WEBVTT 00:00:00.000 --> 00:00:01.940 Things like press releases have relevance again, 00:00:02.000 --> 00:00:04.120 as long as you frame the press releases in the 00:00:04.120 --> 00:00:06.820 form of answers to an extent. Make it more powerful. 00:00:06.940 --> 00:00:10.080 But I heard that PR doesn't help drive revenue, 00:00:10.259 --> 00:00:12.619 Perry. That's what I heard. No, no. You feel 00:00:12.619 --> 00:00:16.339 it in your gut. That was sarcasm for anybody 00:00:16.339 --> 00:00:20.199 who didn't pick up on it. Anyway, carry on. Folks, 00:00:20.260 --> 00:00:23.800 that's a joke. Welcome to Under Embargo, the 00:00:23.800 --> 00:00:26.339 podcast that says the quiet part out loud about 00:00:26.339 --> 00:00:28.940 PR, communications, and the chaos in between. 00:00:29.820 --> 00:00:32.240 Hosted by two of LinkedIn's most connected comms 00:00:32.240 --> 00:00:35.539 pros, Becca Chambers, a no -BS corporate comms 00:00:35.539 --> 00:00:39.020 exec with raging ADHD, and Pari Hedrick, founder 00:00:39.020 --> 00:00:42.159 of tech PR firm Crackle PR. This show brings 00:00:42.159 --> 00:00:45.320 it all. Real talk, sharp insights, and the kind 00:00:45.320 --> 00:00:47.579 of industry truths they don't put in press releases. 00:00:47.960 --> 00:00:51.880 No fluff, no filters, no f***s given. This is 00:00:51.880 --> 00:00:58.100 Under Embargo. Let's get into it. All right, 00:00:58.200 --> 00:01:02.179 welcome. to the under embargo podcast i am your 00:01:02.179 --> 00:01:05.640 host becca chambers cmo of scale venture partners 00:01:05.640 --> 00:01:08.819 and i'm joined as always by my co -host perry 00:01:08.819 --> 00:01:12.739 hedrick founder of crackle pr this is our first 00:01:12.739 --> 00:01:15.819 episode back after a couple months off for the 00:01:15.819 --> 00:01:18.480 summer of fun and summer of work and all the 00:01:18.480 --> 00:01:21.140 things but we're more energized than ever to 00:01:21.140 --> 00:01:24.840 be back and i'm excited for today's guest chris 00:01:24.840 --> 00:01:30.200 g who is um Talking about one of the things that 00:01:30.200 --> 00:01:33.340 I am most excited to talk about, which is AI 00:01:33.340 --> 00:01:37.159 strategy for PR and communications people, which 00:01:37.159 --> 00:01:41.439 is a very specific use of AI, which is rad. But 00:01:41.439 --> 00:01:43.680 it's probably the most relevant topic to most 00:01:43.680 --> 00:01:45.840 of the people who are listening to this podcast 00:01:45.840 --> 00:01:47.879 right now, because we are all trying to figure 00:01:47.879 --> 00:01:51.319 out how to use AI to do our jobs better. Before 00:01:51.319 --> 00:01:54.519 Chris started his AI consultancy, he was also 00:01:54.519 --> 00:01:58.099 head of digital for corporate affairs at Edelman 00:01:58.099 --> 00:02:00.780 and managing director of digital strategy for 00:02:00.780 --> 00:02:03.599 FGS Global, head of digital strategy at Burson, 00:02:03.659 --> 00:02:05.859 among other digital communication leadership 00:02:05.859 --> 00:02:10.060 roles over the past 20 plus years. So I am super 00:02:10.060 --> 00:02:12.439 excited to welcome Chris to the show. Welcome. 00:02:13.039 --> 00:02:15.900 Welcome. Thank you. Thank you. Thank you. It's 00:02:15.900 --> 00:02:18.819 a pleasure to be here. Boy, listing off all those 00:02:18.819 --> 00:02:21.719 PR agencies I've worked for makes me feel like 00:02:21.719 --> 00:02:24.759 I should get a punch card. But, you know, really 00:02:24.759 --> 00:02:27.300 excited to be here and, you know, really kind 00:02:27.300 --> 00:02:29.639 of an interesting transition for me. My career 00:02:29.639 --> 00:02:32.780 has been all about transformation. And so we're 00:02:32.780 --> 00:02:35.659 in a very significant period of transformation 00:02:35.659 --> 00:02:38.659 as we speak. And that's kind of the world where 00:02:38.659 --> 00:02:41.120 I sit in. I started Chris G Consulting two years 00:02:41.120 --> 00:02:46.219 ago. really kind of to address the AI adoption 00:02:46.219 --> 00:02:50.259 curve and gap as it relates to public relations 00:02:50.259 --> 00:02:53.159 and communications professionals, both PR firms 00:02:53.159 --> 00:02:56.719 and also internal corporate communications folks 00:02:56.719 --> 00:02:58.780 and everything. And you're right. I mean, this 00:02:58.780 --> 00:03:00.680 is really an interesting time. This is a big 00:03:00.680 --> 00:03:02.500 topic in our industry, and I'm excited to have 00:03:02.500 --> 00:03:05.080 it with you guys. It's awesome. And I actually 00:03:05.080 --> 00:03:07.620 think that's like the perfect segue to my first 00:03:07.620 --> 00:03:10.770 question for you, which is. How did you decide 00:03:10.770 --> 00:03:14.270 to make it specifically for communications, PR, 00:03:14.590 --> 00:03:19.509 corp comms folks? That's a great question. You 00:03:19.509 --> 00:03:23.530 know, I think for me, it really started, I had 00:03:23.530 --> 00:03:26.830 the idea before I left Edelman in, I guess, the 00:03:26.830 --> 00:03:30.430 early part of 2023, right? When we were all just 00:03:30.430 --> 00:03:34.550 coming out of the holidays and ChatGPT and MidJourney 00:03:34.550 --> 00:03:37.210 and other generative AI tools really captured 00:03:37.210 --> 00:03:41.740 our imagination. I just was trying to absorb 00:03:41.740 --> 00:03:43.719 as much as I could to learn as much as I could 00:03:43.719 --> 00:03:46.240 about AI at the time because I knew most of our 00:03:46.240 --> 00:03:48.039 clients were really going to want to know about 00:03:48.039 --> 00:03:50.500 it. I also wanted to see what the use cases were 00:03:50.500 --> 00:03:55.060 for us. I just really found that there was a 00:03:55.060 --> 00:03:58.500 lot out there, but there was very little that 00:03:58.500 --> 00:04:02.159 was relevant. I was looking for use cases that 00:04:02.159 --> 00:04:04.520 were specific to my job, that would help me do 00:04:04.520 --> 00:04:08.159 my job. I was looking for tools that would be 00:04:08.159 --> 00:04:12.120 specific to comms and PR. And what I found was 00:04:12.120 --> 00:04:14.599 a lot of stuff that was really hot up here and 00:04:14.599 --> 00:04:17.199 very little that was more specific to what we 00:04:17.199 --> 00:04:20.899 do. So when I decided to start my consultancy. 00:04:21.370 --> 00:04:23.089 that's where that's the first place i decided 00:04:23.089 --> 00:04:25.910 to focus i mean i know the space i knew what 00:04:25.910 --> 00:04:28.170 i would need if i were still in one of my previous 00:04:28.170 --> 00:04:31.410 roles i knew what my clients would need and what 00:04:31.410 --> 00:04:33.930 they were clamoring for as i was talking to them 00:04:33.930 --> 00:04:36.569 and doing workshops about ai and helping them 00:04:36.569 --> 00:04:39.629 learn about it so that became my focus and i 00:04:39.629 --> 00:04:42.129 decided to boil the ocean start with the community 00:04:42.129 --> 00:04:44.189 that i knew and it just really kind of grew from 00:04:44.189 --> 00:04:48.670 there Amazing. Can we get a little bit specific 00:04:48.670 --> 00:04:51.790 about some of the kinds of things from a comms 00:04:51.790 --> 00:04:54.569 perspective you were looking for AI to help you 00:04:54.569 --> 00:04:57.110 with and others in turn would be looking to apply 00:04:57.110 --> 00:05:00.649 to their lives? Asking for a PR agency. Myself. 00:05:02.990 --> 00:05:04.949 Yeah, you're trying to figure out how can we 00:05:04.949 --> 00:05:08.610 use AI to do timesheets, right? We don't do timesheets. 00:05:08.889 --> 00:05:13.810 How am I just learning about this now? No, you 00:05:13.810 --> 00:05:18.389 know, well, for one was, for instance, one of 00:05:18.389 --> 00:05:22.670 the things that really bogged down and just took 00:05:22.670 --> 00:05:25.889 hours out of my day and week when I was in my 00:05:25.889 --> 00:05:29.509 previous roles was creating slides. slides and 00:05:29.509 --> 00:05:31.269 spreadsheets. I'm sure neither of you guys have 00:05:31.269 --> 00:05:34.790 ever had to deal with those, right? And so, you 00:05:34.790 --> 00:05:36.730 know, and I would just, my day would just get 00:05:36.730 --> 00:05:39.230 derailed. I'd get an email. Hey, Chris, we're 00:05:39.230 --> 00:05:41.370 chasing this particular opportunity. We need 00:05:41.370 --> 00:05:43.529 slides about your practice. And I would have 00:05:43.529 --> 00:05:46.170 to sort of take 10 slides and crunch them down 00:05:46.170 --> 00:05:48.230 to three or take three slides and expand them 00:05:48.230 --> 00:05:50.959 out to 10 or 11. or take a deck that had been 00:05:50.959 --> 00:05:54.699 tailored for one vertical let's say uh retail 00:05:54.699 --> 00:05:59.540 and and morph it over to financial services and 00:05:59.540 --> 00:06:01.500 they would just take it just would you know just 00:06:01.500 --> 00:06:04.660 derail my day uh and it wasn't something it it's 00:06:04.660 --> 00:06:07.079 a necessary evil but at the same time wasn't 00:06:07.079 --> 00:06:09.160 one of the things that i was trying to strategically 00:06:09.160 --> 00:06:12.079 do to drive impact for the the firm or for my 00:06:12.079 --> 00:06:15.519 practice so um when i first learned how to use 00:06:15.519 --> 00:06:18.319 chat gpt to help me accelerate the production 00:06:18.319 --> 00:06:21.339 and creation of slides and slide content, it 00:06:21.339 --> 00:06:23.920 was a revelation, right? It literally shaved 00:06:23.920 --> 00:06:27.000 about at least an hour off of my day. And I figured 00:06:27.000 --> 00:06:30.420 out how to hack ChatGPT to give me a link to 00:06:30.420 --> 00:06:33.120 downloadable slides, which then really took my 00:06:33.120 --> 00:06:35.379 productivity up here. So it got to the point 00:06:35.379 --> 00:06:37.100 where, I mean, honestly, it felt like I was cheating. 00:06:37.240 --> 00:06:41.259 I would schedule send slides to my colleagues 00:06:41.259 --> 00:06:42.839 so they wouldn't know how quickly I could turn 00:06:42.839 --> 00:06:46.990 them around. In other words, you would be done, 00:06:47.129 --> 00:06:48.889 but you would delay it for a while. So they thought 00:06:48.889 --> 00:06:50.810 you were working your butt off. Yeah, that's 00:06:50.810 --> 00:06:52.850 right. I'd be done in about 30 minutes and I'd 00:06:52.850 --> 00:06:54.709 probably make it like, oh, it took me three hours 00:06:54.709 --> 00:06:58.410 to do this. You know, and I started to learn 00:06:58.410 --> 00:07:04.750 how to train the AI and AI chats and everything 00:07:04.750 --> 00:07:09.910 for more comm specific tasks. So say, for instance, 00:07:10.129 --> 00:07:14.129 for thought leadership. I created the logic within 00:07:14.129 --> 00:07:16.829 back then a chat and then eventually custom GPT 00:07:16.829 --> 00:07:20.310 because the dreaded thing that all of us have 00:07:20.310 --> 00:07:22.589 had to deal with is an executive who says, you 00:07:22.589 --> 00:07:24.930 haven't learned my voice after they have blown 00:07:24.930 --> 00:07:26.769 off every single meeting with you, not giving 00:07:26.769 --> 00:07:28.790 you any kind of feedback, haven't given you anything 00:07:28.790 --> 00:07:32.970 to go off of, right? And so I was able to sort 00:07:32.970 --> 00:07:35.389 of cobble whatever I could together. upload it 00:07:35.389 --> 00:07:38.889 into ChatGPT and just ask it to break down the 00:07:38.889 --> 00:07:41.209 attributes of that particular person's speaking 00:07:41.209 --> 00:07:43.389 or writing style. And then I would just basically 00:07:43.389 --> 00:07:45.910 train it, say, hey, use this for any time I give 00:07:45.910 --> 00:07:49.610 you an assignment. And then it gave me much better, 00:07:49.629 --> 00:07:52.509 more predictable outputs. And then, of course, 00:07:52.529 --> 00:07:54.910 I could tweak it because there's how an executive 00:07:54.910 --> 00:07:56.670 sounds and how they want to sound. So it's like, 00:07:56.709 --> 00:08:00.089 okay, we'll use this. And in one particular executive, 00:08:00.410 --> 00:08:03.949 I said to give it a little twist and insert some. 00:08:04.490 --> 00:08:07.810 uh bits and pieces from other people uh and then 00:08:07.810 --> 00:08:09.550 that we kind of nailed it and we kind of arrived 00:08:09.550 --> 00:08:12.110 at that so so really being able to kind of like 00:08:12.110 --> 00:08:15.740 take the ai effectively prompt it and train it 00:08:15.740 --> 00:08:18.139 for specific use cases like slide production 00:08:18.139 --> 00:08:22.639 for new business pitches, like training it for 00:08:22.639 --> 00:08:24.899 thought leadership content really kind of helped 00:08:24.899 --> 00:08:27.939 me plus up my productivity. So then once I started 00:08:27.939 --> 00:08:31.019 to do my own thing and train other clients on 00:08:31.019 --> 00:08:33.700 how to do that, you know, really what I started 00:08:33.700 --> 00:08:37.340 to do was incorporate other tactics and other 00:08:37.340 --> 00:08:40.940 things that were just sort of common needs. That's 00:08:40.940 --> 00:08:45.000 awesome. What like LLMs do you most rely on? 00:08:46.259 --> 00:08:49.460 Yeah. So this is one that really makes it fun 00:08:49.460 --> 00:08:52.620 when I talk to my accountants because I have 00:08:52.620 --> 00:08:56.440 so many subscriptions. I subscribe to Claude. 00:08:56.519 --> 00:09:00.700 I use ChatGPT. I use Gemini. I use Perplexity. 00:09:00.980 --> 00:09:04.960 And probably like most of us, I find that there's 00:09:04.960 --> 00:09:09.200 no one killer. app each of them has their strengths 00:09:09.200 --> 00:09:11.299 and minuses they have different things that are 00:09:11.299 --> 00:09:13.620 unique about them that make them really good 00:09:13.620 --> 00:09:20.159 i really like the deep research agent for from 00:09:20.159 --> 00:09:25.059 chat gpt and open ai i really like the creative 00:09:25.059 --> 00:09:28.700 output from claude Although I think ChatGPT is 00:09:28.700 --> 00:09:32.019 catching up with it. So there are different things 00:09:32.019 --> 00:09:34.360 that I like and kind of keep me every month saying, 00:09:34.460 --> 00:09:37.840 I still need to subscribe to this. So I don't 00:09:37.840 --> 00:09:39.580 think there's any one in particular that's head 00:09:39.580 --> 00:09:41.179 and shoulders above the other, depending on what 00:09:41.179 --> 00:09:42.700 it is you're looking for. And that's typically 00:09:42.700 --> 00:09:44.320 what I tell clients when they're trying to figure 00:09:44.320 --> 00:09:48.320 out what should be in their stack. So my agency 00:09:48.320 --> 00:09:50.740 is spending a lot of time, Chris, trying to figure 00:09:50.740 --> 00:09:54.720 out. how to best position our clients to be delivered 00:09:54.720 --> 00:09:58.399 in the answers on these LLMs, right? Because 00:09:58.399 --> 00:10:00.940 PR is having a bit of a moment now where everybody 00:10:00.940 --> 00:10:03.740 was writing us off saying, oh, who needs PR? 00:10:03.879 --> 00:10:06.120 AI can do all that. Well, it turns out earned 00:10:06.120 --> 00:10:08.620 media and third -party validation is more important 00:10:08.620 --> 00:10:10.799 than it ever has been. And things like press 00:10:10.799 --> 00:10:12.559 releases have relevance again, as long as you 00:10:12.559 --> 00:10:15.100 frame the press releases in the form of answers 00:10:15.100 --> 00:10:18.940 to an extent, make it more probable. PR doesn't 00:10:18.940 --> 00:10:21.340 help drive revenue, Perry. That's what I heard. 00:10:21.620 --> 00:10:26.019 No, you feel it in your gut. That was sarcasm 00:10:26.019 --> 00:10:28.460 for anybody who didn't pick up on it. Anyway, 00:10:28.539 --> 00:10:33.480 carry on. Folks, that's a joke. So Chris, I guess 00:10:33.480 --> 00:10:36.440 my question is, to what extent are you kind of 00:10:36.440 --> 00:10:39.139 deep in the world of helping either your clients 00:10:39.139 --> 00:10:42.080 or yourself figure out best practices relative 00:10:42.080 --> 00:10:45.159 to being found in those? LLMs at the point of 00:10:45.159 --> 00:10:47.899 need? And what can you kind of say about the 00:10:47.899 --> 00:10:51.200 state of that? Yeah, well, I mean, it's a big 00:10:51.200 --> 00:10:55.000 area because as you guys know, what comes up 00:10:55.000 --> 00:10:57.679 in these searches is really important. And there 00:10:57.679 --> 00:11:03.139 was a stat I saw from Gartner that said that, 00:11:03.179 --> 00:11:05.120 well, first of all, let me just paint the picture. 00:11:06.840 --> 00:11:09.360 In my previous roles, I would go in and I would 00:11:09.360 --> 00:11:12.539 talk to clients and tell them that the first 00:11:12.539 --> 00:11:14.559 page of google results that is your corporate 00:11:14.559 --> 00:11:16.879 home page you can have all the fancy branding 00:11:16.879 --> 00:11:20.700 and copywriting all you want but people are going 00:11:20.700 --> 00:11:22.299 to search your company they're going to search 00:11:22.299 --> 00:11:24.700 your executives names etc and then whatever pops 00:11:24.700 --> 00:11:27.480 up in that first 10 results on the first page 00:11:27.480 --> 00:11:29.860 of google is going to heavily influence what 00:11:29.860 --> 00:11:33.019 they believe about your brand and the last stats 00:11:33.019 --> 00:11:35.159 i saw something around in the neighborhood of 00:11:35.159 --> 00:11:38.580 93 or 94 percent of audiences believe what they 00:11:38.580 --> 00:11:42.399 find in google search results right okay so that 00:11:42.399 --> 00:11:44.100 was sort of then that was the lay of the land 00:11:44.100 --> 00:11:45.740 that we've been operating under probably for 00:11:45.740 --> 00:11:49.679 the last 20 years or so now as we all know a 00:11:49.679 --> 00:11:52.559 lot of us are starting to put those same types 00:11:52.559 --> 00:11:57.679 of queries into ai into directly into llms and 00:11:57.679 --> 00:12:00.399 get the results out they don't give you 10 results 00:12:00.399 --> 00:12:04.019 they give you one answer right and even with 00:12:04.019 --> 00:12:07.159 google a lot of times our searches end with the 00:12:07.799 --> 00:12:10.679 the AI answer, the AI mode answer at the very 00:12:10.679 --> 00:12:12.879 top, and then you'll see the 10 results at the 00:12:12.879 --> 00:12:16.799 bottom. And I've seen anywhere from around up 00:12:16.799 --> 00:12:19.039 to 40 % of the time, people will look at the 00:12:19.039 --> 00:12:21.679 answer that's provided in AI mode and not even 00:12:21.679 --> 00:12:23.360 look at the rest of the results at the bottom. 00:12:23.679 --> 00:12:27.279 So it's really becoming an important driver in 00:12:27.279 --> 00:12:29.399 terms of driving reputation. And then on top 00:12:29.399 --> 00:12:31.179 of that, I started to mention the Gartner research. 00:12:31.539 --> 00:12:36.929 Gartner predicts that by 2026, search volume 00:12:36.929 --> 00:12:40.590 will have declined by about 25%. Okay. So that's 00:12:40.590 --> 00:12:43.649 massive because now we're going to see a much, 00:12:43.710 --> 00:12:46.809 much larger number of people relying on that 00:12:46.809 --> 00:12:48.710 single answer. So that really raises the bar 00:12:48.710 --> 00:12:50.870 para for what you mentioned. It's got to be right. 00:12:50.950 --> 00:12:53.950 We've got to make sure that we get how our brand 00:12:53.950 --> 00:12:57.649 needs to be represented well into these LLMs. 00:12:57.750 --> 00:13:01.809 And that's a challenge because they all kind 00:13:01.809 --> 00:13:04.220 of. uh work a little bit differently and they 00:13:04.220 --> 00:13:06.799 all are trained on different information and 00:13:06.799 --> 00:13:09.919 new sources right uh for instance uh you may 00:13:09.919 --> 00:13:12.320 get a great hit about your company in the new 00:13:12.320 --> 00:13:14.539 york times well guess what they're they're currently 00:13:14.539 --> 00:13:18.440 in litigation with llms like chat gpt and so 00:13:18.440 --> 00:13:21.139 they're not allowing those tools to scrape their 00:13:21.139 --> 00:13:25.059 content it's important Right. It's, it's so, 00:13:25.200 --> 00:13:29.279 so what shows up in these LLMs in some part will 00:13:29.279 --> 00:13:31.320 be influenced by who's allowing that information 00:13:31.320 --> 00:13:33.980 to be ingested. And you know, what's an interesting 00:13:33.980 --> 00:13:37.019 thing that I think is I used to never recommend 00:13:37.019 --> 00:13:40.500 that people like just apply for lists or awards 00:13:40.500 --> 00:13:43.220 or whatever, because just to do it, like, that's 00:13:43.220 --> 00:13:45.159 just a checkbox. There's no outcome. Like, what 00:13:45.159 --> 00:13:47.799 does that get you? Now I'm like, do it. If you 00:13:47.799 --> 00:13:49.700 can get on a list, because if somebody is asking 00:13:49.700 --> 00:13:52.000 what's the best, blah, blah, blah. It is going 00:13:52.000 --> 00:13:54.759 to pull from some random list. Even if that list 00:13:54.759 --> 00:13:57.200 is bullshit, it's going to show up. So like, 00:13:57.200 --> 00:14:01.059 it may not be important to you for X, Y, or Z 00:14:01.059 --> 00:14:03.139 outcome or selling more, whatever. But if it's 00:14:03.139 --> 00:14:05.820 the thing that's going to show up in the LLM, 00:14:05.860 --> 00:14:08.879 that is a massive value. So like, maybe we can't 00:14:08.879 --> 00:14:11.559 like just thumb our nose at these lower quality 00:14:11.559 --> 00:14:14.379 listicles anymore that I don't know. I used to 00:14:14.379 --> 00:14:16.259 often just tell people, let's not spend our time 00:14:16.259 --> 00:14:17.919 on that kind of crap. And let's try to actually 00:14:17.919 --> 00:14:20.299 focus on getting our message out. But now sometimes 00:14:20.299 --> 00:14:23.200 it's like, Maybe you do just want to be mentioned 00:14:23.200 --> 00:14:25.820 alongside other things. I feel like anywhere 00:14:25.820 --> 00:14:28.860 you can pollinate your brand's message, you pollinate. 00:14:29.000 --> 00:14:31.000 Like that's, that's the new organic sources. 00:14:31.240 --> 00:14:35.100 Yes. Correct. Yes. I agree. And it's such a different 00:14:35.100 --> 00:14:39.659 thing than PR used to. Well, then strategic PR 00:14:39.659 --> 00:14:43.120 people used to like, we, it used to be very like, 00:14:43.159 --> 00:14:45.860 let's take a targeted approach. We're doing X 00:14:45.860 --> 00:14:49.080 to achieve Y outcome. Now it's kind of like. 00:14:49.470 --> 00:14:52.029 spray and pray is back baby let's get it all 00:14:52.029 --> 00:14:57.429 over the place yeah we have we have um prospects 00:14:57.429 --> 00:15:00.190 calling us specifically now i mean like directly 00:15:00.190 --> 00:15:02.950 for this for the expressed reason of getting 00:15:02.950 --> 00:15:05.649 cited in llms when people are searching like 00:15:05.649 --> 00:15:07.629 i would say probably 20 to 30 percent of the 00:15:07.629 --> 00:15:10.389 inbound we get is they lead with that like we 00:15:10.389 --> 00:15:13.009 understand search is waning we understand the 00:15:13.009 --> 00:15:16.549 ascent of llms how do we best position ourselves 00:15:16.549 --> 00:15:19.139 now to get ahead of the competition Full stop. 00:15:19.220 --> 00:15:20.860 And so that's how we're beginning our kind of 00:15:20.860 --> 00:15:23.179 thing. Like a spray and pray offering. That's 00:15:23.179 --> 00:15:24.899 like, it's not even quality media we're going 00:15:24.899 --> 00:15:27.360 for. We're just going for hits. Would you like 00:15:27.360 --> 00:15:29.879 a fire hose special? Because that's for sale 00:15:29.879 --> 00:15:36.320 today. Gatling gun, right? Listen, it's the wild, 00:15:36.360 --> 00:15:39.559 wild west. So it's not a bad analogy. You know, 00:15:39.559 --> 00:15:41.500 and I think you kind of touched on it. The old 00:15:41.500 --> 00:15:45.590 is new again. uh blogs uh blogs are now you know 00:15:45.590 --> 00:15:48.830 involved because newsletters newsletters right 00:15:48.830 --> 00:15:52.549 yeah um even reddit i mean uh open ai did a big 00:15:52.549 --> 00:15:56.389 uh deal with reddit and it's it's that content 00:15:56.389 --> 00:15:58.389 those conversations are helping train the model 00:15:58.389 --> 00:16:01.129 and so you know we might start seeing amas back 00:16:01.129 --> 00:16:04.350 in vogue right they are they are back in vogue 00:16:04.350 --> 00:16:08.090 i've been invited to do reddit amas lately and 00:16:08.090 --> 00:16:13.179 i'm like what What? Written AMAs? Cool. I did 00:16:13.179 --> 00:16:15.379 an AMA on Reddit. I was asked to do it for the 00:16:15.379 --> 00:16:17.779 public relations, you know, channel basically. 00:16:17.960 --> 00:16:20.220 And I did that about a year and a half ago. But 00:16:20.220 --> 00:16:22.580 the Reddit thing is so interesting to me because 00:16:22.580 --> 00:16:25.000 anybody who spent any time there knows that their 00:16:25.000 --> 00:16:27.559 bullshit detectors are like all over the place. 00:16:27.620 --> 00:16:29.559 Like you can't go there as a marketer and like 00:16:29.559 --> 00:16:31.500 talk about how awesome you are and expect not 00:16:31.500 --> 00:16:34.799 to get skewered. Only adding value. Yeah. Oh, 00:16:34.799 --> 00:16:38.000 yeah, exactly. So so but everybody now knows 00:16:38.000 --> 00:16:39.899 that Reddit is instrumental in delivering those 00:16:39.899 --> 00:16:41.860 answers. And so everybody's trying to infiltrate 00:16:41.860 --> 00:16:44.679 Reddit. And boy, oh, boy, is a war shaping up 00:16:44.679 --> 00:16:47.720 between people preserving the integrity of Reddit 00:16:47.720 --> 00:16:49.919 and the marketers who want to infiltrate it. 00:16:49.960 --> 00:16:52.820 And it's just a really interesting time. It is 00:16:52.820 --> 00:16:55.879 interesting. Super interesting. It really is. 00:16:55.960 --> 00:16:58.179 You know, and I would also say, you know, so 00:16:58.179 --> 00:17:00.600 it's interesting, I think, to pay attention to 00:17:00.600 --> 00:17:03.460 those media deals. You know, for instance, another 00:17:03.460 --> 00:17:06.099 one with OpenAI has been very aggressive. They've 00:17:06.099 --> 00:17:08.920 done a media deal with Axios. And I think that 00:17:08.920 --> 00:17:10.759 the centerpiece of that deal is they're going 00:17:10.759 --> 00:17:14.119 to be doing, starting at least with local content 00:17:14.119 --> 00:17:17.779 and for markets, for trial markets to start. 00:17:19.079 --> 00:17:21.859 I think that that's really important is to research 00:17:21.859 --> 00:17:26.619 and figure out where those LLMs are getting their 00:17:26.619 --> 00:17:30.059 news, where they have those deals, and bake that 00:17:30.059 --> 00:17:34.759 into your media strategy. Can I ask a question 00:17:34.759 --> 00:17:38.720 about how you see young talent in the comm space 00:17:38.720 --> 00:17:41.980 and PR space right now? Because I know that there's 00:17:41.980 --> 00:17:43.940 been some studies that show there's something 00:17:43.940 --> 00:17:46.880 like, I don't know, the first two or three years. 00:17:47.309 --> 00:17:49.170 Typically, when students are out of college, 00:17:49.349 --> 00:17:51.950 those jobs that they would have are kind of being 00:17:51.950 --> 00:17:54.390 erased by AI across various industries. I'm hearing 00:17:54.390 --> 00:17:56.430 that all the time. But when I look at what's 00:17:56.430 --> 00:17:59.170 actually happening at my agency and what you've 00:17:59.170 --> 00:18:01.009 described as the things you've been using AI 00:18:01.009 --> 00:18:04.089 for, those don't really sound like super junior 00:18:04.089 --> 00:18:07.109 tasks. In fact, the slide presentation stuff, 00:18:07.329 --> 00:18:09.150 the... you know all the things you mentioned 00:18:09.150 --> 00:18:12.230 seem more like at least middle management to 00:18:12.230 --> 00:18:15.289 senior management type tasks so are you fearful 00:18:15.289 --> 00:18:18.309 or up or optimistic about the young folks considering 00:18:18.309 --> 00:18:20.670 getting into comms or getting into pr against 00:18:20.670 --> 00:18:22.910 the backdrop of what you've seen and done in 00:18:22.910 --> 00:18:27.339 the ai space Well, I think it's a mistake. I 00:18:27.339 --> 00:18:30.299 know that there are agencies who have said that 00:18:30.299 --> 00:18:33.420 part of how they're planning their roadmap with 00:18:33.420 --> 00:18:37.119 AI is to reduce their reliance on junior staff. 00:18:37.299 --> 00:18:39.339 And I think that's a mistake. And I'll just be 00:18:39.339 --> 00:18:41.440 completely honest. I never thought that we in 00:18:41.440 --> 00:18:45.119 the comms industry knew how to deploy. junior 00:18:45.119 --> 00:18:49.539 staff in the first place. It never made sense 00:18:49.539 --> 00:18:52.960 to me, guys, that we'd have these armies of young 00:18:52.960 --> 00:18:57.160 people that you recruit out of the best colleges 00:18:57.160 --> 00:18:59.420 and universities in the country, and you say, 00:18:59.500 --> 00:19:03.210 great. uh get up at 5 00 a .m and start with 00:19:03.210 --> 00:19:05.049 media clips and all this kind of stuff i just 00:19:05.049 --> 00:19:06.750 thought that there's a better way to analyze 00:19:06.750 --> 00:19:09.269 media than that you know i didn't come up in 00:19:09.269 --> 00:19:12.009 pr i came up in marketing and advertising and 00:19:12.009 --> 00:19:14.690 i learned how to analyze media coverage without 00:19:14.690 --> 00:19:16.609 having to go through that process so i know it's 00:19:16.609 --> 00:19:21.059 possible um and then you know You guys have been 00:19:21.059 --> 00:19:23.160 around this business for a long time. We've also 00:19:23.160 --> 00:19:26.000 gotten in the bad habit of making them pseudo 00:19:26.000 --> 00:19:28.880 assistants and having them take notes during 00:19:28.880 --> 00:19:31.039 calls and things like that. You don't even get 00:19:31.039 --> 00:19:32.700 to speak until you've gotten a couple of stripes 00:19:32.700 --> 00:19:35.059 on your shoulder. So I don't think we ever really 00:19:35.059 --> 00:19:38.500 knew how to deploy our young talent in the first 00:19:38.500 --> 00:19:41.099 place. It's a whole separate conversation about 00:19:41.099 --> 00:19:43.240 us not knowing how to deploy senior talent either. 00:19:43.380 --> 00:19:47.019 But I think give them... Give them meatier things 00:19:47.019 --> 00:19:50.140 to do. Yes, take some of those rote tasks off 00:19:50.140 --> 00:19:53.980 their plate. Use AI to accelerate the process 00:19:53.980 --> 00:19:57.059 of media coverage. But how about have them start 00:19:57.059 --> 00:20:00.319 doing analysis, right? It's never too early to 00:20:00.319 --> 00:20:02.700 train their brains to be able to be good media 00:20:02.700 --> 00:20:05.220 analysts, right? What are you seeing? What are 00:20:05.220 --> 00:20:06.940 the threads? What are the through lines? What 00:20:06.940 --> 00:20:08.779 do you think is the implication? That's what 00:20:08.779 --> 00:20:11.140 clients want. They could get the clips themselves, 00:20:11.539 --> 00:20:13.839 right? They want to know what to do with them. 00:20:14.190 --> 00:20:17.309 Now AI makes it so you don't have to do any clipping. 00:20:17.630 --> 00:20:20.269 You just have the data. What are you going to 00:20:20.269 --> 00:20:22.690 do with the data? And I think now you could come 00:20:22.690 --> 00:20:25.730 straight out of college and start using LLMs 00:20:25.730 --> 00:20:28.130 and be the one who is analyzing data because 00:20:28.130 --> 00:20:31.490 they are they. We are trained how to do that 00:20:31.490 --> 00:20:34.670 in college. We are trained how to look at information 00:20:34.670 --> 00:20:37.730 and come up with conclusions based on that information. 00:20:38.009 --> 00:20:40.809 So to your point, it's not like you just have 00:20:40.809 --> 00:20:43.559 to do. bitch work in order to learn how to do 00:20:43.559 --> 00:20:46.200 pr it's like apply what you already know to the 00:20:46.200 --> 00:20:48.599 tools that we have available to us now and then 00:20:48.599 --> 00:20:51.539 learn as you go for all of the other stuff for 00:20:51.539 --> 00:20:56.359 sure i actually think like unless we start looking 00:20:56.359 --> 00:21:02.359 like creating new kind of opportunities for this 00:21:02.359 --> 00:21:04.740 just out of school generation it's going to be 00:21:04.740 --> 00:21:08.230 really weird for a few years because We're just 00:21:08.230 --> 00:21:10.890 expecting people to go into the same roles that 00:21:10.890 --> 00:21:13.990 existed before, even though those roles are going 00:21:13.990 --> 00:21:17.730 away. But there is still a need for people to 00:21:17.730 --> 00:21:21.769 utilize said new tools and manage those. And 00:21:21.769 --> 00:21:24.650 that shouldn't necessarily fall on the CMO or 00:21:24.650 --> 00:21:26.549 the director of comms or whatever, because their 00:21:26.549 --> 00:21:29.430 time might be better spent elsewhere. Why are 00:21:29.430 --> 00:21:31.289 we not saying, hey, this is a perfect role for 00:21:31.289 --> 00:21:33.150 the 21 -year -old straight out of college? I 00:21:33.150 --> 00:21:36.089 don't know. But this is a great point. Absolutely. 00:21:36.250 --> 00:21:39.369 You know, and I think also that we need more 00:21:39.369 --> 00:21:41.690 talent. I mean, the industry is going through 00:21:41.690 --> 00:21:45.109 and is just at the very beginnings of a massive 00:21:45.109 --> 00:21:47.150 transformation. None of us knows what it's going 00:21:47.150 --> 00:21:48.990 to look like on the other side. We're going to 00:21:48.990 --> 00:21:51.289 need all the brainpower we can throw at this. 00:21:51.390 --> 00:21:54.369 So it's almost like the Lakers saying, hey, we're 00:21:54.369 --> 00:21:57.049 not going to. take the best prospects out of 00:21:57.049 --> 00:21:59.170 college anymore we're just going to use veterans 00:21:59.170 --> 00:22:02.509 right um it just doesn't make any sense these 00:22:02.509 --> 00:22:04.329 are your young these are your future leaders 00:22:04.329 --> 00:22:09.089 and your future stars so like i think that there's 00:22:09.089 --> 00:22:13.670 something like there is something to understanding 00:22:13.670 --> 00:22:17.730 how to brand people companies whatever you have 00:22:17.730 --> 00:22:20.930 to understand like generational nuance and like 00:22:21.589 --> 00:22:24.509 40 year olds and 20 year olds are from two different 00:22:24.509 --> 00:22:26.630 planets in a lot of ways right now, you know? 00:22:26.630 --> 00:22:30.089 And so you need to have 40 year olds, 20 year 00:22:30.089 --> 00:22:32.950 olds, 58 year olds, all working under the same 00:22:32.950 --> 00:22:36.349 roof in order to be able to like hit the cultural 00:22:36.349 --> 00:22:38.950 nuance and understand the things that matter 00:22:38.950 --> 00:22:41.470 to people these days, because everyone's attention 00:22:41.470 --> 00:22:44.029 is so fragmented. It's just not as easy as it 00:22:44.029 --> 00:22:46.690 was before to be like, here's a swath of people 00:22:46.690 --> 00:22:49.059 that I'm trying to target because. that swath 00:22:49.059 --> 00:22:51.299 that you're trying to target now exists in a 00:22:51.299 --> 00:22:53.700 thousand different places than they used to before 00:22:53.700 --> 00:22:55.299 well you know what's interesting becca about 00:22:55.299 --> 00:22:58.039 that so um the digital native piece i totally 00:22:58.039 --> 00:23:00.720 agree with and and they would be best equipped 00:23:00.720 --> 00:23:02.579 you would think to embrace this new technology 00:23:02.579 --> 00:23:05.940 and and quickly you know grok how to install 00:23:05.940 --> 00:23:08.500 it and everything they do grok no pun intended 00:23:08.500 --> 00:23:11.839 um but but there's kind of an existential battle 00:23:11.839 --> 00:23:16.369 isn't there between colleges who are sort of 00:23:16.369 --> 00:23:20.329 anti -AI in a sense, wanting kids to prove out 00:23:20.329 --> 00:23:23.190 and use their own thinking and not relying too 00:23:23.190 --> 00:23:26.130 heavily upon AI to get their education and get 00:23:26.130 --> 00:23:28.269 their answers. And yet the real world is demanding 00:23:28.269 --> 00:23:31.869 they do just that. Like the real world is screaming 00:23:31.869 --> 00:23:35.450 and crying for talent coming out of college who 00:23:35.450 --> 00:23:37.410 already has a baseline understanding of all these 00:23:37.410 --> 00:23:39.829 tools. So I don't know how you rectify that without. 00:23:40.460 --> 00:23:42.380 Having a broader, you know, sort of educational 00:23:42.380 --> 00:23:45.000 campaign saying these colleges need to embrace 00:23:45.000 --> 00:23:48.220 AI, lean the hell into it and educate the living 00:23:48.220 --> 00:23:50.319 shit out of kids that are about to enter the 00:23:50.319 --> 00:23:52.079 workforce. Because otherwise there's going to 00:23:52.079 --> 00:23:53.880 be a two or three year delta between when they 00:23:53.880 --> 00:23:55.900 get out of school and when they get hired. Wasn't 00:23:55.900 --> 00:23:59.259 that the same though for like dot com time? Like 00:23:59.259 --> 00:24:01.740 everybody's saying like these kids are leaving 00:24:01.740 --> 00:24:04.279 college and they have no idea how to like. The 00:24:04.279 --> 00:24:06.579 reason why everyone was dropping out of college 00:24:06.579 --> 00:24:08.299 because college wasn't teaching them how to do 00:24:08.299 --> 00:24:11.039 the stuff in order to create the next big company 00:24:11.039 --> 00:24:15.440 using code or go work at a startup being a coder 00:24:15.440 --> 00:24:17.539 because colleges weren't doing it. Colleges are 00:24:17.539 --> 00:24:20.460 10 years behind and I'm all about higher education. 00:24:20.680 --> 00:24:24.440 Don't get me wrong. But it is a miss when AI 00:24:24.440 --> 00:24:27.859 is clearly the revolutionary technology right 00:24:27.859 --> 00:24:29.539 now and colleges are just spending all their 00:24:29.539 --> 00:24:32.690 time trying to figure out how to... moderate 00:24:32.690 --> 00:24:35.029 it which yes they do have to do that but they 00:24:35.029 --> 00:24:36.930 also have to figure out how to teach it from 00:24:36.930 --> 00:24:39.990 an ethics perspective from a like a use just 00:24:39.990 --> 00:24:43.809 practical use perspective it's like being 1997 00:24:43.809 --> 00:24:47.269 and like saying you guys cannot use google for 00:24:47.269 --> 00:24:49.390 your assignments no internet for you library 00:24:49.390 --> 00:24:52.640 and only do that there and it's like yeah no 00:24:52.640 --> 00:24:55.839 you need you colleges need to get go find some 00:24:55.839 --> 00:24:58.400 ai experts to come and teach some courses so 00:24:58.400 --> 00:25:01.039 that you guys can get up to speed with the times 00:25:01.039 --> 00:25:03.599 you kids can use encyclopedias and microfiche 00:25:03.599 --> 00:25:07.799 and that's it that's it oh my goodness boy they 00:25:07.799 --> 00:25:09.720 got a whole bunch of gen z -ers like what is 00:25:09.720 --> 00:25:13.460 googling microfiche um you're no you're 100 right 00:25:13.460 --> 00:25:16.509 i don't know how you justify charging what could 00:25:16.509 --> 00:25:19.609 be you know 250 000 for a four -year education 00:25:19.609 --> 00:25:22.210 and have that kid graduate knowing nothing about 00:25:22.210 --> 00:25:25.809 ai in 2025 it's almost criminal um but i think 00:25:25.809 --> 00:25:27.609 that we've seen just as you guys have mentioned 00:25:27.609 --> 00:25:30.230 we've seen these patterns repeat themselves where 00:25:31.479 --> 00:25:33.680 universities are way behind and it takes some 00:25:33.680 --> 00:25:35.960 time to catch up. It happened with myself. I 00:25:35.960 --> 00:25:37.759 mean, like I said, you know, I was a graphic 00:25:37.759 --> 00:25:40.599 design major. I went to art school. I'm a little 00:25:40.599 --> 00:25:42.440 bit older than I look. And so, you know, when 00:25:42.440 --> 00:25:44.700 I graduated, it was right at the beginning of 00:25:44.700 --> 00:25:47.299 the desktop publishing revolution. Problem, I 00:25:47.299 --> 00:25:49.779 learned how to do all the graphic design stuff 00:25:49.779 --> 00:25:52.700 in college with the old school T -square and 00:25:52.700 --> 00:25:54.880 triangle and flexible curve and all that stuff 00:25:54.880 --> 00:25:58.359 they don't use today. And when I graduated, effectively, 00:25:58.440 --> 00:26:01.299 my degree was obsolete. My first job interview, 00:26:01.519 --> 00:26:02.960 they asked, do you know how to use a computer? 00:26:03.180 --> 00:26:06.500 And I lied. And I said, yeah, I do. And I figured, 00:26:06.640 --> 00:26:09.200 well, I know more than they do. And I'll just 00:26:09.200 --> 00:26:11.279 figure it out. And I figured it out. And I taught 00:26:11.279 --> 00:26:14.059 myself. So I think that we're almost kind of 00:26:14.059 --> 00:26:16.079 in a similar inflection point. But back then, 00:26:16.099 --> 00:26:19.720 they wouldn't let us use computers in our program 00:26:19.720 --> 00:26:21.880 in college, just like they're not letting people 00:26:21.880 --> 00:26:25.019 use AI today. Now, eventually, they came around. 00:26:25.140 --> 00:26:27.140 And eventually, they'll come around in universities 00:26:27.140 --> 00:26:30.099 today. But you're right. I think that we have 00:26:30.099 --> 00:26:32.579 to think about new ways to challenge students. 00:26:32.799 --> 00:26:36.099 If they can use ChatGPT or Claude to create a 00:26:36.099 --> 00:26:39.859 paper, great. How about have them do a mock debate, 00:26:40.099 --> 00:26:42.940 right? How about take them and say, all right, 00:26:42.960 --> 00:26:45.059 well, the paper that you created, go over there 00:26:45.059 --> 00:26:46.819 and take your colleague who knows nothing about 00:26:46.819 --> 00:26:48.779 this topic and teach them about it, right? Show 00:26:48.779 --> 00:26:50.380 how much they understand. Or how about this? 00:26:50.660 --> 00:26:54.420 Show me your prompts. right let me see how creative 00:26:54.420 --> 00:26:56.839 and how intricate your prompts were to generate 00:26:56.839 --> 00:26:59.480 this paper and i'll grade you on that i think 00:26:59.480 --> 00:27:01.279 that there's so much we could be doing rather 00:27:01.279 --> 00:27:02.859 than wringing our hands that they're using it 00:27:02.859 --> 00:27:04.900 to cheat they've always been cheating right if 00:27:04.900 --> 00:27:07.480 you want to cheat you've always done it the teachers 00:27:07.480 --> 00:27:10.240 need to evolve too it is really up it's like 00:27:10.240 --> 00:27:13.119 it's both things because there's new tools like 00:27:13.470 --> 00:27:16.670 okay people used to use a hammer and now we have 00:27:16.670 --> 00:27:20.089 a plug -in that you can use to break up the concrete 00:27:20.089 --> 00:27:21.970 you know they shouldn't have to just keep using 00:27:21.970 --> 00:27:23.529 the hammer because the teacher only knows how 00:27:23.529 --> 00:27:25.690 to use the hammer and i just like it boggles 00:27:25.690 --> 00:27:27.950 the mind i think what you said before and this 00:27:27.950 --> 00:27:30.029 is always my number one piece of advice i give 00:27:30.029 --> 00:27:32.390 to anybody who asks like what advice would you 00:27:32.390 --> 00:27:34.890 give me it's to be the i'll figure it out person 00:27:34.890 --> 00:27:37.670 because especially during these times of transition 00:27:38.430 --> 00:27:40.369 Nobody knows what the hell they're doing. We 00:27:40.369 --> 00:27:42.650 are all just trying to figure it out. And if 00:27:42.650 --> 00:27:45.069 you have the drive to figure it out and raise 00:27:45.069 --> 00:27:46.960 your hand and be like I don't really know how 00:27:46.960 --> 00:27:48.420 to do that, but I'm sure I can figure it out. 00:27:48.660 --> 00:27:50.779 Go figure it out. And that's how you learn how 00:27:50.779 --> 00:27:53.519 to do it better, right? You learned AI by playing 00:27:53.519 --> 00:27:56.079 with AI. You didn't go to a class and say, hey, 00:27:56.099 --> 00:27:58.099 teach me how to do this so that I can go teach 00:27:58.099 --> 00:28:00.920 other people. You literally started just consuming 00:28:00.920 --> 00:28:03.740 information and tinkering with it and talking 00:28:03.740 --> 00:28:06.180 to people and blah, blah, blah. And that is really 00:28:06.180 --> 00:28:09.079 where we're at right now with AI. And the more 00:28:09.079 --> 00:28:12.960 you can just trial and error and figure things 00:28:12.960 --> 00:28:15.980 out. you will be light years ahead of your peers. 00:28:16.119 --> 00:28:18.779 Even if you don't have anybody come and train 00:28:18.779 --> 00:28:21.160 you how to do it, just start watching YouTube 00:28:21.160 --> 00:28:24.279 videos or reaching out to people who do know 00:28:24.279 --> 00:28:26.660 how to do things and like tinker, tinker, tinker, 00:28:26.759 --> 00:28:28.200 because that is really the thing that's going 00:28:28.200 --> 00:28:31.339 to, I think, win the day right now and make you 00:28:31.339 --> 00:28:33.079 way more marketable in your next job. And you 00:28:33.079 --> 00:28:36.539 can even ask AI to teach you how to use it. Exactly. 00:28:37.579 --> 00:28:41.259 Exactly. And just kind of pivot a little bit. 00:28:42.990 --> 00:28:45.650 All the organizations I see, the CEOs are talking 00:28:45.650 --> 00:28:47.569 about like, we need to reduce headcount because 00:28:47.569 --> 00:28:51.390 like, it's almost like a badge of honor to have 00:28:51.390 --> 00:28:54.490 reduced headcount right now. Like they are super 00:28:54.490 --> 00:28:56.490 psyched to report to their shareholders, like 00:28:56.490 --> 00:28:58.609 we've reduced headcount, our profits are up. 00:28:58.869 --> 00:29:01.089 But yet against that backdrop, I'm hearing about 00:29:01.089 --> 00:29:04.410 a lot of leaders hiring back people of the roles 00:29:04.410 --> 00:29:07.049 they eliminated to your earliest point, Chris, 00:29:07.170 --> 00:29:10.059 about you think it's a mistake letting off. letting 00:29:10.059 --> 00:29:13.079 go of the youngest folks. Where do you see it 00:29:13.079 --> 00:29:15.599 shaken out in corporate America writ large as 00:29:15.599 --> 00:29:18.640 it relates to CEOs being compensated for reducing 00:29:18.640 --> 00:29:21.839 headcount and being glorified leaders because 00:29:21.839 --> 00:29:24.759 they've become so efficient and the reality on 00:29:24.759 --> 00:29:26.779 the ground that you need human beings to actually 00:29:26.779 --> 00:29:28.740 interact with these tools to bring out the best 00:29:28.740 --> 00:29:30.920 of what it's got? How do you see this playing 00:29:30.920 --> 00:29:33.839 out in the next five years? Well, I think that 00:29:34.799 --> 00:29:37.480 There have been somewhere I've read that somewhere 00:29:37.480 --> 00:29:40.640 in the neighborhood of like two trillion dollars 00:29:40.640 --> 00:29:44.099 worth of money invested in AI over the past couple 00:29:44.099 --> 00:29:46.599 of years or so. So I think a lot of investors, 00:29:46.740 --> 00:29:48.759 first of all, I think there's been a lot of overpromising 00:29:48.759 --> 00:29:53.940 and sort of outlandish expectations about what 00:29:53.940 --> 00:29:55.700 those efficiency gains would look like and how 00:29:55.700 --> 00:29:58.420 quickly they would happen. And in a lot of cases, 00:29:58.440 --> 00:30:00.940 the investors are saying, look. You spent all 00:30:00.940 --> 00:30:02.839 this money. We need to start seeing some cost 00:30:02.839 --> 00:30:04.400 savings and we need to start seeing some returns. 00:30:04.880 --> 00:30:06.440 So I think that there are going to be a number 00:30:06.440 --> 00:30:09.259 of executives, CEOs who are going to feel pressured 00:30:09.259 --> 00:30:11.400 to cut costs. So I think we're going to keep 00:30:11.400 --> 00:30:13.579 seeing that because in some cases they're going 00:30:13.579 --> 00:30:15.900 to have to do it. If for no other reason, just 00:30:15.900 --> 00:30:18.140 to have a Klarna type moment, like, look, we 00:30:18.140 --> 00:30:21.039 tried it. That didn't work out very well. Now 00:30:21.039 --> 00:30:23.940 we got to start bringing people back. Similar 00:30:23.940 --> 00:30:27.859 type of thing with Duolingo. I think we're probably 00:30:27.859 --> 00:30:29.480 going to see some of that because I think some 00:30:29.480 --> 00:30:32.779 of them just feel kind of boxed in. But I think 00:30:32.779 --> 00:30:34.980 a lot of this is due to a lot of these predictions. 00:30:35.279 --> 00:30:37.119 And again, we've seen this, we saw it with the 00:30:37.119 --> 00:30:42.039 dot -com boom as well. Retail or brick and mortar 00:30:42.039 --> 00:30:43.259 is going to go away. You're never going to go 00:30:43.259 --> 00:30:45.480 to a store again. Well, how'd that work? So I 00:30:45.480 --> 00:30:47.140 think that people saying, hey, we're going to 00:30:47.140 --> 00:30:49.700 have all these capabilities within the next six 00:30:49.700 --> 00:30:52.259 months or 12 months or so, set up unrealistic 00:30:52.259 --> 00:30:54.500 expectations and we're kind of paying for that. 00:30:54.920 --> 00:30:57.519 I think realistically, I always say co -pilot, 00:30:57.539 --> 00:31:01.579 not autopilot. The ideal pairing is humans and 00:31:01.579 --> 00:31:05.099 AI working together. And ultimately, whether 00:31:05.099 --> 00:31:06.839 they have to learn it the hard way or the easy 00:31:06.839 --> 00:31:09.119 way, I hope they learn it the easy way, I think 00:31:09.119 --> 00:31:11.859 companies are going to figure that out. What's 00:31:11.859 --> 00:31:16.400 so interesting is I hear so very little about 00:31:16.400 --> 00:31:20.099 companies investing in... upskilling their workforce 00:31:20.099 --> 00:31:22.980 for this so very little and i mean this is what 00:31:22.980 --> 00:31:25.160 you do for a living so people are hiring you 00:31:25.160 --> 00:31:27.640 to come in and upskill their workforce but like 00:31:27.640 --> 00:31:32.380 the conversation is let's cut people Or let's 00:31:32.380 --> 00:31:34.720 hire people who know how to do AI stuff so that 00:31:34.720 --> 00:31:37.779 we can do AI stuff here. And there's no conversation 00:31:37.779 --> 00:31:41.460 about let's give time to our employees so that 00:31:41.460 --> 00:31:44.460 they can go tinker with AI. And let's give them 00:31:44.460 --> 00:31:48.460 X number of hours a day to go learn this or have 00:31:48.460 --> 00:31:50.619 somebody come in and teach them or help them 00:31:50.619 --> 00:31:53.599 build an agent or whatever. It comes down to 00:31:53.599 --> 00:31:56.059 people really needing to go out and do it themselves. 00:31:56.519 --> 00:31:59.440 And so if you do want to have a workforce who 00:31:59.440 --> 00:32:01.779 is going to multiply, apply what they're doing 00:32:01.779 --> 00:32:04.980 right now by using AI, then freaking teach them 00:32:04.980 --> 00:32:07.920 how to do it, right? Give them permission to 00:32:07.920 --> 00:32:11.980 go try and fail and waste time on a project that 00:32:11.980 --> 00:32:14.200 might lead to nowhere because that's how they're 00:32:14.200 --> 00:32:16.140 going to learn how to do it. And you're never 00:32:16.140 --> 00:32:19.200 going to be able to have an AI workforce that 00:32:19.200 --> 00:32:22.019 is productive. using AI if you don't teach people 00:32:22.019 --> 00:32:25.059 how to use it? And that I don't understand. Like, 00:32:25.059 --> 00:32:27.759 where is the conversation around upskilling rather 00:32:27.759 --> 00:32:30.660 than just like - No, that reminds me of the old 00:32:30.660 --> 00:32:33.920 CFO saw, which is, well, what happens if we spend 00:32:33.920 --> 00:32:35.680 all this money training all these people and 00:32:35.680 --> 00:32:37.619 they leave? And then they leave. Yeah. And then 00:32:37.619 --> 00:32:38.859 somebody's like, well, what if we don't train 00:32:38.859 --> 00:32:41.039 these people and they fucking stay? That's right. 00:32:41.579 --> 00:32:44.440 That's kind of where we're at. You're right. 00:32:44.700 --> 00:32:47.779 Well, it's incredibly short -sighted. It's the 00:32:47.779 --> 00:32:50.039 first place, unfortunately, in our business culture, 00:32:50.200 --> 00:32:52.119 which has been this way for the past 40 years. 00:32:52.259 --> 00:32:54.480 It's the first place we go, right? Cut costs 00:32:54.480 --> 00:32:58.319 and costs are us, right? People. But you're 100 00:32:58.319 --> 00:33:00.819 % right, Becca. The companies that are going 00:33:00.819 --> 00:33:03.539 to endure, like, you know, if we could transport 00:33:03.539 --> 00:33:07.140 10 years into the future. The strongest, best 00:33:07.140 --> 00:33:08.799 performing companies are going to be companies 00:33:08.799 --> 00:33:10.539 that have more people who understand how to use 00:33:10.539 --> 00:33:14.200 AI, not fewer people using AI. So I think that 00:33:14.200 --> 00:33:16.839 upskilling really needs that. We've got to think 00:33:16.839 --> 00:33:19.259 about that again. And it wasn't that long ago 00:33:19.259 --> 00:33:22.539 that that training, corporate training, et cetera, 00:33:22.640 --> 00:33:24.759 was the norm in corporate America. But I think, 00:33:24.799 --> 00:33:26.799 you know, I don't know, my theory is somewhere 00:33:26.799 --> 00:33:30.190 around. the Great Recession, you know, 2008, 00:33:30.710 --> 00:33:33.410 2009, a lot of those budgets got cut and they 00:33:33.410 --> 00:33:35.490 just never came back. But I think we're going 00:33:35.490 --> 00:33:37.950 to have to really dust that off because, you 00:33:37.950 --> 00:33:40.759 know, you can... you're going to have to have 00:33:40.759 --> 00:33:43.680 a lot of training. You can come in and you can 00:33:43.680 --> 00:33:46.579 hire someone like me, please do. But I think 00:33:46.579 --> 00:33:48.720 you can't just take one bite at the apple. It's 00:33:48.720 --> 00:33:51.380 a continuous thing. No one can take one golf 00:33:51.380 --> 00:33:53.079 lesson and think that they know how to play golf 00:33:53.079 --> 00:33:55.140 or one French lesson and think they know how 00:33:55.140 --> 00:33:57.640 to speak French fluently. You've got to keep 00:33:57.640 --> 00:33:59.700 doing it. You've got to keep training. And I 00:33:59.700 --> 00:34:01.119 think companies are going to have to continue 00:34:01.119 --> 00:34:03.859 to invest in training, even for people who know 00:34:03.859 --> 00:34:06.119 how to use AI, because it's a moving target. 00:34:06.980 --> 00:34:09.079 Totally. I mean, I felt like a year ago I was 00:34:09.079 --> 00:34:11.500 on top of it. And now I just always feel like 00:34:11.500 --> 00:34:13.400 I'm behind because like, it really is a matter 00:34:13.400 --> 00:34:15.679 of like, how many hours in the week do you have 00:34:15.679 --> 00:34:18.079 to spend on learning the new things or understanding 00:34:18.079 --> 00:34:20.219 the new models or whatever, whatever, because 00:34:20.219 --> 00:34:23.139 there is constantly something new, constantly 00:34:23.139 --> 00:34:25.880 a new way to add value, but like it takes time, 00:34:25.980 --> 00:34:29.340 right? You have to go devote time to learning 00:34:29.340 --> 00:34:32.739 it, tinkering with it, refining it. Like it's 00:34:32.739 --> 00:34:36.389 not just a snappy, easy. you know i think the 00:34:36.389 --> 00:34:40.030 permission piece um organizationally like internally 00:34:40.030 --> 00:34:42.090 like if we if our ethos is like we are going 00:34:42.090 --> 00:34:43.789 to get smart we're going to do what it takes 00:34:43.789 --> 00:34:46.570 to to learn uh as best we can as these things 00:34:46.570 --> 00:34:48.710 morph and change i think that's the winning strategy 00:34:48.710 --> 00:34:51.260 like i'm trying to incorporate that with So we 00:34:51.260 --> 00:34:53.780 have a dedicated channel on Slack that is all 00:34:53.780 --> 00:34:56.960 about AI learnings, use cases, where everybody 00:34:56.960 --> 00:34:58.920 has just constantly got this little laboratory 00:34:58.920 --> 00:35:01.380 of things they're trying. And they tried to upload 00:35:01.380 --> 00:35:03.340 this file into that. And this is the format, 00:35:03.480 --> 00:35:05.500 just constantly sharing what we're learning, 00:35:05.619 --> 00:35:08.920 iterating and tweaking. But it's got to be blessed 00:35:08.920 --> 00:35:10.900 at the highest levels of the organization to 00:35:10.900 --> 00:35:13.039 let people experiment, to Becca's point, and 00:35:13.039 --> 00:35:16.260 have that time and space and grace and latitude 00:35:16.260 --> 00:35:20.800 to learn as they fly. companies don't have the 00:35:20.800 --> 00:35:23.039 patience or the time unless there's a measurable 00:35:23.039 --> 00:35:25.960 outcome or output. Sound familiar, PR and comms? 00:35:26.619 --> 00:35:29.900 Associated with that. So I think that's another 00:35:29.900 --> 00:35:32.760 problem we have to solve. How do you give soft 00:35:32.760 --> 00:35:35.679 time to people that you can't directly measure 00:35:35.679 --> 00:35:37.860 the outcome? How do you do that organizationally 00:35:37.860 --> 00:35:42.480 and have the key stakeholders bless that? We're 00:35:42.480 --> 00:35:45.559 in that transition right now. And on the note 00:35:45.559 --> 00:35:48.699 of transition. I'm going to transition us to 00:35:48.699 --> 00:35:51.820 our ending questions because they're my favorite 00:35:51.820 --> 00:35:54.500 part of the show because they show everybody's 00:35:54.500 --> 00:35:58.000 personality. Although we will not be asking our 00:35:58.000 --> 00:36:00.679 old question of, would you rather have the hiccups 00:36:00.679 --> 00:36:03.420 for three months or cut your finger off? Because 00:36:03.420 --> 00:36:06.019 we've asked that 10 times now. And it's good 00:36:06.019 --> 00:36:08.760 every time. Wait, but wait, what is your answer 00:36:08.760 --> 00:36:10.460 to that one quickly? Would you rather have hiccups 00:36:10.460 --> 00:36:13.409 for three months or cut off your pinky? Oh, man. 00:36:14.269 --> 00:36:16.570 I mean, I'd have to go with the hiccups. I rely 00:36:16.570 --> 00:36:20.110 way too much on my pinkies. I'm attached to them. 00:36:20.349 --> 00:36:23.550 There it is. Okay, so first we'll go with the 00:36:23.550 --> 00:36:25.150 normal ones, and then I'm going to let Perry 00:36:25.150 --> 00:36:27.230 go with an unhinged question for you. And I don't 00:36:27.230 --> 00:36:28.889 even know what Perry's unhinged questions are, 00:36:28.969 --> 00:36:30.769 but I also have some unhinged questions if you 00:36:30.769 --> 00:36:36.210 want to do this. Okay. You get to erase one corporate 00:36:36.210 --> 00:36:40.110 buzzword from existence. What is it? Low -hanging 00:36:40.110 --> 00:36:46.800 fruit. Love it. Agreed. You're about to give 00:36:46.800 --> 00:36:49.480 a presentation in front of 50 ,000 screaming 00:36:49.480 --> 00:36:53.960 fans. What is your walkout song or walk -up song? 00:36:54.920 --> 00:37:01.139 My walk -up song? I would say Public Announcement 00:37:01.139 --> 00:37:04.880 by Jay -Z. Allow me to reintroduce myself. I 00:37:04.880 --> 00:37:09.940 like that. That's a good one for sure. Okay, 00:37:10.059 --> 00:37:14.519 if you had to do a TED Talk on anything but AI, 00:37:14.699 --> 00:37:18.260 communications, marketing, what would your topic 00:37:18.260 --> 00:37:23.400 be? I would say the need for better leadership 00:37:23.400 --> 00:37:27.579 in corporate America. Ooh, I would watch that 00:37:27.579 --> 00:37:32.320 and agree, probably wholeheartedly. And you get 00:37:32.320 --> 00:37:35.900 to use one emoji only for the rest of your life. 00:37:36.239 --> 00:37:40.530 Which emoji? Oh man, that's a tough one. I like 00:37:40.530 --> 00:37:44.210 my emojis. It would have to be the sort of flexing 00:37:44.210 --> 00:37:48.289 arm emoji. Yeah, I use that a lot. Yeah, that 00:37:48.289 --> 00:37:50.570 means that things are good. Yeah, most of my 00:37:50.570 --> 00:37:54.969 friends should grin. Love it. Solid. Perry, do 00:37:54.969 --> 00:37:56.849 you have an unhinged question for Chris? It's 00:37:56.849 --> 00:37:59.050 not quite as unhinged as cutting off one's pinky 00:37:59.050 --> 00:38:02.030 or being eaten by a shark, but... You have a 00:38:02.030 --> 00:38:05.389 choice, Chris. You can be given $20 million free 00:38:05.389 --> 00:38:08.409 and easy today to use as you see fit for the 00:38:08.409 --> 00:38:11.429 remainder of your days. The problem with that 00:38:11.429 --> 00:38:14.469 is you can never sleep in the same dwelling two 00:38:14.469 --> 00:38:16.750 nights in a row for the rest of your life. And 00:38:16.750 --> 00:38:18.590 it can't be two houses you buy next door. You 00:38:18.590 --> 00:38:21.130 just go next door. It's got to be vastly different 00:38:21.130 --> 00:38:23.949 locations. Do you take the money? And say, screw 00:38:23.949 --> 00:38:25.570 it, I'll just fly all over the world and sleep 00:38:25.570 --> 00:38:28.750 wherever the hell I want. Or is routine, structure, 00:38:29.269 --> 00:38:33.849 normalcy more important than that $20 million 00:38:33.849 --> 00:38:36.829 cash infusion that would undeniably change your 00:38:36.829 --> 00:38:41.469 life for the remainder of your days? Man, everybody's 00:38:41.469 --> 00:38:42.769 going to say. That's a hard one. That's a hard 00:38:42.769 --> 00:38:44.530 one because I love my bed. Everybody's going 00:38:44.530 --> 00:38:46.329 to say, this guy's insane. He's going to say 00:38:46.329 --> 00:38:50.010 no to $20 million, but no. I am a creature of 00:38:50.010 --> 00:38:53.349 habit and structure, and I would much, much rather 00:38:53.349 --> 00:38:56.409 have my routine and my friends and my bed and 00:38:56.409 --> 00:39:00.389 all that stuff. I'm with you, Chris. This is 00:39:00.389 --> 00:39:02.190 the first time we've ever all agreed on something. 00:39:02.309 --> 00:39:05.429 I 100 % would also not take the money. What am 00:39:05.429 --> 00:39:09.280 I going to do? I have a wife and kids. What am 00:39:09.280 --> 00:39:12.800 I going to do? See you in a couple days. I just 00:39:12.800 --> 00:39:17.380 value my sleep too much. I would hate the idea 00:39:17.380 --> 00:39:20.579 of waking up one day and being so tired and just 00:39:20.579 --> 00:39:22.420 wanting to get in my bed later that night and 00:39:22.420 --> 00:39:23.989 being like, where the fuck am I? Am I sleeping 00:39:23.989 --> 00:39:26.250 tonight? I don't even know. I got to get there. 00:39:26.349 --> 00:39:29.230 This sucks. I bet if we ask this question a few 00:39:29.230 --> 00:39:31.090 more times to some guests, some people will be 00:39:31.090 --> 00:39:32.650 like, you know what? I'm taking the goddamn money. 00:39:33.750 --> 00:39:40.070 Careful what you ask for. Chris, I want to give 00:39:40.070 --> 00:39:42.869 you an opportunity to share anything with our 00:39:42.869 --> 00:39:44.949 audience that you would like to share, how they 00:39:44.949 --> 00:39:47.550 can find you, any parting words, whatever you 00:39:47.550 --> 00:39:50.769 want to leave us with. Yeah, parting words. I 00:39:50.769 --> 00:39:52.590 just want to echo what you said. I mean, you 00:39:52.590 --> 00:39:55.110 know, I tell people all the time when they say 00:39:55.110 --> 00:39:56.550 they're not an expert, they don't know anything 00:39:56.550 --> 00:39:58.190 about this stuff. I say, well, play around with 00:39:58.190 --> 00:40:01.949 it a bit and wait six months, right? So this 00:40:01.949 --> 00:40:04.190 is evolving. We're on the ground floor of this 00:40:04.190 --> 00:40:06.989 amazing technology. Please jump in. We need all 00:40:06.989 --> 00:40:09.469 the brains we can get. So that's the first thing. 00:40:09.530 --> 00:40:12.679 Where they can find me, you know. Not as much 00:40:12.679 --> 00:40:15.420 as you guys, but I try to maintain a presence 00:40:15.420 --> 00:40:19.539 on LinkedIn. So you can find me there. And also 00:40:19.539 --> 00:40:24.880 my website, chrisg .me, where I blog semi -regularly 00:40:24.880 --> 00:40:28.019 and have a newsletter where I do a prompt of 00:40:28.019 --> 00:40:30.599 the week every Saturday. Ooh, that's a good one. 00:40:30.639 --> 00:40:34.210 I didn't know that. Well, this has been... Very 00:40:34.210 --> 00:40:37.369 enlightening, very fun. And I love this topic. 00:40:37.409 --> 00:40:39.610 So Chris, I will continue to be in touch with 00:40:39.610 --> 00:40:41.530 you and hope to learn more from you in the future. 00:40:41.570 --> 00:40:45.789 And everyone go visit Chris G's website and connect 00:40:45.789 --> 00:40:48.369 with him on LinkedIn. And we'll see you all next 00:40:48.369 --> 00:40:50.730 time. Thank you. Appreciate the discussion. Thank 00:40:50.730 --> 00:40:52.210 you, buddy. Thanks for having me.