Podcast: Play in new window | Download | Embed
AI’s Role in B2B: Tool vs Leader – Setting Clear Boundaries” This episode dives into the critical distinction between using AI as a supportive tool versus letting it drive your brand strategy. We explore Resound’s core philosophy of maintaining AI in an assistant role while keeping human decision-making at the forefront. Learn practical frameworks for establishing clear boundaries in your AI usage, including when to leverage its capabilities and when to rely on human expertise. We’ll discuss real-world examples of companies that successfully maintain this balance and those that have struggled by giving AI too much control. The episode offers actionable guidelines for B2B brands to harness AI’s power without compromising their strategic vision, including specific checkpoints for reviewing AI-generated content and maintaining brand authenticity. Perfect for marketing leaders and content strategists who want to develop a responsible AI implementation plan that enhances rather than replaces human creativity.
Contact: Mike Jones mike@resoundcreative.com
Discuss at https://www.linkedin.com/company/resoundagency
The show is recorded at the Resound offices in ever-sunny Tempe, Arizona (the 48th – and best state of them all).
Show Transcript
Mike Jones (00:43.557)
Welcome everybody to another episode of the Remarkable Brand Podcast. This is Mike Jones with my double co-host today, Sam Bagel and Chris Stadler. We’re going full radio voice today, I think. I think that’s what’s happening. All right. We’re going to be talking all about AI today. Everyone’s favorite topic. We have not yet burned ourselves completely out on the AI topic since our last episode on Inbound. I think maybe we’re getting close though. I think we might be a little tired of AI, but we got to keep talking about it.
Sam (00:54.286)
Woo! I like it, Mike. I like it.
Mike Jones (01:13.425)
because it’s really important and we have some really interesting things to consider when talking about AI with your writing, your blogging, your content that you’re putting together. So we’re really excited to talk about that. And before we do, Sam.
Sam (01:28.802)
we gotta name 10 things and this is a good one today because I think this actually will, for our listeners and our viewers, have some practical real world applications, guys. Because we’re gonna name 10 use cases that you should not use AI for, okay? 10 things you should not use AI for, Mike.
Mike Jones (01:45.231)
Hmm
Chris (01:48.592)
So are we even gonna have anything to talk about after this?
Mike Jones (01:49.266)
mow in your lawn.
Sam (01:52.576)
Yes, I think so. Mike said mowing your lawn.
Mike Jones (01:54.181)
Don’t AI your lawn or you might end up with a hairless dog.
Sam (01:59.519)
Mmm, that’s good. That’s a good tagline right there. I’m going to go ahead and say, do not use AI to raise your children. That probably won’t work.
Chris (02:00.496)
that’s really good.
Mike Jones (02:01.841)
Mmm.
Sam (02:12.462)
Chris, what do you got?
Chris (02:13.779)
Well, I was going to say don’t use AI to write your wife a love note, actually might be a good idea to start with.
Sam (02:18.082)
Good one, yeah, classic, yeah.
Mike Jones (02:19.217)
But you did, didn’t you?
Sam (02:24.944)
Just to unblank the page.
Chris (02:25.86)
Give me some ideas. I’m blank the page from the AI.
Sam (02:28.75)
All right, we’re generating some good ideas here already. Mike, what do you got?
Mike Jones (02:31.409)
feed in your wife’s persona. Give it some some info. I’m gonna go don’t AI your next cup of coffee. Just don’t do it. Coffee is too important people.
Sam (02:34.389)
Mmm.
Sam (02:41.512)
yeah, mm, yeah. Yeah, probably not good. Probably not good. Don’t let AI diagnose your car problems that you had this morning.
Mike Jones (02:58.009)
One of these days, maybe.
Sam (02:58.062)
Actually might work though. Yeah. Okay.
Chris (02:59.152)
That’s so close to my one, Sam. I was gonna say don’t let AI change your oil.
Sam (03:04.347)
that’s good too. Yep.
Mike Jones (03:05.509)
Hmm. It always strips that that nut in there and leaks everywhere. I mean, just like just like the people do.
Sam (03:09.967)
It does. It’s a big problem.
Mike Jones (03:15.857)
you
Sam (03:18.75)
All right, Mike, what do you got?
Mike Jones (03:19.729)
I’m gonna go with don’t use AI to write your congressman.
Sam (03:30.958)
Okay.
Mike Jones (03:33.007)
I don’t really know why you wouldn’t do that, but I just think like, hey, you know, on principle, just don’t.
Sam (03:39.853)
Yeah.
Chris (03:40.09)
Well, I just think if, if they’re not gonna read it, I’m not gonna write it. So you might as well use AI, right?
Sam (03:45.25)
Hmm. Yeah.
Mike Jones (03:45.393)
Don’t use AI to read all of your constituents’
Sam (03:50.57)
Okay, that’s, now there’s, yes. Well, that actually might be a good idea too, Mike. You’re coming up with good ideas. I’m gonna say don’t let AI be your dentist. We all know what AI does with teeth.
Chris (03:50.894)
You
Mike Jones (03:52.791)
hahahaha
Hahaha!
Chris (04:07.706)
Don’t use AI to feed your dog, because I feel like it would be a softie and it would overfeed your dog and you’d have a fat dog.
Mike Jones (04:16.049)
Why is AI a softie? We’ll get into that later. I’m gonna go, don’t AI your next holiday Christmas family photo. Because your children need just two arms, I promise you.
Sam (04:19.309)
Last one, Mike.
Chris (04:20.206)
It’s always saying yes.
Chris (04:28.484)
Good one.
Sam (04:30.914)
That’d be a nightmare. That’d be a good Halloween card. Because everybody sends out Halloween cards, right?
Mike Jones (04:34.927)
Mmm.
Chris (04:35.044)
Yeah.
Mike Jones (04:37.741)
Everything’s just like a little bit off like every every person’s arms are like just a hair too skinny Too long like alien arms All right, well that was awesome all right Chris Why don’t you give us a little intro and ask us? Our first question since you’ve been kind of thinking a lot about AI and content writing
Sam (04:39.874)
Yeah.
Sam (04:44.686)
And too long. yes.
scary stuff.
Mike Jones (05:06.179)
and how brands should be thinking about that.
Chris (05:09.104)
You know, I have been thinking about it a lot and I’ve been working it into writing processes and I’ve learned a lot from that. yeah, this is one of my favorite topics right now. How do you get more, how do you get AI to do stuff for you but then still do your job and do great stuff, right? Do great work. And so, but first I wanted to ask you guys, why are we doing this podcast? Why?
Why do we want to help other people, our clients, the audience out there? Why do we want to help people understand how to do this?
Mike Jones (05:43.153)
You mean with AI?
Chris (05:44.752)
Yeah, why do we want to help everybody? What are we trying to do?
Sam (05:50.7)
I think it’s, you we talked about this a little bit on the last episode, but I think we’re starting to see that it’s helpful to have a policy, whether that’s you personally or you within your firm to know and understand what do we do with AI as a firm or what do I do? How comfortable am I with it? And then start to build some policies around it. And we know that a lot of firms have already done that and have started to do that.
to be careful, but really kind of goes down to your core ethos. Like what do you believe? And then where does AI fit into that?
Mike Jones (06:30.203)
Yeah. And I think just, obviously AI is hot, right? It’s, a popular tool. It’s a topic. It’s kind of coming up a lot for a lot of firms, not only just in their marketing, but really in all of the work that they’re doing, and especially for anyone doing B2B services, you’re in professional services or financial services, especially tech companies. you know, AI is getting embedded in almost everything that you do, or at least should be considered.
or it’s being forced on you to some degree. And so like, think part of the reason we’re talking about is just because it is this issue that everyone is dealing with. And we want to really think about how do we as people who believe in authentic brands and brands that are true and honest to who they are, how should they approach AI as a tool in especially their like communication and how they develop their identity?
through all of their content that they’re putting out into the universe. Cause I think there’s a lot of pitfalls that a lot of firms are gonna start running into if they’re not already. I’ve just starting to lose their sense of identity, lose their sense of personality, lose their sense of distinctiveness and what makes them remarkable and special and interesting if they come about using AI in the wrong kind of ways. And so I think for me, that’s a really big reason why I wanna have this conversation, why I think
we need to like continue to have this conversation. I doubt this is a single episode conversation. Obviously we’re honing in on content today and content writing, particularly with like blogs and articles and that kind of thing. But I mean, this has much deeper implement implications for every brand, all the way from like the creation of your brand, right? And how you establish your primary like attributes, who you are as a firm.
Sam (08:11.32)
Yes.
Mike Jones (08:23.825)
through your core assets like logo and name and tagline, colors, fonts, your core messaging, your verbal guidelines, all the way through into how you execute every one of your communication and marketing deliverables, especially for those on the marketing side of the business. I mean, just, think there’s a lot of opportunity to use AI and to make your processes more efficient.
But I also think there’s a ton of opportunity to just kind of lose your way really quickly when it comes to establishing a really unique and special brand. So that’s why I’m excited about talking about.
Sam (08:56.152)
You know, we talked, we, joked about writing your wife a love note using AI and there’s a little bit of truth in that, right? Cause like, you don’t want to just use a tool and lose your soul a little bit. Right. And that’s kind of what we, what we mean about, you know, when it’s branding and AI, how do you stay true to the brand while harnessing this great tool? We use it a lot, but you have to make sure that
Mike Jones (09:01.745)
Yeah.
Mike Jones (09:09.346)
Hahaha!
Sam (09:24.94)
you’re not using this cool new tool and at the same time you’re kind of letting it run away and make decisions for you that ultimately might mean, yeah, we’re just not who we thought we were anymore.
Chris (09:38.832)
I mean, didn’t you guys ever wish when you’re in school, you could just send like, create a robot of you and send it in place of you and then it could come back from school and just kind of give you the high points.
Mike Jones (09:49.019)
Sure. Yeah. mean, we all want to cheat at some level when we’re not invested in the thing we’re doing. Right. And so I think to your example, Chris, it’s like if you’re going to treat your brand like school, this thing that you dread and you don’t really want to do, and you want to go to market without doing any work, then AI seems very tempting to just kind of offload the whole process to it and sit back and go, I don’t have to touch my marketing anymore.
I think that there’s a fundamental problem I have with that philosophy. And that is that you’re not creating anything at that point, at least not anything that’s true and not anything that’s like really real AI is not capable. We’re, clearly seeing this AI is not capable of creating things that are fully original. it can’t think for itself. It is.
you know, very replicative, everything it’s doing is derivative of something else that’s already been created, often by a human being. And so as it gets used more and more, what’s happening, at least what I’m seeing is everything starts to look and sound the same that comes out of AI, because it’s only as good as the inputs that are given to it. One, this like bank of knowledge of content and imagery and like
pieces of data that have already been fed to it. And I have heard from people on like the data science side of this, that they’re already concerned that there’s not enough knowledge categorized and collected by human beings for AI to get much smarter than it is today and more original than it is today. Like they are hungry. Like they are like, we’ve absorbed most of the internet that’s public into these AI engines.
and fed them these data sets. And now they’re looking at like, well, where else can we get data? And they’re looking at private libraries or private institutions who have lots of content like Reddit, Wall Street Journal, X or Twitter. But even then a lot of that work is derivative, right? It’s not all that interesting. And so I think there’s a hard limit for anyone using AI to say, we can just have it create for us, like wholesale.
Chris (12:10.105)
Ahem.
Mike Jones (12:10.159)
And then beyond like the data that’s already been fed into the model to train it and that it’s pulling from as a resource. the other issue is it’s only as good as the inputs you give it, right? As a user. I think a lot of users are lazy and they don’t, you know, copy paste inputs, prompts. I can’t tell you how many times I’ve seen people selling whole sets of prompts for marketers. and it’s like, man, that’s just going to get very similar answers every single time.
regardless of which brand, you know, plops that into their chatbot or their AI bot.
Chris (12:46.468)
Yeah, well, and I thought of another benefit too of us talking about this because as I think about the people we know who are in marketing, how do you craft your career and how do you basically, where is AI’s place in your work? And then how do you have to build so that you’re gonna be really valuable so that you’re not kind of doing work that AI could do, but you’re focusing more on the original thinking.
creative process might actually become more of a topic from now on. Like how do you actually come up with new ideas and things like that.
Mike Jones (13:24.954)
Yeah.
Chris (13:27.342)
Yeah. So, so, so I have a question for you guys. Why shouldn’t we just trust AI to write blog posts? Just set it free, man. Just let it go. Why can’t we, why can’t we trust it?
Mike Jones (13:36.567)
Ha ha ha!
Mike Jones (13:43.077)
Sam, you wanna take a first stab?
Sam (13:45.782)
Well, I mean, it of goes back to what we were just talking about. You can absolutely do that, Chris. You can put in five to seven words and boom, you’ve got a 600 to a thousand word blog post. That’s going to sound like a thousand other blogs online. And again, you know, hey, maybe that’ll get some people to come to your website. Maybe the SEO will be decent on there, but.
There’s a couple reasons why I would not do that. Number one is people are starting to see AI writing like they see stock imagery. They can sense it right away in the first sentence or two and boom, they’re gone. Because honestly, people don’t want to read something a robot wrote. They don’t. Just like, you know, I was just talking about this before we recorded, I cannot stand, you know,
going through YouTube and you come across this video that looks interesting and right away it’s an AI voice. And it’s like, no, I don’t care about this. I’m not here for robot chatter. I want to read or listen to something that a real human created. And…
Chris (15:02.96)
I was looking at a HubSpot blog post and I was like, wait a second, this article only has four fingers. I could tell it was AI. And I was just like, I’m gonna read it anyway, like, you see what it says. Yeah.
Sam (15:12.963)
Yeah.
Mike Jones (15:15.442)
It’s because it has so much depth to it, didn’t it?
Sam (15:21.368)
Yeah, and the other thing is, just, you know, if you’re gonna do a, the fastest way to get a blog post is to have ChatGPT or Claude write me something real fast. I just don’t have time to do it. Again, like, there’s just gonna be no, there’s gonna be no personality to it. It’s not gonna give you a point of view. So it’s not gonna be interesting. It’s just gonna be, you know, some random facts or some very vague facts.
that actually might not be true. So I’ll start with those two things. What do you guys think?
Chris (15:58.064)
I have a follow up question, if I may, for both of you guys. You mentioned point of view. Why is that the secret sauce? Why is that kind of a part of what AI, why is it so important, I guess?
Mike Jones (16:15.747)
I mean, that goes back to brand, right? A brand without a point of view is not really a brand. It’s just a show. And we see that, you know, lot of industries we work in that that point of view is hard, right? It’s really hard to come to a singular point of view as a company, as an organization. It’s hard to be willing to like take the risk to lean in on that and make that your central message. You know, I think there’s a lot of fear of like, well, if we’re not
If we’re not really broad and we don’t talk about everything, which means you really don’t talk about anything. you know, if we don’t cover every single industry and every single solution and every single topic that we possibly can and not offend anyone. So we don’t want to have a point of view. Cause we might offend like a potential client. I think that fear holds up a lot of brands. So the challenge I have with AI in this problem is AI won’t give you that point of view. Right? So it’s a,
lowest common denominator generator when it comes to whether it’s imagery or content or video or anything else you’re doing, it’s giving you kind of like aggregate of all of humanity. Here’s everything humanity has created over the last 40 years. We’ve trained the AI model on it and it’s going to execute kind of what it thinks the general populace expects when you put that prompt in, right?
And so it’s not interesting. It’s not necessarily unique. It’s gonna feel derivative. It’s gonna feel very much like, this feels like something else that is unique, but not quite, right? And for your firm in particular, it’s just not gonna have any kind of interesting point of view. And the reality is like point of view can only come from human beings, right? To have a point of view means you have a singular view.
Like I am a singular individual and I see things a certain way because of the unique set of experiences and history and biology and environment around me and how I think and the values that I hold, the things that drive me. And AI doesn’t have any of those things. What AI has is aggregate data, right? So it has many points of view and then flattens those out.
Mike Jones (18:38.517)
That’s just how it works. That’s fundamentally how it works. maybe there’s some data scientists out there that’ll tell me I’ve got it totally wrong. But as far as I can tell, that’s it. And you can kind of fake a little bit of point of view. You can say like, think like this kind of person and have a point of view and that’ll get you closer. But that’s still an abstracted point of view from a real person or even from a company.
because organizations can have a pretty unique point of view when they’re really tight on a core set of values and on purpose and on the things that matter to them, right? The experience that they’ve had that usually does flow from founders, right? And if they’ve done a really good job of building a tight culture, they’re going to have more opportunity to have that point of view really like staked down and not stray from it or not let it become kind of fuzzy and really blurred out. So
Yeah, I don’t know, Chris, if that totally answers your question, but I…
Chris (19:40.814)
So are you guys saying that that point of view, having a point of view actually makes you believable, interesting, convincing, a legitimate strategic advantage in communication?
Mike Jones (19:54.117)
Yeah. mean, who do you want to hang out with that doesn’t have a point of view? Right. In your personal life, why would you want to hang out with brands that don’t have a point of view? Right. Especially in purchasing decisions that you feel like are really important to you or to your organization. Would you rather buy from like, kind of like some amorphous, like unknown, doesn’t have a point of view, isn’t really clear on anything?
Sam (19:59.416)
Yeah. Yeah. Right.
Chris (20:21.113)
Right.
Mike Jones (20:21.817)
Like time and time again, like marketing studies have shown like clarity comes from being specific and one way to become specific is to have a singular point of view and say like, this is what we see in our industry. Here’s what we see as right. Here’s what we see as wrong. Here’s how we do business. we do it this way and not that way. And we hold to these values, these fundamental beliefs about how the world works and they don’t have to be like,
You know, I think some people are like scared of like, well, you know, we don’t want to be like a religion. We don’t want to have like that kind of specificity. And I don’t think anyone’s arguing that you need to be a, a religion to be a business. in fact, that probably be a bad idea. although some of them do get kind of cultic, right? they borrow things from religion and how religions have been built. well, part of that is having a point of view, right? Like that’s interesting to people. People go, yeah, I agree to that. I subscribe to that.
Chris (21:01.776)
call.
Mike Jones (21:19.513)
And a company doesn’t have to have like some fundamental, like super like intense point of view on the entire world. They just need to have a point of view on their industry, right? On something that they are involved in. it doesn’t have to be this like world changing cause either. think some companies have gotten really hung up on that in the last like 20 years and thinking that like somehow we have to be like fundamental change makers across the entire world in every category. And it’s like, no, just
pick something and hopefully it actually represents your culture and your brand and where you’ve come from and your history and the experience that you have and maybe has something to do with your products and services, I hope. So I’m getting a little ranty here, but, and we got an offer from AI, but AI won’t give you those things. It will not give you those things.
Chris (22:03.865)
Yeah.
Chris (22:08.229)
Yeah.
Sam (22:09.474)
Yeah. You have to remember too, that people want to purchase from and work with other people. Right. And, that’s especially in B2B, those personal relationships are everything. And so if the first impression is, this company, this firm, didn’t have the time to like actually write this article and provide value to me and what I’m looking for. Am I, am I going to get the time?
that I need in this relationship. how would your wife feel? you didn’t even have time to write me a love note for Valentine’s day. Okay. Right. Like it’s kind of that same energy of like, Hmm, you don’t have time to like write this out of like how you would approach this problem. Then what am I actually getting here? And the second thing is, you know, to provide that value, if you have, you know, a prospective client or
somebody in your audience that is looking for something, why would they come back to you if that was their first impression? Couldn’t they just do that themselves in chat GPT? Because more and more people are using it, right? So, this is chat GPT, I could just go do that myself. Why am I even here, right? Like there’s no value provided.
Chris (23:31.044)
Yeah, but how cool is it though that you could just start a company and not know anything about the industry? All you have to do is just have chat GPT create all your stuff.
Sam (23:42.728)
Mm-hmm. Yep, that would totally work.
Mike Jones (23:44.197)
Yeah, I think there’s probably, and this is probably going to get to a question we need to start answering, which is, when, when can I use, you know, AI, when can I plug that into my process? If it’s so like, not helpful with my brand and helping me develop a point of view, it’d be interesting to stand out to be specific. Well, then what can I use it for? Can I use it? Should I use it? and I think that’s maybe where like,
If you’re doing research or you wanna understand not a point of view, but understand kind of like the aggregate point of view, right? The kind of humanity view on some topic. I think actually AI is really good at giving you that. If you wanna like hone that in and kind of be like, okay, I kind of have this assumption, let me test that out a little bit.
Let’s say I have an assumption about like, think these are the challenges that this particular type of person I’m trying to target and get in front of. These are the kinds of challenges that they have. Well, go ask AI, right? Now it’s not going to give you really interesting answers, but it’s going to give you aggregate answers. I think that’s a great way to use AI is kind of test things, hypothesize things, kind of like have a sounding board of like, okay, I have some, I have an assistant who has access to like,
A lot of information, right? Now it can’t give me like really specific answers. I can’t go like be like, okay, what’s the smartest answer to this question or the most nuanced answer, but I can at least get the aggregate answer. And so if you if you’re using it as a research tool, you know, test hypotheses, maybe generate like an initial idea or like a sounding board for an idea that you have, right?
I think one of my favorite things to do with AI in terms of like content generation is like, ask it a question and tell me, give me 10 answers to this or 10 things. And then I say, great, give me 10 more. Great. Give me 10 more. And by the time I’m at like 30, 40, 50, then I might get something that’s actually interesting. Maybe because it’s like, by that point, it’s like really digging. And there’s times I’ve actually like, I took it to a hundred on like one question.
Sam (25:50.722)
Yep.
Mike Jones (26:06.929)
And it’s just stopped it like chat GPT quit on me at 75, like answer number 75. was just like, but can you do it? and it was like, you know, I, I probably a little bit notorious in our organization for saying this, but like, I feel like, you know, chat GPT or maybe co-pilot some of these other like kind of basic content tools, content AI tools are probably as good as a college intern.
Sam (26:14.946)
Hahaha
Mike Jones (26:36.825)
Like it, you know, they have a wider breadth of things to pull from than a college intern and they’re certainly a lot faster than a college intern at like getting an answer. But the answers you’re going to get back are like, yeah, that’s what I would kind of expect from someone who’s just on the internet researching for four hours a question and spitting back answers to me. And even then, I think probably if you had a college intern spend four hours on one question, they probably, if they’re like,
halfway smart and good at being inquisitive, would probably come back with some interesting answers that a chat GPT or co-pilot’s never gonna give you.
Chris (27:17.296)
Yeah, so it’s kind of like, you know, made me think of sometimes I’ll use it for, I’ll think of, okay, before I look at anything, because there’s phenomenon in creative work where, or in just, I don’t know, cognitive psychology or something, where there’s top-down processing and bottom-up. And so if you’re like scrolling the internet, you’re doing bottom-up processing, you’re responding to stimuli, right?
Top down is where you’re like trying to solve a problem and that problem is something you own in your head. And so a lot of times in order to take advantage of top down processing where it’s gonna be more original is I’ll write my list and then I’ll say, okay, once I’m done with my list, I’ll ask GPT to say, am I missing? Because it’s pretty well-rounded, it’s gonna give you some of those things you forgot about and you’re like, duh. Someone like me is like really useful.
Mike Jones (28:13.509)
Yep. Yep. Yeah. I even get concerned with like people who use it to like build an outline from scratch. You know, it’s like, Hey, you know, write me an outline on this topic. I’m curious, like how far you can take that before it’s very derivative of a lot of other articles or like pretty soon it’s like everybody in your industry is writing that same outline because they’re all using the same tool to like generate it.
Chris (28:39.856)
Mm-hmm.
Mike Jones (28:45.33)
It might have like slightly different wording and stuff or like certainly if you take an editing pass on it, maybe you’ll massage it and it won’t be like Incredibly apparent on the first pass that like this is the same article But you know, you’re only gonna get so many variations on it We’ve even seen like we’ve taken podcast transcripts that I thought had really interesting kind of little bites in them, right like little
examples or metaphors or case study type, you know, facts and figures type stuff that we know about and we throw those in there in our conversation. And when we ask like a chat, GPT or a Claude to go back and essentially write an article from this, this podcast, loses all those things like completely overrides them. And yeah, the broad strokes of the article are in accordance with what we said in the podcast.
It kind of follows a general outline, but it loses all of the like really specific, interesting things that I think make for really great content. and, I do wonder like, you know, we get another year down the road. It’s like, many, how many pieces of content do all these companies create that are just flat out AI generated and pretty soon it’s like, nobody’s writing anything interesting anymore. And I think what you’re going to do is train people, just ignore all your content.
That’s what’s gonna happen.
Sam (30:12.834)
Yeah, well, and on a technical side, we’re already seeing that Google, YouTube, even Amazon is, I don’t know if they’re starting to filter it out, but they’re asking about it. So like, for example, YouTube will ask like, there any AI generated or manipulated content in this video when you upload a video? I know Amazon, like we do some self publishing through Amazon, they’ll explicitly ask in the publishing process, is there any AI generated imagery or content? Is it, you know,
Is it light or none at all? So they’re starting to ask about these things. I would guess, you know, hey, YouTube, Google, Amazon, all these companies, these big tech companies, they want to give their customers, which by the way is not AI chatbots, it’s real human people. They want to give real people the best content. So if they start to learn that humans do not prefer AI generated content, guess what’s going to happen? You’re going to be in trouble.
If you’ve been doing this for two to three years. Yeah. The other thing, kind of switching back to like, what would you use it for? I love using AI to help me think differently than I can think, right? Like you said, Mike, give me 10 more, give me 10 more until you start to get into some things that just like, I wouldn’t have thought about that. And so for me, a lot of times it’s like, man, I got to come up with a title for a blog post. Hey, here’s all the content.
Mike Jones (31:10.641)
They’re going to derank it. Yep.
Sam (31:39.256)
Give me 30 out of the box, interesting or creative. And almost every time there’s not a single line where I’m like, that’s it perfectly. But I’m like, ooh, that’s interesting. And that second one, number 28 is really interesting. There’s some interesting things in there that I wouldn’t have thought of. And then you, as the human, you can curate those things and start to piece them together. I’ve had some, I mean, it takes a lot of the time out of like the.
What is going to be the most interesting part of this blog post and how do I put that into a title, right? How do I get people to be interested in the content I’ve just generated? And I think AI does a really good job. If you kind of learn how to ask it the right questions and to just maybe keep pressing it, like give me some more, make it more like this, give it more of this touch.
Chris (32:27.408)
I think that’s a really good point because when you think about writing and when I think about journalism, Sam, you and I both went to a journalism school. A lot of it’s about asking good questions, right? It’s asking the right questions and also seeing points of tension where you’re like, okay, that seems suspicious. Even in brand workshops, we do that and it helps our clients, right? Because we’re seeing things that don’t seem to agree.
that we see because we’re looking for him, but maybe the client doesn’t see because it’s just like, well, it’s just what you do. Right. And, and so that, that helps us really find things that are, that are interesting. I think the same thing is true with writing, writing articles when you’re thinking about your persona. Like if you haven’t given AI, you know, the AI, your persona, or even if you have, they’re not going to be able to really sympathize with the person. They, don’t, they don’t know that person, right. That you’re writing to.
But you do, and you have that in your mind. So you’re looking at it you’re thinking, okay, you’re reading the article that was generated or the outline or whatever. And you’re saying, okay, yeah, I guess that sort of is it, but it’s not completely it. How would I explain this to my uncle? And who is the persona? This is the guy I picture when I picture this persona, right? How would I explain this to Uncle Joe or whatever? And now…
Now you’re getting really conversational in your brain and even your questions as you look at these things are gonna get a lot more conversational. You’re be like, yeah, Uncle Joe, yeah, he sorta cares about this, but here’s what he really cares about, right? And now you have a great question to ask AI.
Mike Jones (34:05.775)
Yeah. Yeah. Yeah. I think that’s even even there’s like, within that idea is that there is a flaw even with a persona, right? Even better than a persona is an actual person. Like if I can have a specific client in mind, when I’m writing, man, I can get even more specific than I can with a kind of somewhat abstracted broader persona, because there’s a level of just like, it’s just not specific.
as, as specific as a, as a singular human person, human being, that I can talk to. And I can think about, like you just said, with your uncle, Chris, like, you know, your uncle in a way, like, you’re just not going to know your persona. you know, I would even argue like, make your persona as tight as a single, like person, right? Like, like make it that tight. and yeah, maybe you can use that with AI and feed it some of that information.
I personally still think that AI is going to abstract that. I think it’s going to get derivative pretty quickly, no matter what you plug into it, only because the way that it thinks is not, it’s not a person. It can’t really think about anyone. It can’t think about even what you gave it. All it’s doing is like a giant word association cloud. That’s what it’s doing and doing probabilities of word associations. So.
Chris (35:32.922)
Yeah, howdy, howdy.
Mike Jones (35:33.617)
It’s just going like hey, I’ve been fed all these all these pieces of words that flow together. I Expect to see this kind of answer when you ask this question with these words in it, right? It’s not cognitively understanding. What what kind of person are you trying to talk to you? It’s just going like okay. Let me go to my bank of data and go. Okay, you said accountant. Okay, so what
what content surrounds the word account, right? And then what content surrounds these additional pieces of information that you’re gonna give me in your prompt in order to generate this. So it’s only as good essentially as the specificity of the data it already has and what you give it.
Chris (36:14.799)
Yeah.
Chris (36:21.806)
Yeah, I saw this quote, AI is no smarter than your toaster. It was in a Cambridge conversation about AI or something like that. And you’re making me.
Mike Jones (36:30.673)
It makes me sound really generous making it sound like a college intern
Chris (36:36.208)
A toaster. Maybe a toaster could be your intern, depending on the quality for your recruiting.
Mike Jones (36:42.065)
Not my toaster at home.
Sam (36:43.214)
Would you rather have a toaster or AI? You can only have one. Dude, I’m not giving up my toaster, man. I love toast too much. I love toast too much.
Chris (36:47.482)
Or the president. A toaster could be the president. Right in candidate. So yeah, so you made me think, we don’t understand how humans look at their sense of self or a relationship with other people. We sort of do, but how do you program a computer or an AI? How do you…
Mike Jones (36:50.865)
I’m not giving up my toaster. I love toast.
Chris (37:14.352)
set up something like that that’s gonna understand those relationships, it’s gonna be very low resolution. It’s just gonna be very ham handed, demographic driven rather than a feel of that person, how you would interact with that person there in front of you. When would you smile? How would you shake that hand or give that hug or whatever? There’s just no way, right? No.
Mike Jones (37:19.643)
Yep. Yep.
Sam (37:35.192)
Yeah. Because AI cannot meet with your clients. You can’t outsource the Zoom calls with your clients. AI can be integrated into that, but you can’t take AI with you to inbound and meet with your friends or your potential clients or your current clients. It doesn’t learn like you do with human interactions and human learning. Every time we meet,
with somebody in person or even on a Zoom call or a phone call, we learned something new about him. this person likes Taylor Swift. that’s really interesting. It just reshapes a little bit about that person you’re talking to and it just continuously kind of feeds that persona in your mind.
Chris (38:20.816)
Mm-hmm.
Mike Jones (38:27.249)
It’s really good.
Chris (38:27.47)
Yeah, have you guys ever asked AI to do like make a joke or have a sense of humor about a topic?
Mike Jones (38:34.533)
No, I don’t believe in jokes or humor, puns definitely. I don’t like those. So what is sarcasm?
Chris (38:39.792)
But you do believe in sarcasm.
Apparently.
Mike Jones (38:47.313)
I don’t think I’ve ever done that. I don’t I’ve done that to like Siri, those kinds of things. But I saw I have seen there was a this was like a year ago, I think there was a guy who entered a joke, like a stand up comedy contest. And he entered it as a representative of an AI bot, where he basically said, I’m not gonna write any jokes. I’m just going to
give you jokes that the bot gives me as I type them in and see how far I can get in the contest. And it was like an epic fail from the standpoint of humor. Like they were not funny. Like now this was a year ago. Maybe they’re a little bit better now. Maybe there’s like an AI bot that’s a little bit more trained on jokes. But like it just really struggled. It really struggled. And it was very interesting.
to see kind of like, kind of get behind like, okay, how does an AI like think quote unquote. But I love this conversation. I think we’re gonna have to do at least another episode. There’s probably a bunch more episodes on AI coming up. But I think this was a really good starting point for us to kind of talk about like the pitfalls of AI, some of the challenges with it, the edges of where it’s useful. But I think our next one should really be about, you know, how can I use it?
Chris (39:53.52)
Thank
Mike Jones (40:11.835)
How can I leverage it to the best ability that I can, especially as a marketer in a professional services firm? How can I really leverage it well? So I don’t know what you guys think, but that would be my proposition. Let’s talk about that next time.
Chris (40:27.728)
That sounds great.
Sam (40:27.777)
I love it.
Mike Jones (40:29.208)
Sweet.
Sam (40:35.554)
Great. No, no, No, no, that was perfect.
Mike Jones (40:36.305)
Do you need a little tag, like an outro tag thing? I don’t think you need that, right?