425 love that. And now I'm like, oh, I need to do this. Yeah. It's 306 00:18:47.425 --> 00:18:51.240 pretty good at it. Yeah. Surprisingly good. I love that. I love that. Okay. We're 307 00:18:51.240 --> 00:18:54.440 gonna take a quick break. And when we come back, we have more AI questions 308 00:18:54.440 --> 00:18:58.280 coming at you. We are long past the good old days 309 00:18:58.280 --> 00:19:01.715 of MySpace. And despite what those Facebook ads are 310 00:19:01.715 --> 00:19:05.475 touting as the magical AI solution antithesis to 311 00:19:05.475 --> 00:19:09.130 social media, social media is not dead. It 312 00:19:09.130 --> 00:19:12.730 is simply evolving, and that's what we're exploring at the 313 00:19:12.730 --> 00:19:16.205 Social Media Day Summit. Join me and my fellow 314 00:19:16.205 --> 00:19:19.585 experts on June 30th as we dive into what's working 315 00:19:19.805 --> 00:19:23.560 for social media marketers here now and today, 316 00:19:23.700 --> 00:19:26.840 we're exploring innovative strategies and timeless 317 00:19:27.140 --> 00:19:30.500 tactics that you can use for both yourself and your 318 00:19:30.500 --> 00:19:33.965 clients. Grab your ticket today for $10, and I'll see you 319 00:19:33.965 --> 00:19:37.325 there. And we're back. So 320 00:19:37.325 --> 00:19:41.080 you talked a little bit about this already, and I wanna dive further 321 00:19:41.140 --> 00:19:44.840 on the topic of generative AI and copyright 322 00:19:44.980 --> 00:19:48.635 and who owns all of this. Because the latest update 323 00:19:48.635 --> 00:19:52.155 to chat g p t, you know, it can create photos and videos 324 00:19:52.155 --> 00:19:55.940 and design things for you. But do we 325 00:19:55.940 --> 00:19:59.620 own these things that it's creating? Short 326 00:19:59.620 --> 00:20:03.434 answer is probably not 327 00:20:03.434 --> 00:20:07.195 or no, depending on where you live in the world. So I'm 328 00:20:07.195 --> 00:20:10.894 in Canada, and in Canada, our copyright law does not 329 00:20:11.240 --> 00:20:14.840 recognize generative AI. Like, our copyright law predates generative 330 00:20:14.840 --> 00:20:18.460 AI. So Innovation Science and Economic Development Canada 331 00:20:18.975 --> 00:20:22.815 has, just wrapped up a consultation process with experts. They're taking 332 00:20:22.815 --> 00:20:26.255 all of the results from that. They're going to be advising and 333 00:20:26.255 --> 00:20:30.050 creating policy advice to update copyright law in Canada. So, 334 00:20:30.450 --> 00:20:34.210 TLDR, it's coming. You know, in Canada, we have an AI 335 00:20:34.210 --> 00:20:37.995 act that is being worked on at the moment, some copyright updates. But 336 00:20:37.995 --> 00:20:41.435 in America, the US copyright office has 337 00:20:41.435 --> 00:20:44.955 said, no, what you create with generative AI 338 00:20:44.955 --> 00:20:48.780 cannot be, copywritten means 339 00:20:48.780 --> 00:20:51.660 like writing copy for a website. I don't think you use that word if you're 340 00:20:51.660 --> 00:20:54.725 talking about copywriting something. So let me let me rephrase that sentence. 341 00:20:55.285 --> 00:20:58.865 The US patent office has said or sorry. The the 342 00:20:59.445 --> 00:21:03.205 you know what I mean. The, the federal copyright office in the United 343 00:21:03.205 --> 00:21:06.960 States has said that, that anything that's created with generative AI 344 00:21:07.500 --> 00:21:11.215 cannot be applied for a copyright, it cannot be applied for a trademark. 345 00:21:12.095 --> 00:21:15.135 So this is this is a big thing. If you use generative AI tools to 346 00:21:15.135 --> 00:21:18.835 make a logo, if you use generative AI tools to make an image 347 00:21:18.975 --> 00:21:22.610 that, you know, you wanna put on social media that will represent a 348 00:21:22.610 --> 00:21:26.130 program that you're running, you know, knock yourself out. But if you wanna turn around 349 00:21:26.130 --> 00:21:29.735 and then say that you own copyright to that image or if you use 350 00:21:29.735 --> 00:21:33.515 chat gbt to write a book or a white paper 351 00:21:33.575 --> 00:21:37.335 or something, it gets really tricky. You know, this thought of, 352 00:21:37.335 --> 00:21:41.080 well, then who actually owns that intellectual property, who owns the 353 00:21:41.080 --> 00:21:44.760 copyright of that material. So, in general, the answer 354 00:21:44.760 --> 00:21:48.475 is no, that that you basically either you probably 355 00:21:48.534 --> 00:21:51.894 don't or you definitely don't, have the copyright to anything you 356 00:21:51.894 --> 00:21:55.650 create. I think it's so important to emphasize that because 357 00:21:55.710 --> 00:21:59.390 this is where that that cheating feeling comes from 358 00:21:59.390 --> 00:22:03.205 because you have to be very careful, especially business owners, marketers, 359 00:22:03.665 --> 00:22:07.105 as you know what I mean? It's for your clients even. Like, there is some 360 00:22:07.105 --> 00:22:10.480 gray area and then there's like, oh, no. Definitely don't do this at 361 00:22:10.480 --> 00:22:14.320 all. And logos, any written copy 362 00:22:14.320 --> 00:22:18.095 that's going on websites and things like that, I just would stay away 363 00:22:18.095 --> 00:22:21.315 from it if I were you because you can get used some serious hot water. 364 00:22:22.095 --> 00:22:25.520 But I wanna talk specifically about, like, client stuff and, 365 00:22:25.520 --> 00:22:29.280 especially, you you mentioned earlier you removed names and email 366 00:22:29.280 --> 00:22:32.260 addresses to not feed it to the machine. 367 00:22:33.040 --> 00:22:36.135 Yeah. Because it's not safe. Right? Can you talk a little bit more about that 368 00:22:36.135 --> 00:22:39.735 as well? Yeah. So, you know, this is where I started to bump up 369 00:22:39.735 --> 00:22:43.090 against my knowledge of, you know, I am not a computer scientist. I I am 370 00:22:43.090 --> 00:22:46.789 not a an AI researcher. I'm also not a, I'm not a cybersecurity 371 00:22:47.169 --> 00:22:50.755 specialist, but I am an information sponge, 372 00:22:50.755 --> 00:22:54.595 and I read everything I possibly can. And I talk to a lot of people, 373 00:22:54.595 --> 00:22:58.275 and I listen as much as I can. And what 374 00:22:58.275 --> 00:23:02.009 I hear out there in the in the kerfuffle, in the 375 00:23:02.009 --> 00:23:05.769 chatter, is that we need to be a little bit concerned and and a 376 00:23:05.769 --> 00:23:09.135 little bit, we we definitely need to be a 377 00:23:09.135 --> 00:23:12.975 little safety conscious about putting any 378 00:23:12.975 --> 00:23:16.370 confidential information into chat GPT. Now chat 379 00:23:16.370 --> 00:23:19.730 GPT and and and all the other systems too, Microsoft Copilot, Google 380 00:23:19.730 --> 00:23:23.570 Gemini, Perplexity, Claude, Meta AI, all of 381 00:23:23.570 --> 00:23:27.335 them, same thing. These systems already know a heck of a lot about me. 382 00:23:27.335 --> 00:23:31.175 Like, anything that that you could find out about me by googling me, I 383 00:23:31.175 --> 00:23:34.730 figure is fair game. So my name, what I do, my 384 00:23:34.730 --> 00:23:38.350 business, etcetera, fine. I don't I don't pull that out. 385 00:23:39.290 --> 00:23:42.975 But anytime that I'm entering any information that has to do with my clients 386 00:23:43.355 --> 00:23:46.955 or people that I work with, that's where, if I 387 00:23:46.955 --> 00:23:50.550 can, I will strip out names, I'll strip out email addresses, 388 00:23:51.010 --> 00:23:54.770 anything that could be, personally identifiable? And that's 389 00:23:54.850 --> 00:23:58.425 it's it's just because the way that these systems 390 00:23:58.425 --> 00:24:01.945 work is a black box. We 391 00:24:01.945 --> 00:24:05.660 know in general how they work, but we don't know 392 00:24:05.960 --> 00:24:09.640 exactly how they come to the exact answer that they come 393 00:24:09.640 --> 00:24:13.400 to. And so when you're feeding information in and and 394 00:24:13.400 --> 00:24:17.225 I know Microsoft and OpenAI and Google, they have all 395 00:24:17.225 --> 00:24:21.065 said that they don't take your material that 396 00:24:21.065 --> 00:24:24.230 you type into the system, that they are not taking that and using it as 397 00:24:24.230 --> 00:24:28.070 their training data for their large language model. But 398 00:24:28.070 --> 00:24:31.750 there's also real world examples, and if you look it up, you'll find them, real 399 00:24:31.750 --> 00:24:35.535 world examples where people will interact with these systems, and all of a sudden, it 400 00:24:35.535 --> 00:24:39.375 it starts spitting out information that it it got from somewhere. Mhmm. 401 00:24:40.095 --> 00:24:43.530 So I I am a bit concerned about putting any 402 00:24:43.530 --> 00:24:47.210 sensitive information in there. Instead, I'll just pull it out or 403 00:24:47.210 --> 00:24:50.970 I'll rephrase. But that being said, there's still a heck of a lot you can 404 00:24:50.970 --> 00:24:54.635 do with these systems without, you you know, like, I'm not gonna I'm not gonna 405 00:24:54.635 --> 00:24:58.395 put, like, all of my client payment information or anything like 406 00:24:58.395 --> 00:25:01.480 that. I mean, do it. Like, you know, I would never do that. Yeah. Yeah. 407 00:25:01.480 --> 00:25:05.020 I mean, the big guys, they have a lot of information on us already 408 00:25:05.080 --> 00:25:08.920 too. So I feel like Yeah. I feel like the the the least 409 00:25:08.920 --> 00:25:12.725 amount I can give it, the better. One of my good friends works 410 00:25:12.725 --> 00:25:16.565 in data privacy now on the government level. And the stuff she 411 00:25:16.565 --> 00:25:20.270 tells me, I'm like, o m g. Like, they know a lot about 412 00:25:20.270 --> 00:25:24.030 us. And some of the stuff we're, like, volunteering, just 413 00:25:24.030 --> 00:25:27.544 to share. So anything that we can be very careful and 414 00:25:27.544 --> 00:25:31.225 mindful of how we give them that information is, I think, 415 00:25:31.225 --> 00:25:34.664 key here, especially with other people's stuff. 416 00:25:34.664 --> 00:25:38.290 Like, if you're a marketer and you're working on your your client 417 00:25:38.350 --> 00:25:42.190 sentiment analysis, for instance, on, you know, reviews that they've had, remove those 418 00:25:42.190 --> 00:25:45.975 names, remove those email addresses for sure because you don't want that 419 00:25:45.975 --> 00:25:48.794 information being leaked out there. I love that. 420 00:25:49.894 --> 00:25:53.650 Okay. So on the same tone of, like, 421 00:25:53.650 --> 00:25:57.410 using generative AI, specifically as a marketer, 422 00:25:57.410 --> 00:26:01.145 as a professional, even, you know, I'm thinking about copywriters, I'm thinking 423 00:26:01.145 --> 00:26:04.125 about web developers, I'm thinking about designers. 424 00:26:05.385 --> 00:26:08.830 Do we need to disclose this for our 425 00:26:08.830 --> 00:26:12.350 clients? Yeah. So that's another big 426 00:26:12.350 --> 00:26:15.894 question that you can come at from a bunch of different angles. So let's start 427 00:26:15.894 --> 00:26:19.414 with the regulatory response. And that is 428 00:26:19.414 --> 00:26:23.170 that in Canada and in America right now, you are 429 00:26:23.170 --> 00:26:26.710 not legally required to disclose the use of generative 430 00:26:26.770 --> 00:26:30.230 AI, like, from, like, like, the federal level at either of those companies. 431 00:26:30.505 --> 00:26:34.184 If you are in the European Union, you do have to disclose the 432 00:26:34.184 --> 00:26:37.840 use because the EU has passed their AI act, which 433 00:26:37.840 --> 00:26:41.360 does have different parameters for the disclosure of use of 434 00:26:41.360 --> 00:26:45.120 AI. Like I said, there's one coming in Canada, there's one coming in 435 00:26:45.120 --> 00:26:48.835 America, all of this stuff, you know, this is typical. Technology moves faster than 436 00:26:48.835 --> 00:26:51.895 the law does, so the law has to, you know, catch up. 437 00:26:52.595 --> 00:26:56.330 But that's just, you know, the the basics. Right? Like 438 00:26:56.330 --> 00:27:00.050 like, regulatory, you know, federal law. Do you have to that 439 00:27:00.090 --> 00:27:03.385 now you've got to kind of get into this this question of, you know, 440 00:27:03.465 --> 00:27:06.925 ethically, do you need to tell your clients? Now 441 00:27:07.225 --> 00:27:11.000 I would assume that your clients know that you use certain tools, 442 00:27:11.080 --> 00:27:14.780 that when you are creating an image for them, 443 00:27:15.080 --> 00:27:18.735 or or some graphic design for their social media posts, or 444 00:27:18.735 --> 00:27:22.195 for their website that you are not, you know, personally painting 445 00:27:22.815 --> 00:27:26.575 a photo, you know, an image that will become, you know, their their Facebook 446 00:27:26.575 --> 00:27:30.389 post, that you are using tools. You're using Photoshop. You might 447 00:27:30.389 --> 00:27:34.149 be using Canva. You might be using Adobe Premiere to cut some 448 00:27:34.149 --> 00:27:37.975 videos. They know that you're using tools and I think it's it's 449 00:27:37.975 --> 00:27:40.075 generally considered okay to use tools. 450 00:27:41.735 --> 00:27:45.275 I think that the real dilemma lies in, 451 00:27:46.600 --> 00:27:50.440 the value of the work. And this is 452 00:27:50.440 --> 00:27:54.235 the thing that becomes really, really thorny, which is that, you've got to 453 00:27:54.235 --> 00:27:57.835 think about why and I know we're we're getting a bit theoretical here, but 454 00:27:57.835 --> 00:28:01.650 but why are your clients hiring you? What is the 455 00:28:01.650 --> 00:28:05.110 value? Is the value the final product? 456 00:28:05.730 --> 00:28:09.135 Is the value the getting to the final product? Is the 457 00:28:09.135 --> 00:28:12.895 value that they, they love working with you because 458 00:28:12.895 --> 00:28:16.435 you're so pleasant to work with, is the value that 459 00:28:18.280 --> 00:28:22.060 you can, you know, do some analysis and some strategy. 460 00:28:23.400 --> 00:28:26.940 But if the value is purely in the final product, 461 00:28:27.455 --> 00:28:31.294 then we've got a problem. Mhmm. Because we've got robots that can get 462 00:28:31.294 --> 00:28:34.840 to the final product. And we got robots that don't sleep, they don't eat. Right 463 00:28:34.840 --> 00:28:38.120 now, a lot of them are free or very, very low cost. So it's gonna 464 00:28:38.120 --> 00:28:41.559 be a race to the bottom really, really fast in terms of just 465 00:28:41.559 --> 00:28:45.235 competition of, you know, how can a human compete with a robot if all you 466 00:28:45.235 --> 00:28:48.995 care about is the final product? So that right there, I think, 467 00:28:48.995 --> 00:28:52.390 is why some people are hesitant to tell their 468 00:28:52.390 --> 00:28:55.910 clients that they're using generative AI because it might open up a 469 00:28:55.910 --> 00:28:59.750 conversation of the client, and hopefully your clients are much more respectful than 470 00:28:59.750 --> 00:29:02.775 this. But the client could look you in the eye and say, well, what am 471 00:29:02.775 --> 00:29:06.455 I paying you for then? Yeah. You know, and you better have a really good 472 00:29:06.455 --> 00:29:09.990 answer to that question. And hopefully, you are 473 00:29:09.990 --> 00:29:13.830 bringing so much more value than just the, like, well, I 474 00:29:13.830 --> 00:29:17.175 sliced and diced and made 10 Instagram posts for you. Yeah. 475 00:29:17.495 --> 00:29:21.335 That that there's more to it than that. But if you can't say that there's 476 00:29:21.335 --> 00:29:24.990 more to it than that, then, like, I mean, they might fire 477 00:29:24.990 --> 00:29:28.830 you. And frankly, I would probably fire you, too. Yeah. You know, everybody wants 478 00:29:28.830 --> 00:29:32.125 to save money, and if the if the robot can do just as good of 479 00:29:32.125 --> 00:29:35.405 a job, then why not let the robot do it? So it's, it's, it's a 480 00:29:35.405 --> 00:29:39.070 real, it's a really meaty question, that 481 00:29:39.070 --> 00:29:42.530 really gets into as marketers, what is our value? 482 00:29:43.230 --> 00:29:46.865 And I I strongly believe that it is not solely in the final 483 00:29:46.865 --> 00:29:50.485 product, but we have to get better at communicating 484 00:29:50.705 --> 00:29:54.549 that value to our clients so they don't assume 485 00:29:54.549 --> 00:29:58.390 that the value is in the final product. You have my wheels spinning over 486 00:29:58.390 --> 00:30:00.809 here because this is something I've been trying to 487 00:30:02.265 --> 00:30:06.045 conceptualize for myself and and especially, I mentor, 488 00:30:06.105 --> 00:30:09.785 you know, marketers, social media managers. And this question does come 489 00:30:09.785 --> 00:30:13.370 up, often, especially for clients who are more 490 00:30:13.370 --> 00:30:17.210 budget conscious and, you know, scrappier, and they're trying to 491 00:30:17.210 --> 00:30:21.054 do the thing. And frankly, there is a lot that these tools 492 00:30:21.054 --> 00:30:24.735 can do. I like to compare it to the Photoshop versus 493 00:30:24.735 --> 00:30:28.255 Canva conversation where, you know, go back 10 years 494 00:30:28.255 --> 00:30:31.710 ago. If you said as a marketer that you primarily 495 00:30:31.929 --> 00:30:35.690 use Canva as your design tool, people were like, what am I paying you 496 00:30:35.690 --> 00:30:39.315 for? Right? Like, oh my gosh. You don't use Photoshop? You don't use the Adobe 497 00:30:39.315 --> 00:30:42.855 Suite? Whereas now, it is very acceptable 498 00:30:43.394 --> 00:30:47.090 and encouraged even to use Canva as the tool for, like, 499 00:30:47.090 --> 00:30:50.610 a very easy way to, like, whip up something really quickly that doesn't need all 500 00:30:50.610 --> 00:30:54.304 the technical aspects of, like, an Adobe product. So I do 501 00:30:54.304 --> 00:30:57.664 think that there's a conversation there. And I also hear a little inkling of, like, 502 00:30:57.664 --> 00:31:01.320 an ethical conversation around what are we telling our 503 00:31:01.320 --> 00:31:05.159 clients, how are we presenting this information. Because if we are 504 00:31:05.159 --> 00:31:08.779 using it, we definitely do need to disclose it, I think, from an ethical 505 00:31:08.840 --> 00:31:12.365 perspective. Like, you know, having, we're working on this with our 506 00:31:12.365 --> 00:31:15.805 agency right now, having, like, a one cheater or, like, even a page on our 507 00:31:15.805 --> 00:31:19.570 website that's, like, here's how we use AI tools in the 508 00:31:19.570 --> 00:31:23.330 business. Yeah. And so I think that the there's, like, a lot of things 509 00:31:23.330 --> 00:31:27.110 at play there. And like you said, it's moving so much faster than, 510 00:31:27.745 --> 00:31:31.105 the laws can keep up with it. So it we're kinda like the wild, wild 511 00:31:31.105 --> 00:31:34.645 west out here making our own rules as we go before the laws catch up. 512 00:31:35.100 --> 00:31:38.460 Yeah. Absolutely. And it's one of those things where, like, you know, this could be 513 00:31:38.460 --> 00:31:42.220 a completely separate podcast episode, and I'm sure you've had this this episode and 514 00:31:42.220 --> 00:31:46.015 will continue to have it. But, this kind of gets you into, again, that big 515 00:31:46.015 --> 00:31:49.055 conversation about what is your value, what is the value that you bring to your 516 00:31:49.055 --> 00:31:52.460 clients, and, also, how is your fee structure 517 00:31:52.460 --> 00:31:56.300 reflecting that value? And is your fee structure aligned with 518 00:31:56.300 --> 00:31:59.820 that value? Because, and now here come here comes some Avery opinions for 519 00:31:59.820 --> 00:32:03.535 you. I don't do a lot of client services these days, but 520 00:32:03.535 --> 00:32:07.215 I did for many, many years, and I was never a 521 00:32:07.215 --> 00:32:11.030 fan of the hourly rate. I am not a fan of the billable hour. I 522 00:32:11.030 --> 00:32:14.870 am not in the business of selling time. I am in the business of selling 523 00:32:14.870 --> 00:32:18.705 expertise. I'm in the business of selling ease. I'm in the business of 524 00:32:18.705 --> 00:32:22.385 selling strategy. I am a bit in the business of selling the final 525 00:32:22.385 --> 00:32:26.140 product because it's gonna be pretty good too. But if 526 00:32:26.140 --> 00:32:29.580 you are doing a billable hour with your clients and 527 00:32:29.580 --> 00:32:33.260 if using generative AI literally gets more 528 00:32:33.260 --> 00:32:36.995 done in less time, you've got a problem. Yeah. Either 529 00:32:36.995 --> 00:32:40.835 your billable hour like, your rate has to go up or you're going to 530 00:32:40.835 --> 00:32:44.215 have to do something to add value in that hour 531 00:32:44.490 --> 00:32:47.450 because you may actually be able to create. And and this this is what a 532 00:32:47.450 --> 00:32:51.290 lot of studies are showing that on average, people that use generative AI 533 00:32:51.290 --> 00:32:55.105 tools are saving between 25 45 percent of their 534 00:32:55.105 --> 00:32:57.985 time. Wow. And so what are you gonna do with that time? Are you gonna 535 00:32:58.385 --> 00:33:01.640 and this gets into a big conversation of, like, as a society, what are we 536 00:33:01.640 --> 00:33:05.240 going to do with that time? Unfortunately, because we're all just 537 00:33:05.240 --> 00:33:09.080 capitalists. We're probably just gonna do more more work with that 538 00:33:09.080 --> 00:33:12.695 time. But wouldn't it be beautiful if we did less work, if we if we 539 00:33:12.695 --> 00:33:15.095 took that time and that was time that all of a sudden we all just 540 00:33:15.095 --> 00:33:18.535 embraced a 4 day work week, or we, you know, spent more time with our 541 00:33:18.535 --> 00:33:21.929 family and our our and our communities and our pets and and the things that 542 00:33:21.929 --> 00:33:25.610 we doing the things that we love. But we'll probably 543 00:33:25.610 --> 00:33:29.285 just do more work. I was just so there the 544 00:33:29.285 --> 00:33:32.725 podcast episode that's coming out next month, we're talking 545 00:33:32.725 --> 00:33:36.220 about systems and automations. And 546 00:33:36.220 --> 00:33:40.059 this is part of that conversation, which is, you know, we're 547 00:33:40.140 --> 00:33:43.919 we are in the capitalistic society of, like, improving things, 548 00:33:44.115 --> 00:33:47.635 But at the same time, there's something so sweet 549 00:33:47.635 --> 00:33:51.235 about having the hobbies. You know, like, having the time to do the 550 00:33:51.235 --> 00:33:54.940 things. Like, we we just do this because we like to frolic and enjoy it. 551 00:33:54.940 --> 00:33:58.780 And so Yeah. I I think that, hopefully, we can also do 552 00:33:58.780 --> 00:34:02.320 that with the power of technology and how quickly it's moving, 553 00:34:02.585 --> 00:34:06.265 you know, all this free time that we're saving. You know, we can read 554 00:34:06.265 --> 00:34:09.930 some romance novels and, like, enjoy a glass of wine in the sunshine. How about 555 00:34:09.930 --> 00:34:13.610 that? That's right. I love it. I love it. I'll be out in my 556 00:34:13.610 --> 00:34:17.450 garden, like, you know, tending to my flowers, which is what I love to do 557 00:34:17.450 --> 00:34:20.775 when I'm offline. Yes. Beautiful. And, 558 00:34:21.635 --> 00:34:25.315 speaking of flowers, y'all should, follow Avery on social media. She'll 559 00:34:25.315 --> 00:34:28.760 she'll post all that stuff there. You can find all the links to Avery's 560 00:34:28.760 --> 00:34:32.600 work, online dreya.com/309. And 561 00:34:32.600 --> 00:34:36.344 hey, guess what? Avery's gonna be speaking at our social media day summit coming 562 00:34:36.344 --> 00:34:40.025 up on June 30th. Avery, can you give us a little teaser 563 00:34:40.025 --> 00:34:43.625 about what your keynote's gonna be about? We're gonna be talking about this 564 00:34:43.625 --> 00:34:47.300 stuff. We're gonna get into these meaty issues. You 565 00:34:47.300 --> 00:34:50.819 know, I I am not an AI hater. I am also not an 566 00:34:50.819 --> 00:34:54.554 AI hype girl. I live in that in between world where I think 567 00:34:54.554 --> 00:34:58.235 most people are living right now. I'm trying to figure out how to how to 568 00:34:58.235 --> 00:35:01.860 wrestle with this stuff, how to use it in a really thoughtful, ethical 569 00:35:02.400 --> 00:35:05.840 way, to to get the benefits, but also to 570 00:35:05.840 --> 00:35:09.600 minimize the risk. So that's the kind of stuff I'm gonna be talking about. 571 00:35:09.600 --> 00:35:13.265 How to as a marketer, how do you use AI in a really thoughtful 572 00:35:13.265 --> 00:35:16.145 way? Yes. It's gonna be a lot of fun. We're gonna it's gonna be a 573 00:35:16.145 --> 00:35:19.650 lively discussion. We're gonna have a good time. Yay. I'm looking forward to it. Make 574 00:35:19.650 --> 00:35:23.410 sure y'all grab your tickets for that. And, Avery, if people wanna connect with you, 575 00:35:23.410 --> 00:35:26.530 tell us a little bit more about Camptech and how else they can connect with 576 00:35:26.530 --> 00:35:30.195 you. Yeah. Sure. So, Camptech, you know, we run, a 577 00:35:30.195 --> 00:35:33.575 number of different, training sessions for companies, 578 00:35:33.875 --> 00:35:37.670 for, small business groups, but we also run a whole bunch just for the general 579 00:35:37.670 --> 00:35:41.430 public. So you can check those out. You can attend them. They're online, so 580 00:35:41.430 --> 00:35:44.714 you can attend from anywhere. If you go to camp tech.ca/ 581 00:35:46.055 --> 00:35:49.815 workshops I think it I think it's workshops, plural. Yeah. I'm 582 00:35:49.815 --> 00:35:53.370 fairly certain it is. Camptech.ca/workshops. Also, like you 583 00:35:53.370 --> 00:35:57.050 said, you can find me. I I'm an oversharer on the Internet. I'm 584 00:35:57.050 --> 00:36:00.810 always online. I think there's somebody named Avery Swartz that lives 585 00:36:00.810 --> 00:36:04.565 in Tennessee, but, too bad for that person because I own that 586 00:36:04.565 --> 00:36:08.405 name online. If you if you look up Avery Swartz, it will be me. 587 00:36:08.405 --> 00:36:12.230 It is nobody else will will show up with that name. I've got it on 588 00:36:12.230 --> 00:36:15.990 lock. I love that. I think there's something too for having a unique ish 589 00:36:15.990 --> 00:36:18.170 name. My name, and I can't get anywhere. 590 00:36:19.815 --> 00:36:23.575 Thank you so much, Avery, for being on the show today. It's been my 591 00:36:23.575 --> 00:36:27.410 pleasure. And all of you listening, thank you for hanging out with us today. 592 00:36:27.410 --> 00:36:29.349 All the links are in the show notes onlinedrea.com/309. 593 00:36:31.410 --> 00:36:34.690 And make sure you give us a 5 star rating on Apple Podcasts and Spotify. 594 00:36:34.690 --> 00:36:38.485 Helps keep us in the top 100 marketing podcasts. That's all because of 595 00:36:38.485 --> 00:36:42.105 your support. I'll be back at you soon with another episode. 596 00:36:42.245 --> 00:36:45.250 Until then, I'll see you on the Internet as well. Bye.