Transcript of the podcast:

John  (00:02): Hello, my name is John Cass, and I’m here with AIContentGen, I’m here today with Chris Penn. And we’re going to be talking about AI and marketing. Chris, thank you so much for joining today on the podcast. I think first met you at one of your pod camps many years ago, and, you know, I think you’re just such an inspiration. You were definitely an inspiration to me. I think the way that you approached marketing, you helped to teach the entire community here in the Boston area and across the country, actually. You know, what social media is, the value of it is. And then also you also had that connection to the PR industry. So thank you for joining me today.

Chris (00:49): Thank you for having me.

John  (00:54): You were one of the sort of the early pioneers in the industry with PodCamp and those, and those sorts of things. You know, what, what’s your sort of philosophy you know, in looking at the latest, latest in technology? Cuz now you’re in ai?

Chris (01:14): You know, I don’t know that I necessarily have a philosophy so much as it’s just what I’m interested in. I like to, to see what machines and tools and stuff can do. I am notorious for finding uses unintended uses of any technology. Just yesterday, I was cleaning out the toaster in my kitchen with the leafblower because I didn’t feel like seeing, I was sitting there shaking this thing and opening those little trays. Just take the leaf floor stick, unplug it, take the leaf floor, stick it in there, just blow all the, the, the crap out. And the same thing applies to, you know, all these things that we’re doing with data science and ai as, as an industry, just taking a look at what’s possible, what the tools can do, and then say, well, what are the, what are the applications that we wish we had solutions for? What kinds of intelligent automation should we be doing? What kinds of answers can we get that previously might have been inaccessible or might have been, you know, overly laborious to get? So that’s, that’s sort of just how things have pivoted over the years, you know, podcasting. Yeah, we were all early on, you know, I started my first podcast in 2005 and, you know, marketing over coffee has been on the air continuously since 2007, but the industry’s evolved and, and ideally everybody evolves with it.

John  (02:33):No, that’s very true. That’s very true. I mean, I also think it’s interesting cuz I mean, do you, do you call yourself a, a digital guy or a, a marketer or, because I think what’s interesting is that you spent quite a bit of time in the PR industry working for a PR agency. But what was fascinating to me is that you worked in tech pr, right? And that was a, you, you worked with a group of people who were very savvy and understood that it wasn’t just about the traditional ways that PR has done, even though it was done in digital context. It is being done in a digital context. So you have to bring in analytics and those sorts of things. How does that sort of color your perception of coming, having worked in the PR industry in relation to you know, digital marketing and ai?

Chris (03:27): I mean, AI really is, it’s just a tool. It’s, it’s, a lot of it is just math. And so the, the question to ask is what problems does any industry, public relations, advertising, marketing, what problems do you have that are mathematical in nature that you need answers for? One of PRS biggest problems is how do we measure this stuff, right? How do we know that public relations is working? And the limitations in the, in that industry have been largely because statistical modeling techniques for what PR does have been around for about 40 to 50 years. These are, these are not new things. However, no one in the PR industry has had the, the, the mathematical background to bring those to life. And so having this arsenal, or this toolkit of other things could bring into the industry helped for a period of time in improving measurement.

Chris (04:28): There are, you know, certain metrics and measures that I think should be in every PR person’s toolkit. And very, very few folks do it. Partly because, again, it requires some expertise and it requires stepping outside your comfort zone and part of it, because sometimes one of the trends that I’ve seen, particularly in the last five years, maybe the last six years, is people, and this stems from general society and politics. People only like data when it says what they want, right? People do not like answers that are uncomfortable. People do not like being told they’re wrong. People don’t like hearing bad news and or hearing something that does not agree with their perspective. And so when you do something like attribution analysis on a PR campaign, and the news is it did nothing, right? You spent 20 grand for nothing, right? People don’t like hearing that <laugh>

John  (05:31): <Laugh>. Well, well, Chris, that, that, that’s one reason why I think as a marketer and in the profession I’ve worked, you know, I don’t know if you know, but I, I worked in the agile marketing community for, you know, quite a few number of years. And part of it is, I think it reframes, it gives another mindset not just for the marketers. And it’s this mindset of, oh, we’re, we’re testing things, right? And so it also, but it also, but the real be, I mean, it definitely helps with the teams and also the leaders in marketing to think like that, right? But then it also gives a methodology that if you explain to the other stakeholders, you know, the C-suite and the other departments, they can say, oh, that’s what this market is doing. You know, they’re, they’re coming to me with this different mindset and they’re, so I think, I think something like a framework or a methodology like agile then re recast the frame. I mean, do you think that is kind of helpful? I know we weren’t going to talk about agile, but you, you, you brought that issue up.

Chris (06:33): It can, if the person understands the point of agile or understands the point of any framework, and it’s aligned with their goals, and this is the problem that, you know my c e o, Katie, Roberta and I have been tr trying to tackle around analytics and data in general for the existence of our company which is data and analytics and stuff. These are like cooking methods, right? You know, broiling, frying, et cetera. If the c e o wants McDonald’s, it doesn’t matter how good, you know, a, a, a stake chef you are or how good a sushi chef you are, if the, your stakeholders have a predefined, preset, inflexible rigid point of view, right? Unless you are aligned with that point of view, you’re not gonna make any headway. So someone who is, we call it opinion driven, someone who is opinion driven, isn’t going to become data driven, right?

Chris (07:33): That’s, it’s like asking somebody who is you know, one religious faith to become another religious faith. It’s very difficult. Something, something has to happen to, to motivate that change. And motivating change in human beings is really hard. It’s much easier to change code, right? Than it is to, to change humans. And so that combined with the systems and structures in businesses is very challenging because a lot of systems and structures reward the wrong things. You know, real simple example, if you are a publicly traded company, your reward is your stock price, right? And so you measure everything to your quarterly earnings. You measure everything to, you know, how much can you goose the stock price this quarter and keep investors happy? Not can I build a, a sustainable business for the long term and make bets that maybe money losing bets and, and not have it be a punishment, you know, for all of the issues and objections and ethical problems that I have with, with Facebook now, now meta you know, mark Zuckerberg trying to do the whole, you know, metaverse virtual reality, that’s a big bet.

Chris (08:43): And so far it’s really not paying off. But I have to at least acknowledge that they’re taking a big bet, even though they’re a publicly traded company. Most other publicly traded companies do not want to rock the boat. They’re unwilling to make those big bets. And as a result, when you start bringing things like analytics and stuff in and saying like, yeah, here’s what’s going on. They’re like, Nope, this, this, this data doesn’t help us make our quarterly numbers. Like, well, <laugh> then, then, you know, if, if that’s your point of view, that’s gonna be a lot harder to make change. It’s one of the driving forces why you see a lot of companies, there’s, you know, reasonably successful companies going back private, right? Privately held investment because they and their investors think, we know we can make some big moves, but we can’t do it if our, if our horizon of vision is measured by the quarter instead of by the decade.

John  (09:36): Right? Right, right, right. So ai, you know, how did you get into this area of, of marketing, you know, and what, and what are you doing in the area at the moment at at the company?

Chris (09:50): So I got into AI through analytics, right? So I, I started really paying attention to analytics and, and embracing it really, like 2004, 2005, you know, as websites started producing a lot of data and we all needed ways to analyze it. Google analytics became available to the public in 2005 after the Google bought a company called Urchin. 2011 is when they first introduced multi-tech attribution and assisted conversions. And it was at that point when like, okay, I guess I need to, you know, shrug off the shame and embarrassment of failing statistics in college and, and relearn this stuff. So when I joined the old public relations parameters with part of that you from 2013 on was reteaching myself, statistics, data science and stuff, and, and, and analytics. And, and AI was a natural outgrowth of that because that’s where machine learning, specifically classical machine learning is where data science and statistics have been going.

Chris (10:45): Well, they, they’ve been there for 50 years, but the computational power has been available to us over the last 20 years. You know, your laptop can do things now that 50 years ago where theoretical only. And so my career progression really started taking that turn into the machine learning space, or starting in 2013, predictive analytics, you know, forecasting regression analysis, complicated regression analysis and things. And then in 20, late 2017 Katie my c e o and I, we, we decided the, the agency we were working with at the time was going one direction. The, we wanted to go in a very different direction. So we, we split off and started our own company and focused around the, the more intelligent use of, of data and analytics. And part and parcel of that is using artificial intelligence. I think it’s, you know, worth pointing out that AI is kind of a, a umbrella term.

Chris (11:42): There’s, there’s three things in machine learning that you’re, you’re three fundamental tasks, right? There’s regression classification in generation regression is supervised learning. Can we figure out, you know, what happened? Classification is, Hey, I’ve got a big bucket of data. Can I make so sense of it? Can I sort it? And, and, and classify it and understand what’s in this giant bucket, which is a problem that many marketers have. And then generation is, Hey, I want machines to make something from the data that we have. This has been the, the talk of the town really for the last 18 months with services like open ai, Dolly and Dolly two stable diffusion mid journey and, and stuff like that. You know, making pictures of dogs and tutus on skateboards and such writing blog posts, you know, with, you know, one click blog post writing. But the underlying technologies are all pretty much the same thing, you know, regression classification and generation.

John  (12:38): While you were at that PR company, were you also looking at social listening? Because, and, and the reason I asked that question, and I had wondered was because I was director of blogging strategies at a company called Backbone Media back in, I think it it’s about 15 years ago. And, you know, we were heavily into that. We weren’t a PR agency, we’re a digital marketing agency, but I, I think PR was some of the first people to look at social listening. And that, to me seemed to be a really ready a application and an early application of machine learning.

Chris (13:13): Yeah. Listening and search listening were the, were the two things that were worth paying attention to. And search listening I think is actually more valuable which is seeing what people are typing into search engines. I mean, anyone can do that. You go to Google Trends, you know, trends.google.com, type in search queries. You can see search volume of, of a given set of terms over time. You know, you don’t need any technical skill for that. But there was a book, I want to say it’s like five or six years ago called Everybody Lies. And it was a, a book they’re talking about from the perspective of search engines, the billions and trillions of questions that people ask surgeons, they would never ask another person, another living human being because it’s too embarrassing, too private, or something that, you know, even in a professional context, if you are in a marketing role and you, and you know, your, your, your stakeholder says, Hey, let’s let, let’s get some attribution analysis going.

Chris (14:06): And you don’t wanna admit, I have no idea what that means, <laugh>, right? You’re probably not gonna say it in public social media unless you have like a throwaway account, but you will absolutely Google it and say, okay, what is attribution analysis? Social listening is really good for qualitative analysis to understand the, the language space of our problem. It’s really poor at quantitative analysis because there’s, generally speaking people only have conversations about things when something’s either very wrong or very right. Most people don’t have conversations about things that were okay, right? You know, if you, if you think about like a restaurant experience, you either leave a one star review or a five star review. You, Jim, people don’t leave three star reviews. It’s like, service was good, food was fine, it was good, right? <Laugh>, that’s, you’re not, not compelled to leave review. It’s like, you know, but the, the waiter threw my food on me and lift the table on fire and <laugh>, you know and you see this like on Amazon too. There are tons of either one or five star reviews on things. It’s, there really is not a, not a lot of middle ground. And so social listening provides you the qualitative context, and then you need to use other methods like surveying and things to quantify the questions that you’ve developed from social listening.

John  (15:22): Right? You know, it, it is interesting cuz I’ve actually used reviews to pull together insights for customer journey mapping where I wasn’t able to initially do surveying. So I’d like to see more tools in, in those areas. Now for trust insights, what, what are you doing in that? How are you, how are you working with clients? What’s, what’s the area of coverage?

Chris (15:49): So, with a lot of our clients, our, we’re a management consulting firm that focuses right now on marketing. Although, you know, the, the techniques can be used to anything. Many of our clients come to us to make better use of the data they have or fix the, you know, the infrastructure problems they have or get insights that they’re not able to extract otherwise. For example, we have one client where we, we process a lot of their NPSs scores. We look at the, their, their net promoter score data, and then do some, you know, fancy math to say, here are the things they’re probably driving, you know, this score that’s rising or this score that’s falling. We had another client in the food and beverage industry that said, here’s, here’s the inbox of our, of our customer service department, right? Here’s all the emails coming in, do some text analysis and tell us if we’ve got any blind spots.

Chris (16:43): And we did. We found, you know, this was 20 20 19, they had no formulation. They made thickeners, they had no formulation for oat milk, right? And, and it was, it was becoming a very hot topic. And so being able to dig into their, their existing data, people asking them, emailing them, Hey, do you have, what’s your solution for oat milk? Was a valuable insight for them. We did some work with a, a recruiting agency, and we, we sh they were like, we can’t get people to, to, you know, fill out the application forms as much as we want. So we, again, we did some analysis. Here’s all the things that you say in your 5,000 job listings, here are the transcripts of 17,000 calls of candidates with your recruiters. Notice that the conversations, the questions candidates are asking are not any answered anywhere in any of your job listings.

Chris (17:29): Like, you know, well, you know, what’s for the vacation time? What’s the, the starting pay and stuff like that. And we said, if you just put those common questions into your job descriptions, you’ll do better. They did. And they literally increased their conversion rates 40% within, like, overnight, just, just saying, okay, here’s the answers that people are, are really after a lot of our other clients. Right now Google Analytics four is, is the hot topic. You know, the universal analytics system is coming to an end on July one of the, of next year. And a lot of people are realizing that it’s not just a a, an upgrade. It’s not like going from Microsoft Word 2020 to, to Word 2022. It’s a totally new piece of software. And so there’s a lot of marketing operations changes we’re helping companies make to, to, to make that pivot to the new system and make it useful because there are a legion of gotchas that, that are not obvious. So that’s, that’s kind of who, who we try to help.

John  (18:25): Great. Great. Great. how do you see the future of AI marketing? Where do you think it’s gonna go? You know in relation to where we are today and in, in the next well, two years or five years.

Chris (18:40): <Laugh>, if I knew that we would not be having this conversation because I would already be retired, <laugh>, phenomenally wealthy. Here’s what we, here’s what we see and, and what you can see in the marketplace, machine learning of all kinds is just getting baked into products, right? Very few companies other than like really big companies with really big budgets are building their own. Most companies are waiting for vendors to build machine learning into their products. And it’s in, it is in everything now, right? We’re using Zoom. Zoom has, you know, live captions, which is using text speech to text transcription, which is AI based. That is, you know an example of AI just kind of snuck its way in there. We’re gonna see much more of that much more intelligent automation, just machinery being used to, to speed things up.

Chris (19:33): The thing that’s captured people’s imagination right now is generation. You know, all of the, the AI generated artwork pe we are seeing a mechanistic automation of certain parts of marketing, you know, content generation is, is, is a big part of it. And there’s a real serious dangers with that. Particularly on the intellectual property side. I just did an almost hour long interview with attorney Ruth Carter. They specialize in IP law, and they were saying like, yeah, there’s, there’s a whole bunch of, of legal issues with AI generated content that nobody seems to be aware of, but it’s gonna bite some people really, really, really hard. So those questions are gonna need yet resolved to the ex to some extent in the next couple years. There are some there, what’s happening in the industry overall right now is there is a trend towards privacy.

Chris (20:26): And this is not gonna let up anytime soon. We have legislation popping up all over the place next year in California. The California the consumer privacy act, C P R A, it takes effect January 1st, 2023. That’s gonna restrict companies from sharing personal, personal information of consumers, but it also allows consumers to know when there data’s being used by machines for decisioning, a k a machine learning, and to opt out of it. So there’s gonna be a lot of frantic scrambling to become compliant once the first lawsuits start rolling in a about the improper use of people’s data for ai. So a big portion of what is likely to happen in the next couple of years will be increased interest in things like synthetic data that’s modeled off of real data, consumer data, but is not using the, the consumer data to, to build, you know, functional models.

Chris (21:26): And there will be a lot of focus on behavioral data, because behavioral data doesn’t contain personally identifying information. I don’t care who you are, I just know that if you visit my website, you go to the services page, the about page and the contact page, you’re probably gonna convert. So anything I can do to nudge that along, again, with the no personal information at all, is gonna be a money maker. So those are kind of the, the, the areas where there’s definitely gonna be growth coming along. But there’s, there’s also so many unknowns. There’s a whole thing on from stability AI this morning from the founders saying we are holding off and releasing the next model be of our, our image generator, because we have a whole bunch of very serious unanswered questions about what people have done with the existing open source model. Some people are doing really bad things with it because it’s open source, right? It’s, it, it is like, you know, putting out a pile of kitchen appliances and somebody, yes, there is one person who’s running around with a cleaver hacking people to bits. You don’t have control over that once you open source something. So they’re, they’re struggling with that right now to figure out, okay, how do we live up to the ethos of open source, but also reduce harm?

John  (22:41): No, that makes sense. And I think you’ve seen that happen with some other players where they’ve taken a much more careful approach. And I also thought in this discussion, it was interesting you were talking about how companies are thinking about how AI applies to their setting, right? I I very much, I’m starting to see that as well. You know, you know, you, you might be I, I think what’s interesting, you, you bring up AI content generation, but I think that you know, you might be in one particular industry where you’ve already got industry players who are really successful, and then all of a sudden they’re saying, oh, we might have a competitive advantage if we start developing it. But what do you do? Do you build, do you start building it yourself? Or is it better to go and partner with someone or buy someone? You know, what, what, what are your thoughts about that in terms of AI content generation and the opportunities that apply to existing leaders in their particular industry?

Chris (23:41): You know, one of the things we have, we have a five part framework we call the five Ps purpose people, process, platform performance. And the big question for any company that’s looking at AI is, you know, number one, are you using it as, as, you know, incremental improvement, efficiency building and things like that? If that’s, if that’s your purpose, it’s probably safe to just buy it from a vendor. If it is gonna be part and parcel of your secret sauce as a company, you probably should own it because you don’t want your, the, the fate of your company in another vendor’s hands, right? Unless you’re just outright by that vendor. The second question you have to ask is, do you have the people with the right skills internally to support a build decision? If, if you go the buy route, then you don’t have to worry about that.

Chris (24:26): That’s the vendor’s problem to, to acquire the necessary talent. Do you have the right processes in place, you know, data, governance, operations, et cetera, to not only construct AI models successfully, but also adhere to the labyrinth of regulations that are cropping up around it, right? They’re, there are some regulations about the use of consumer data around the world that have intimate personal consequences, right? China has, its P I P L law, people’s internet privacy law. If you violate that law and you’re, you’re, you’re, you’re found guilty of it. Not only do you get whopping fines from the government of China, but if your executive set foot on Chinese soil, they go to jail, right? So, you know, the, the process and the governance is, is really important. Do you have the right platforms and technology stack to, to support ai? Some companies do, some companies don’t.

Chris (25:18): And then, you know, ultimately, can you, can you successfully use AI as to generate the outcomes you want? There are whole swaths of problems that AI is really bad at solving, right? Because they’re, they’re problems that either don’t have a lot of data, they have very sparse data, very poor quality data or they’re just a problem that you can’t solve with AI because it’s not a, it’s not a, a mathematical problem, right? Think about how, you know, the, the most common examples, think about something like getting people to wear masks, right? That is not a, a, a mathematical problem. That is a cultural problem, right? There’s a an education problem. You’re not gonna solve that with ai. There’s no way to, to automate or to, to even, to, to effectively measure why something as simple as, you know, stick this cool looking thing on your face is, is happening or is not happening. So that, you know, to, to look at the adoption of AI and, and making decision on it. You gotta look at all five of those factors.

John  (26:26): Great. Great. So what about your thoughts on AI content generation? And you also mentioned images, cuz I, I think to a certain extent, what’s interesting to me over the last couple of months is that that’s the aspect of AI generation, content generation. That’s really blown up partly because of that, that incidence in Colorado where you know someone was using one of the AI tools to develop an image that they put into an art competition and, and they won the art com competition mm-hmm. <Affirmative>. And there’s been a, a lot of blowback on that. But there’s also text generation as well. And I think what’s interesting to me is that there are different you know, there are different approaches. A lot of these companies, they might succeed on blog writing or general article writing, right? But then you’ve got companies like the PR industry where I don’t see companies doing as well as that type of content. And I, I wonder if, if that’s partly because there isn’t the demand for PR agencies in, in, in terms of what they’re asking. Although I’ve, I’ve, I’ve spoken to some that, that do. So what are, what are your thoughts about AI content generation, where that’s going and, and then maybe specifically that example with pr. Cause it goes back to your, you know, your work over time.

Chris (27:55): As the tools become easier and more accessible, PR will be probably one of the first consumers to use extensive text content generation, because a lot of the tech, they generators, boiler plate, right? Machines can already very capably write press releases that are better than what your average junior account coordinator is gonna crank out. The, the challenge with the content generation in general is that these models are trained on very large data sets, right? Eluthra, AI’s the piles 800 million documents, basic mathematics, most, most content is mediocre, right? The most content’s, okay, it’s not great, it’s not bad. Like there, there are are a few, you know, winners there a few like total losers, and it’s a whole bunch of meh in the middle. And all of these models are trained on as much data as can be acquired.

Chris (28:52): Most of that data is gonna be in the middle. So the, what the models produce right now is perfectly adequate mediocre content, right? The you, you use any of the major tools right now that they’re on the market and they create readable, coherent, okay, content it fits what Google calls. Nothing wrong, but nothing special, right? And that’s problematic because if everyone and their cousin is creating nothing wrong and not, but nothing special, nothing makes you stand out, right? You don’t have that unique creative edge that humans tend to bring. Now here’s the bigger issue for all of these companies and all these, these companies using AI generated content. Again, I talked to, to attorney Ruth Carter about this AI generated content cannot be copyrighted, period, because copyright can only be held by a human. Even if you are using the tool, the tool did the work.

Chris (29:54): And there are a number of court cases that we went through that illustrated very, a very clear chain of evidence by a court saying, if you use AI degenerate content, you cannot copyright it. And if it if someone rips it off, you can’t do anything about it because it is inherently in the public domain. There was a case, Naruto versus peta where a chimpanzee took a camera, took a selfie, and the photographer tried to copyright it, and the court ruled human didn’t create the content. No copyright. That image is, is in the public domain. So the, the person who created the mid journey image that won that art contest, that’s public domain. You can use that image as much as you want, and the artist can do nothing about it because they cannot hold a copyright on it. And this is one of the challenges that people don’t realize when it comes to AI generated content is legally it is a very different than human led content.

Chris (30:47):Now you can do things, for example, like have a machine provide you an outline, and then you write the content from that. You can take a first draft from a piece of con of AI generated content and substantially rewrite it and improve it. And generally speaking, if you think of AI generated content as being in the middle, literally it’s the, the, it’s the definition of, of average. If all of your staff are below average writers, then guess what? AI is actually going to improve your business to, to get it to mediocre. But if you, if you aspire to anything above mediocre, you’re going to need to, to continue investing in humans to not only deal with the copyright issues, but also to stand out. Because right now, again, machines create stuff that is okay, even in the image generation side. You know, there are still enough artifacts and enough oddities that you can tell the difference between machine generated and human generated.

Chris (31:46): Now, the machine generated stuff some way it’s pretty darn and cool. It is, it’s fun to look at and think, we’re starting to see models generating images and animations and stuff. But we are a long way off still from providing a prompt, having a machine write a novel that is, you know, coherent or producing a motion picture that will, it’ll be a little while for what companies should be thinking about though. A, you should be talking to your lawyers A S A P to figure out how to integrate AI processes in your company without endangering your ability to protect your intellectual property. That’s a big deal. And, and expect to spend a lot of money with your legal team to do that. And then b, figuring out from a purpose perspective, what problem does this does, does this technology solve for you?

Chris (32:35): And does it actually solve that problem? Like if, you know, if you’re in the example of a PR firm, if you wanna relieve an account coordinator of doing the first draft on press releases, that’s a good application, right? You can take that person and then have them become an editor and, and level up those, those releases. Or maybe not. I mean, if no one reads press releases anyway but if you are, you know, creating a content shop you’ve gotta create content that is helpful and useful. One of the things that has happened recently is Google pretty much straight up said, Hey, we’re not gonna allow content that is low quality to rank anymore machine generate or human generated, which we’re, we’re, we’re trying to curtail any benefit that that content mills and spam farms generate. And that bar isn’t just a one-time announcement that’s, they’re gonna keep ratcheting that up to say, okay, we expect your content to be higher and higher quality and AI won’t solve that for companies.

John (33:28): Right? And, and Chris, you know, a point there, it doesn’t mean that Google is saying you know, you can’t use AI generated content. But I think the point that you are making and is in the industry, and as I spoke, speak to people across the industry, you have to have that editor. You have to have that marketer, that communicator there who is coming up with additional ideas. I mean, that’s really the value, right? With the with these tools. It’s, to me it’s that initial process of what do I need to write? You know, you, you made that earlier point about SEO and search analytics. So I, I, I think you’re making a mistake if you’re, you’re only, you know, you’re only using the tool cuz the production and the content that you get back is, is just not gonna work. You have to have a human involved. But the thing is, it’s a cultural change, right? For those writers, you know, in AI content donation. Have you have you seen reluctance on the part of companies to do this? Not for some of those issues, but for the problems with, you know, that digital transformation on the part of employees,

Chris (34:38): People, you know if this was before March of 2020, I would say that companies are, you know, were obviously being held back, but beginning into March of 2020, digital transformation was kind of forced on the entire planet, right? We spent three months all trying to figure out how to take every possible business and make it work from our living rooms because we had to, and if you look at the history of digital transformation stuff, what happened during the early years of the pandemic was we basically accelerated some companies in some industries, 10 or 25 years in three months, right? You take old school, like manufacturing companies, like, okay, now you’re gonna learn how to be a hybrid company because your accounting team can’t work in the office anymore. And, you know, seeing something, companies that were a little more forward thinking say, yeah, you know what, our company doesn’t need an office.

Chris (35:30): We can, we can still get work done from wherever. And that bleeds over into the use of other things like AI and things. Because the more accepting you are of change and the more tolerant you are of, of change in new technologies, the faster you’ll get benefit out of any kind of machine learning tools, right? If you are willing to adopt, you know data-driven attribution models for your a your analytics, you will pivot faster and deal with unexpected changes in the market faster than the competitor that is purely opinion led that doesn’t see the, the, you know, the ground changing underneath them. The, the last three years have been massively disruptive to every single company, you know, regardless of, of industry. But those companies that had the resilience and the agility to deal with it are also inherently the kinds of companies that will benefit more from ai.

John  (36:30): Great. Well, Chris, I, I really appreciate you spending some time with me today in the marketing ai chat podcast. Thanks so much.

Chris (36:42): Thanks for having me.