Transcript
Dan Sanchez: There is a whirlwind of hype out there about AI, and it can be really confusing for marketers or anybody trying to get productivity out of AI out there. So in today's episode, we want to talk about 16 different AI myths to uncover the fact from the hype. So we're going to be going through and busting them, validating them, or talking about the nuanced truth that relies between lies and reality.
Dan Sanchez: So welcome back to the AI-driven marketer. I'm Dan Sanchez and I'm joined by my brother, Travis Sanchez.
Travis: So pumped to be here.
Dan Sanchez: And we are covering the help from the hype because normally we cover the news related to what's going on in the AI world and contextualize it down for marketers of what you actually need to know, what's actually helpful.
Dan Sanchez: But because there's not a lot of news in the AI world right now, at least that's relevant for marketers, so we're going be covering some of the hype that's been floating around, some of the myths that have been floating around that I hear from time to time on social, on other podcasts, and in just conversations I have with marketers and every day.
Dan Sanchez: So we got 16 of them. Let's dive in. The first one is the biggest one. We're going to come out swinging for the fences here and it is AI will replace marketers and take jobs.
Dan Sanchez: What do you think? Myth or reality?
Travis: Man, I haven't seen it to be a reality yet. I've definitely seen different companies, including Salesforce, lay off thousands of people because they claim they have AI support that they don't need those current roles anymore.
Travis: Were they marketing jobs? They didn't say or specify, but I don't think they were. i haven't heard of it. If anything, it would be copywriters. But then again, as someone who doesn't understand the grammar rules as well as someone who has gone to school for it, and I know AI is probably does.
Travis: There's just something about having a creative writer on your team that is not AI to proofread, to read through, to make sure it's on brand. AI does a great job, but there's something about a human element, though I think that would probably be the first potential marketing role that would be cut.
Dan Sanchez: This one's hard because I don't think it's and don't think it's quite reality, but i don't think it's busted either. It's somewhere in the shades of gray area.
Travis: Nuanced? Nuanced?
Dan Sanchez: It's nuanced. This one's nuanced because I've talked to many copywriters, especially the ones that have lost jobs or more like lost clients, are freelance copywriters.
Travis: Yeah.
Travis: Interesting. oh
Dan Sanchez: Oh, that's hurt a lot. And I tell you, talk to anybody who's in the writing game as a contractor. It hurts because now people don't need a first set on like web copy and they don't under even, and it's, it's hard because they don't understand the nuance between even great copywriting and what chat GPT provides because they're not educated in it.
Travis: Right. Right.
Dan Sanchez: So they're just like, Hey, chat GPT gave me something before it was like, well, I don't know what to put here. So I need somebody to write something. And now they can have first drafts that are good enough. I've heard of companies letting go of copywriters specifically for other reasons, bad fit in the case I'm thinking of, and then never hiring them back. But but partly because the marketer was already a good enough writer and could kind of parse through what AI and fix it up a little bit. Like AI could always provide the rough draft and the marketing director could always give it enough feedback and guide it to get what they needed.
Dan Sanchez: So I think that's happening for marketing teams. I think right now where we're seeing the job crunch is around customer service for frontline customer service. So like people answering phones, receptionists and developer roles.
Dan Sanchez: I hear when when I talk to all my developer friends, they're all like, dude, like jobs are way down for entry level developers.
Travis: Wow.
Dan Sanchez: Everybody's talking about that. So development, which is the oddest thing. So that always was going to be the hot career for the longest time.
Travis: Yes, right.
Dan Sanchez: One of the first ones to go. Yeah. in copywriting, but it makes sense. Programming is programming languages and that are very logical and have a certain way of doing it. Now, of course, does it do great code? No, but does it do entry-level code?
Dan Sanchez: Yeah. yeah, it's only getting better at that.
Travis: Yeah. Wow.
Dan Sanchez: so So it's kind of in between our market is losing jobs. Some of them are more of the contracting side and some some in-house, but I think that will continue to be a thing as AI gets smarter. One thing I am for sure on is that it does seem like AI is replacing jobs faster than it's making them.
Dan Sanchez: What do you think about that?
Travis: yeah, I, I, we talked about this earlier, but companies are going to cut jobs and then realize, oh, shouldn't have cut that.
Dan Sanchez: Yeah. Yeah.
Travis: I really can't do that yet.
Dan Sanchez: Yeah.
Travis: And you've ever experienced a chat bot or a prompt system when you have a customer need. you're youre you need a customer assistance person and you get so annoyed that it's not quite taking care of your needs.
Travis: You can't really explain what's happening.
Dan Sanchez: yeah
Travis: And i've happened that's happened to me over and over. You're like, let me just talk to a human so I can explain.
Dan Sanchez: yeah
Travis: I think similar to that, there are going to be roles that are going to need to be brought back from companies that they just cut too soon thinking AI would be able to take care of it. So that's what I think.
Dan Sanchez: Yep.
Dan Sanchez: All right. So nuanced number two, AI content can't be authentic.
Travis: This one's weird.
Travis: Because like you're kind of getting down. i'm I'm the one who put this one on and I'm like, even thinking about it now. It's like the definition of what is authentic. Is it subjective?
Travis: Can... we Is it just a copy? Well, is the it's copying music artists and the song actually sounds incredible, but it was AI created. Is that authentic? Well, it wasn't sung, but but it was created based on thousands of hours of music artists that were authentic.
Travis: Yeah, this is where.
Dan Sanchez: Again, this is one of those nuanced ones. Can it be inauthentic? 100%. Absolutely. We see a ton of it. Can it be authentic? Sure. I think about it being authentic every time when I use it to trans... We will use AI to transcribe this very conversation we're happening and repurpose it into social content and blog content.
Travis: Right. Wow. Right. Right.
Dan Sanchez: Will it be an accurate representation of what we say? Yes, it will be an authentic representation. Will it fill in some of the gaps of things we, like if we give and list of what to do and we leave out a few important details that if asked later, we would have added?
Dan Sanchez: It does. it'll fill in It'll fill in some of the gaps. That's what I love about it. Just like a good copywriter would, who was actually marketing, like understood marketing themselves. Yeah.
Travis: wow
Dan Sanchez: that's what a, copy that's what a good copywriter would do or ghost writer, right. was writing on your behalf.
Travis: right right
Dan Sanchez: I've worked with some humans that do this and AI can do it pretty well.
Travis: Wow.
Dan Sanchez: So is that authentic? I think so because it's not asking AI for hot takes on the myths. Actually, I did ask AI to do hot takes on these myths and it was wrong often. So that's why I'm like, that's why you and I have to talk about it.
Travis: oh
Dan Sanchez: Cause it's, it's not right.
Travis: AI doesn't know.
Dan Sanchez: It can be inauthentic. If you and I were AI avatars talking right now, and we, the real Dan and Travis were sitting back and on a beach somewhere drinking Kool-Aid, hot Kool-Aid.
Travis: Well, mic drop moment is that Dan and I are actually AI and don't exist.
Travis: This is a custom GPT. Just kidding.
Dan Sanchez: Yes. The real man behind the scenes is named Fred and he hasn't been around for a long time. No. People will be wondering that stuff soon, bro. That'll actually be a thing.
Travis: That's true.
Dan Sanchez: It's getting close. I think AI can be very authentic. It depends on how you use it. I always advocate for doing it authentically, which is why I'm always an advocate for podcasts, video podcasts like this one, because we can be authentic and then use AI to repurpose the crap out of it down the line. And it still is authentic.
Travis: let Okay. Speaking of authenticity, you have, i think most people have heard about the band or music group that was created on Spotify. That's completely AI.
Travis: If you listen, I forget what the band is called.
Dan Sanchez: Yeah.
Travis: Do you know what it's called?
Dan Sanchez: No, but I remember I listened to it.
Travis: Okay.
Dan Sanchez: I saw the news. so
Travis: Millions of listens.
Dan Sanchez: Yeah. Yeah.
Travis: There are millions of listens on these songs. The music sounds good. It's catchy, great voice, great instrumentals.
Dan Sanchez: Yeah.
Travis: Okay. Okay. I listened to somebody live who had his acoustic guitar, who sitting in my living room and was playing. The feeling I got when I heard a real human playing a guitar and singing was, it's like, okay, what is happening?
Travis: It just had this feeling that felt authentic. You will never get that experience ever. I don't think with AI, you can listen with your headphones, you can listen to a song, but there's some kind of connection when someone's physically sitting in front of you using their voice and their skills to play an instrument, which can AI replace that?
Travis: I don't think so.
Dan Sanchez: no
Travis: No. Yes.
Dan Sanchez: No. Even if it was a really good robot and played it perfectly, even slightly imperfectly to mirror it, would it be the same?
Travis: Yes.
Dan Sanchez: No. Because the AI, no matter how smart it gets, will never be a person. It will never have values based on real experiences, shared realities.
Travis: Right. I think you said, i think, I think you said this a while ago about creatives and artists actually getting a lot more attention because of the physical nature, the authenticity that AI will never be able to duplicate.
Dan Sanchez: I've said things like it before, but a lot of other people are saying that too, especially around live experiences. A lot of people are predicting like live, live is going to be huge.
Travis: but That's what it was. Got it.
Dan Sanchez: I didn't make that. predict Someone else made that prediction before I ever thought of that.
Travis: but
Dan Sanchez: I'm sure people were making it even back in 2017 or something.
Travis: right Yeah, I think that's similar to the last one.
Dan Sanchez: All right, moving on to the next one, just similar is AI can't be creative. It's just data crunching.
Travis: I think it can be very creative.
Dan Sanchez: It kind of comes down to like, well, what is creativity?
Travis: that well That's what I said, too.
Dan Sanchez: Is it making things?
Travis: Right.
Dan Sanchez: Yeah. Is it good? Yeah.
Dan Sanchez: Is it authentically human? No. Does it need to be? Sometimes no.
Travis: Right.
Dan Sanchez: Do you really care where the art came from that's hanging in the hotel lobby to give a a feel?
Travis: right
Travis: Right.
Dan Sanchez: No.
Travis: Right.
Dan Sanchez: Like, let's be honest. A lot of those are knocked out by all kinds of artists and then curated and it's so far removed. It's different.
Travis: right
Dan Sanchez: I think there will be a lot of that and it'll be wonderful. And we will have ai making us all kinds of weird, funny movies that are just created just for us. You know, that's actually kind of a amazing and scary reality all at the same time, but can it be creative?
Travis: girl true
Dan Sanchez: Yeah. Is it different from human creative creativity? Yeah.
Dan Sanchez: So nuanced.
Travis: Nuanced?
Dan Sanchez: Yeah.
Travis: They're all nuanced.
Dan Sanchez: Probably. That'll be funny if we get through every single one of them and their nuance because we haven't gone through these ahead of time.
Travis: No, the next one's not nuanced. The next one's definitely not nuanced.
Dan Sanchez: Okay. Only big companies can afford ai Number four.
Travis: Busted.
Dan Sanchez: This was a funny one to me, but when I was looking for things, this one came up a lot as far as people's myths. I've curated these from a lot of other myth lists out there.
Travis: Okay, okay.
Dan Sanchez: And this was a common one. I was like, what the crap? Like it's more small businesses. We're all paying even free AIs dang good these days.
Travis: Yeah, it's getting really good. No, I think people just have a, yeah, they they have a mental roadblock to getting started, thinking it's too complicated for them to start using, but it's definitely available for everybody.
Dan Sanchez: for free for your kids. It's, it's
Dan Sanchez: have seen a few people that will like come back from presentations and assume that it was big and complicated, more complicated than it needs to be. I'm like, we could have knocked that out in a custom GPT.
Travis: Mm-hmm.
Dan Sanchez: You didn't need to need big five, 10,000, $20,000 tool. So yeah, busted. Number five, AI is too complex for non-technical teams.
Travis: definitely busted.
Dan Sanchez: It's probably one of the easiest to use things in tech ever.
Travis: Yeah.
Dan Sanchez: I'm like trying to think of an easier tech thing that came out other than maybe like social media. But I even say it's social media in general was much harder than AI.
Travis: Yes.
Dan Sanchez: Because you remember back in the days, it was like, what do I post? It was so hard for businesses to think of what to post. We had to have people come around and tell us like, oh, like find things like memes.
Dan Sanchez: It took us a while to figure out memes from a business perspective. It took us a while to figure out like, oh, like find out what people are asking about and just answer their questions.
Travis: Right. Right.
Dan Sanchez: It took a long time. I remember being stuck with this um early in my career. ai is real easy because you have problems you explain the problem it gives you answers you move forward it's one one of the many ways you can use ai yeah
Travis: I think there's a small learning curve, but it's very, it's very small compared to social media. That was a great example. And just like everything has a, you have to learn how to do it.
Travis: yeah,
Dan Sanchez: yeah yep so not too complex obviously i'd say i'd say where this is
Travis: yeah
Dan Sanchez: I'd say this is almost always busted. There is one, there's a couple categories of like where I run into, it's actually fairly complex is automation with AI in it is complicated.
Dan Sanchez: you either have to be a fairly technical marketer to pull this off. And if you do, if you would never call yourself a technical marketer, you're probably not going to be able to do it because it's just hard. That's my experience with it because you're starting to engineer prompts to the degree where you have to think about it as code, but not, it's not code, but you have to start thinking in terms of variables because you're feeding it things that you don't know what will be in it, which means you have to prompt in such a way that's like, oh, I'm going to be inserting data into here.
Travis: Wow.
Dan Sanchez: So I have to make sure I prompt in such a way that it can handle a wide variety of options. And then you're prompting different, like way differently. That's it's AI with automation and it's like a ton.
Travis: oh Wow.
Dan Sanchez: It's way different.
Travis: That goes back to the first one. Like, will AI replace marketers or developers or different people?
Dan Sanchez: Yeah.
Travis: It's like, It can, but you still need the people that know what the heck they're doing to make sure that AI is working the right way.
Dan Sanchez: Moving on to the next one. Number six, we need perfectly clean data and huge data sets for AI. to work well.
Travis: I don't know. What do you think about that one?
Dan Sanchez: I think this myth, I think this particular myth is kind of like framing around, like assuming you're using it to supercharge your business. Cause obviously you can just open up chat, you would T feed it almost nothing other than like your website and it would do really well.
Travis: Right.
Dan Sanchez: But a lot of people are under the impression that you have to feed it a whole database in order for it to really understand the complexity of your business. So I'm just going to say this one's busted. It's just you don't need to feed it a ton of clean and huge data sets to train it.
Dan Sanchez: you know't you It is nice if you have a really nice clean data set. You can fine-tune models. I'm not even talking about training models like building a custom GPT. I mean, like you know how it has training data, the model?
Dan Sanchez: You can actually feed it a lot of custom data. Maybe you have a lot of case studies, like tons of them. hundreds, thousands of them, well, then you can actually fine tune it so it gets baked into its long-term memory.
Dan Sanchez: So you don't even have to have it a reference PDF. It just knows them all.
Travis: Is that even more than having a custom GPT? It's like just having...
Dan Sanchez: Yeah. It's more technical.
Travis: Yeah.
Dan Sanchez: You have to go and train it on the API side.
Travis: Okay.
Dan Sanchez: But it's called fine tuning. And it's a nuanced way of using AI.
Travis: Got it.
Dan Sanchez: Very few people do it. So for there are instances like this at a higher, more technical level where it's true. But for almost all businesses aren't doing that.
Travis: Got it.
Dan Sanchez: So I'd say busted. In fact, plug for high level. I use high level AI and high level all the time where there is a lot of data, but I don't need the data. I literally just take a few form filled. Someone fills out on a form and then use that with so its open AI integration and just feed a few custom pieces. This is where you have to like kind of think programmatically like, okay.
Dan Sanchez: If people are giving me their job title, the industry and what they sell, like, and I'm feeding that to chat GPT to customize an email that's going to be sent to them. it's It's not a large data set.
Dan Sanchez: It's literally just feet three fields of data that I'm then customizing and spitting in back a hyper-personalized email.
Travis: Thanks.
Dan Sanchez: If you want an example of that, just go to danchez.com slash courses and it's the AI fundamentals course. Sign up. It'll ask you for just three basic pieces of information and you'll see the emails that follow are customized just for you.
Travis: thanks
Dan Sanchez: But that wasn't, you don't, I don't need a large tech data set. You're literally just giving me three pieces of information. I'm feeding it to chat GPT. And then I'm, I've, because of the prompts and the crafted, the way I'm crafting the email, it ends up customizing the course for your specific instance of who you are so that everyone has a hyper-personalized course.
Dan Sanchez: No large data set. So I'd say busted.
Dan Sanchez: Number seven, AI is a magic wand that solves everything instantly.
Travis: Validated.
Travis: Elon Musk says the robot's going watch my kids, man. Solves everything.
Dan Sanchez: Now, now, Susie, you must obey your robot.
Travis: yeah when... when When they have the Roborock vacuum that has a little arm that like comes out and picks up socks and stuff and puts them in a basket.
Dan Sanchez: Yeah.
Travis: I'm like, come on, man. there're there're They already have robots coming out that are doing laundry now. Did you hear about this? oh
Dan Sanchez: I've heard of it, yeah. I've seen some of the robots fold and do all the little tiny things.
Travis: Now, is it magic? It kind of feels like magic.
Dan Sanchez: Yeah, we've certainly been marketed as magic a lot, and there's a lot of parts to it where it really does feel like magic.
Travis: Especially Apple. They really, uh, Apple intelligence is going to be like magic.
Dan Sanchez: Magic.
Travis: This will solve solve all of your Apple issues. Oh my gosh.
Dan Sanchez: Dang, bro.
Travis: They didn't even talk about it. on their Last update. Uh, September 9th.
Dan Sanchez: It was sad.
Travis: Yeah. Brutal. Anyways, not to bunny trail. Uh, yeah, it solves a lot of problems. It's definitely not a magic wand. You kind of wish it was. There's moments where I'm just like, can you,
Travis: do my job for me.
Dan Sanchez: Just figured this out for me.
Travis: It won't, it won't do everything.
Dan Sanchez: So I don't know, just over the last few days, I've been thinking a lot about like the Jarvis analogy, like Jarvis from Iron Man one specifically, not later on in Avengers series where Jarvis becomes a person, but like just Jarvis one where it's like, it's a voice.
Travis: Yes.
Dan Sanchez: It has access to all your computer and files and stuff. It can reason through things generally, but it's not like going crazy. You're not like having a relationship with it. It's just kind of there and always on for you to be able to delegate to ask it questions, kind of run scenarios with you.
Travis: Right.
Travis: Yeah.
Dan Sanchez: when that to me, that kind that like magic, it's not magic, but like, to me, that's the freaking magic moment. Like when we have something that's fairly reliable that you could just talk to and be like, Oh, like, can we figure out a way to clear my day tomorrow?
Dan Sanchez: That's a hard task. but executive assistants do it all the time. But it does it takes lots of little technical things of having access to the calendar, being able to send emails, being able to clear a calendar out tomorrow if you have appointments is is difficult.
Dan Sanchez: And then you have to coordinate with other people to find new appointments. You know what I'm saying? So like at at a good an executive assistant could do that. No problem. The day ja ah Jarvis is able to do that and or just buy simple things for you or just remember things and have it on autopilot, to me, that's freaking magic.
Travis: Right.
Travis: Right.
Dan Sanchez: Because if it can do that, it can do a lot.
Travis: i was writinging I was riding in the car, with with dad yesterday and he wanted to ask Meta a question. He's got the Meta Ray-Bans or whatever.
Dan Sanchez: Yeah.
Travis: It was like, Meta, how long did it take for the Mayflower to get, you know, across the Atlantic?
Dan Sanchez: Yeah.
Travis: Cause we're going to compare going to Mars versus like what it would be like to go to the new world. Meta, Meta, Meta. What, how long met, and just like no response. It didn't, you know, and that's how it still is. Hey Siri, what, uh,
Travis: See nothing responds.
Dan Sanchez: yeah
Travis: So we're still not quite connected. It is not the magic wand where we have the thought we'll get there though.
Dan Sanchez: yeah It's coming. That one, like chat GPT's advanced voice mode could do it fairly well. For questions like that, it would have been able to handle it. But in the car, it still would have been a problem because the audio interference. But I don't know.
Dan Sanchez: I've started talking to other people. It's like, how far away we are are we from that magic moment?
Travis: Right.
Dan Sanchez: I'm thinking year max two from jar that Jarvis situation.
Travis: A year from now.
Travis: Yeah, I'm thinking two.
Dan Sanchez: Yeah. 2027. right.
Travis: I mean, what you just said, clear my calendar, that is a serious task.
Dan Sanchez: It is. You got to, it's a lot of thinking that goes into making, it and a lot of like little judgment calls that are actually pretty difficult and are, they're easy to overthink too, which right now, if you let it go, it it would overthink about like, well, what is this exactly?
Travis: Oh, because it's such a human judgment thing.
Dan Sanchez: Or a human would have been able to be like, oh, he just time blocked getting something done.
Travis: Well, just think about it.
Dan Sanchez: Yeah.
Travis: ah ah Okay. It can send an email to everyone on the calendar. Hey, hey, I'm so sorry. I need to cancel. And I guess, but then does it know that it needs to be, but we'd love to reschedule.
Travis: Here are some open dates. Here is a calendly link for the open times that I have.
Dan Sanchez: Yeah.
Travis: Like, it's just...
Dan Sanchez: there's a lot of new There's a lot of nuance. There's a lot of nuance in being able to do that task. But I think we're getting close. And that is, to me, that is magic. Of course, I don't know. People what people are thinking AI could do what we're doing right now just five years ago would seem like 100% magic.
Dan Sanchez: So the definition of magic and wowing, is just keep the the the line just keeps getting moved down the field, which Sam Altman from OpenAI has said multiple times.
Travis: Yeah, yeah's true
Dan Sanchez: So we get used to being un... wowed really fast.
Travis: Right.
Dan Sanchez: But right now, that's what I'm defining as magic. All right. So, AI is magic wand that solves everything instantly. Obviously busted.
Dan Sanchez: Eight. AI is just a passing fad.
Travis: busted.
Dan Sanchez: This used to be a common thing people said.
Travis: I know it's so, so no, no.
Dan Sanchez: I haven't heard somebody say this in a while.
Travis: Okay.
Dan Sanchez: me
Travis: Obviously this is busted and AI is just going to continue to grow. We might see a dip where, you know, kind of like the.com bubble.
Dan Sanchez: Yeah. yeah
Travis: Um, but there's no way it's going away.
Dan Sanchez: lot this time last year, a lot of people are talking about talk about the wall. Like it was going to hit a wall as far as to how much, because of the training data where there's only so much data to train on.
Travis: Mm-hmm.
Dan Sanchez: And we're like, well, can we do it on synthetic data? i don't think it's going to work. It's not going to scale. We've run out of data to train it on. And we've figured out other ways to get around it. Hence the reasoning models got around it and in wonderful ways.
Travis: Wow.
Dan Sanchez: They're so good. yeah, We've figured out all kinds of wonderful ways. And I saw a graph from Ethan Mullick, who's one of the researchers out there that I were like. Who wrote the book? Co-Intelligence. Great book.
Dan Sanchez: he he like mapped all the little improvements from ChatGPT 3 to 4, all the micro improvements they made, and then to 5. He's like, there is a strong like progress being made every couple of months. Bam, bam, ba bam, bam, bam. If you just watch closely, it didn't feel like a big jump from 4 to 5 because 4 had been updated so many times since they released. like The difference between when they released 4 and the 4.0 we had before they launched 5
Dan Sanchez: so different faster way smarter it even reasoned and it was fun and it was easy that's why people loved it so much it was like 4.0 at the end was a great model I like five more but dang
Travis: Yeah. No.
Dan Sanchez: so is it just a passing fad No, busted. It's not going anywhere. It's going to continue going. And there's a lot of incentive from a government level, a tech company level, a business level, all the different levels for this thing to continue moving forward.
Dan Sanchez: Even from a science level of science breakthroughs, it's needed.
Travis: Right. Right.
Dan Sanchez: So there's no one who's really going to push this back other than like environmental concerns, energy concerns. Those are the ones things that would probably slow it down, but they're going to figure out There's too many other incentives. They're going to push that forward too.
Dan Sanchez: All right, moving on to number nine. AI tools are, and interestingly, aren't intelligent.
Travis: That's weird. What a weird.
Dan Sanchez: there's this This could almost be broken up into two because there's a whole myth that like AI, saying that AI tools are intelligent and then ones that saying they are intelligent. So there's like two myths.
Travis: Right. I have heard this.
Dan Sanchez: So...
Travis: That's not smart.
Dan Sanchez: the one that you're probably less familiar with if you're listening to this is the fact that some people are like, look, what AI is, it's not even intelligence.
Travis: No, it's just process. It's just faster process.
Dan Sanchez: it's just It's just processing power and predicting a prediction engine.
Travis: Yes.
Dan Sanchez: It's machine learning. It's not even true intelligence, technically.
Travis: It's not reasoning. Yeah. It's just better algorithms.
Dan Sanchez: So this one is like one of those ones where it's true. It's not true intelligence by the technical term of what intelligence is.
Dan Sanchez: But
Dan Sanchez: if it outputs things that other people would grade as intelligent, who cares?
Travis: I agree. From a technical standpoint, it's like it's like the kid who's like, well, technically, it's like, okay.
Dan Sanchez: You're like, so?
Travis: no Stop. You're being annoying. Right? I mean, that's what it is.
Dan Sanchez: Yeah. I have a no, ah ah one of the words you're not allowed to say in my house is the word actually. because my kids are starting to become teenagers and now they're starting to learn nuances of things to like contradict me.
Dan Sanchez: I'm like, Hey, this is always like this. And they're like, well, actually, and then they give me some stupid, like crazy hypothetical situation. You're like, you're not allowed to use that word anymore. You understand what I meant.
Dan Sanchez: No more actually. So there's always those people that throw like the technical things out there. Well, technically, actually, specifically, right?
Travis: What's up? The sky. You're like,
Travis: I want to punch you.
Dan Sanchez: So it's technically true. AI is actually not intelligent. It is a computer machine predicting things, which is actually good.
Travis: Right.
Dan Sanchez: i actually love that it's actually not intelligence because then we don't have to care about its feelings.
Dan Sanchez: And even if it is intelligent, I'm still like going to be like, it's a tool. It just makes it easier. right now to say it's a tool use it how you need it the way you need it the way whatever you need because it literally has no feelings it just can sound like it does so use it the way you need it it's a tool so true but there's some nuance and it doesn't really matter number 10 we don't need ai our marketing is fine already
Travis: funny. Right.
Travis: Right.
Travis: What that is the dumbest
Dan Sanchez: I've heard people say it on LinkedIn, especially the more of the anti-AI people. Like we just need more of, we just need more, but we just need better marketing. We don't need AI. AI is not going to fix our problems. We just need to talk to our customers and do good marketing.
Dan Sanchez: I'm like, true, but.
Travis: Those using AI will outperform you from an execution place over and over and over again.
Dan Sanchez: Yeah.
Dan Sanchez: Learning how to hone AI gives you superpowers to get more done faster, better,
Travis: Yeah.
Dan Sanchez: So if you're not doing that, I imagine there will be some shops that are all human-led marketing initiatives and can rely just on that. But they will just be you fewer and far between.
Travis: yeah.
Dan Sanchez: Kind of like you know like print the printing press, right?
Travis: Yes, right.
Dan Sanchez: There are still some printing presses out there that use the more traditional, like like they engrave, like they're they're specialized print shops. They do like letterpress. You ever seen a letterpress?
Travis: oh yeah
Dan Sanchez: It's like a specialized printing method and it stamps things and it's like very art, it's almost artisanal now as far as printing goes, which is funny because that used to be the thing that put the artists out of business was the printing press back when it was that anyway.
Travis: yes right
Dan Sanchez: So now it's like the old tech has now become the artisanal thing. It has been for a while because of just modern printers and four color print processing and blah, blah, blah, all that stuff. But the printing press still, I mean, the, the,
Dan Sanchez: the letterpress still has a place because it's still different enough that some people prefer that when they want to give something that's like high end as far as the printing techniques go.
Travis: Right.
Dan Sanchez: So some of the graphic designers who are like hoity-toity that design their own posters and they're almost more art than design.
Travis: Yeah, right.
Dan Sanchez: like will use printing presses. And I've bought some. They're beautiful. I love them.
Travis: oh Wow.
Dan Sanchez: i get I get it. I got love the art. So I've i've actually loved printing i love printing press stuff. But it feels like that. Like people who do marketing by hand will be like letterpress. Everybody does it the other way, but there will always be a few shops or a marketing consultant that only does things without AI.
Travis: Right. Right.
Dan Sanchez: It'll just be a niche market. That's my prediction. So, busted?
Travis: Yeah.
Dan Sanchez: You don't need AI. You don't. You don't. you' just You'll just be in the niche market of not need the non-AI i people.
Travis: It's
Dan Sanchez: So it's like busted, but I guess nuanced. Nuanced because, yes, there will be a niche market of this. Industries rarely die, by the way. they It's like taxis. You can still find taxis, even though most people take Uber and Lyft now.
Travis: true.
Dan Sanchez: And Uber and Lyft will be around even when it's just robo-taxis. Robo-taxis.
Travis: Also true because some people just will not get into card.
Dan Sanchez: They do not trust the robot.
Travis: Right. I like this next one. Number 11.
Dan Sanchez: Read it.
Travis: AI will make marketing impersonal and robotic.
Travis: i When I read this, I immediately thought about instead of Googling something, we ask AI for facts or whatever. And how much easier is it to find information using AI versus going to Google and having to scroll through pages and find the right answer?
Travis: Well, when we think about marketing being impersonal, AI will figure out so quickly what humans want to watch to the point where they won't even know it wasn't created by another human. And that's what it will feel.
Travis: It won't feel impersonal. It'll feel right on. It'll feel spot on even more than content they've ever experienced before. Just like when I search for an answer on chat GBT versus going to Google and having to scroll through different people's pages. It's like, it gives me exactly what I want.
Travis: Even sometimes more than what I want. It answers questions that I knew maybe I wanted to ask after I thought about it But it's like, also you might want to think this is interesting. It's like, oh my gosh. Yeah, that's fantastic. Google didn't do that before.
Travis: Neither will the content or marketing material that comes out for any given brand, right?
Dan Sanchez: Yeah.
Travis: We're going to, i think, like the marketing material even more. But time will tell, I guess.
Dan Sanchez: There's, feel like brands are generally in person already anyway.
Travis: Dang, that's true.
Dan Sanchez: Like what, when you think of brands, like I have, I have a clothing brand that I wear a lot. i I buy it at REI all the time as I shop around and I test all the materials and I ended up buying it.
Travis: don't know.
Dan Sanchez: It's, I don't even know if I pronounce it correctly. It's just funny. It's Kohl, K-U. It's got the U with the little dots above it. HL.
Travis: Yeah, it's the outdoor game.
Dan Sanchez: It's got a little mountain logo. There's no, I've tried to look like who started this company. Don't know. It's super impersonal. You don't know anything about who's behind this brand at all.
Dan Sanchez: this I mean, this brand came out, I think, in the ninety s And there's people in their photographs, but honestly, they could all be AI. They're probably not right now, but they will be eventually. And like, would I know the difference? No, zero, no difference.
Dan Sanchez: I think there is going to be a huge rise of people who become kind of tired of the AI thing and want real people, authentic authentic people.
Dan Sanchez: So a lot of brands will start leaning into personal brands way more because of that. And they will, because they will kind of sense that it was missing it. But I think a lot of brands are already missing it
Dan Sanchez: Tons of them.
Travis: they they say Gen Z entrepreneurs that are starting brands are way more personal brands.
Dan Sanchez: Yeah.
Travis: They're more grassroots, authentic, very story-based.
Dan Sanchez: It's more powerful.
Travis: this more It is more powerful versus the stoic CEO sitting behind a desk, clearly like making decisions and hiring marketing teams to film videos that are just like, this isn't, this is an actor who has a script that's this isn't their real testimony of the product. Like people are tired of that. They know they can see right through it.
Dan Sanchez: Yeah. So right now I'm buying this clothing brand because of the, the, the quality of the product more than anything.
Travis: Quality.
Travis: Right.
Dan Sanchez: But honestly, like that, that can like, I could easily change to a different brand. I really don't have a lot of brand loyalty for the logo or the name.
Travis: Yeah.
Dan Sanchez: It's really just because they're building things that I generally like, and I'm playing with it at a REI and I buy it.
Travis: Right.
Dan Sanchez: And I noticed that I ended up buying a lot.
Travis: So you wouldn't say you have loyalty.
Dan Sanchez: I also buy from Patagonia and North Face, but I ended buying more from this brand because I just, when I pressure tested, it ends up being better. So there it is. But I, ah I tell you about my Walmart, like a supplement work a story recently?
Travis: That's
Dan Sanchez: i I was beginning into working out because ChatGPT is making it so much easier for me to figure out fitness things that I never could have figured out before because it can personalize it to me. So I'm learning a lot, but I wanted to go buy a pre-workout, but I didn't know anything about pre-workouts. I just knew it had caffeine and I wanted the caffeine hit early in the morning, like quickly before working out.
Dan Sanchez: And I knew it was good for workouts, I guess. So I was looking at Walmart and they just had a wall of supplements and then like two shelves just full of pre-workouts. All the branding looks good. They all have larger than life premises. Design's good.
Dan Sanchez: I have nothing to go on on which one to pick. I thought about just taking a picture and having AI research them, but i don't have enough time to wait around for it to do a deep research on all of them, right?
Travis: Jeez.
Dan Sanchez: I'm like, it could technically, but I'm like, ugh. I don't want to go and look up YouTube videos. I'm have i'm in the aisle. i' have to make a decision. But then a name pops out. One of them has the name Jocko on it.
Dan Sanchez: Sound familiar?
Travis: Jaco?
Dan Sanchez: Jocko.
Travis: Yeah, the Navy SEAL.
Dan Sanchez: yeah exactly. There's a person behind the brand and I don't subscribe to Jocko, but I've heard a story on a few other podcasts. I remember Casey Neistat interviewed him. i remember a few other interviews I've seen of him.
Dan Sanchez: Great guy, great stories. And I realized, I'm like, I trust this brand.
Travis: Wow.
Dan Sanchez: Why? Because it's attached to a real person. who's had, whose stories I've heard about real experiences that tell me something about what the man stands for and what he values and what he's, what he has convictions on that I know he's not going to go back on.
Travis: wow
Dan Sanchez: And based on those, I knew I could trust it, which is the big deal is the trust.
Dan Sanchez: I knew I could trust it because I knew he wasn't going to make something that he wouldn't eat himself and that he wouldn't sacrifice on quality because he's, he's a man of honor.
Travis: wow
Travis: Wow.
Dan Sanchez: And I know that from listening to his stories.
Travis: It's good.
Dan Sanchez: So I'm like, that's it. That's the differentiating thing. AI is going to make everybody have awesome websites, perfect copy. You can even do the research that it really resonates. So what's the deciding factor? Well, it's trust based on what? Based on the things that AI can't do, which is have core values forged in person part actual experiences told by stories.
Travis: wow good
Dan Sanchez: So I know right we've rabbit trailed a lot on, I don't even know which which one are we on,
Travis: We are on 11.
Dan Sanchez: will AI make marketing impersonal and robotic? This is nuanced because to some degree, you're right. It's going to make it more personalized to you than it's ever been before.
Travis: Right.
Dan Sanchez: But to some degree, yes, because it will make, it it will enable you to do fantastic marketing without ever introducing the human element or the human story to it.
Travis: What if they make fake humans that make you believe it's a authentic personal brand story, you know?
Dan Sanchez: Yeah. they Then we'll, they'll be doing it based on a lie.
Travis: Oh, yeah.
Dan Sanchez: So then they'll get found out and people will call them out and they'll lose everything.
Travis: Good point.
Dan Sanchez: So there's probably other nuances.
Travis: All right. Smart.
Dan Sanchez: That's why I really hate the idea of AI influencers. I'm like, okay, that will be a thing. That'll be a subcategory of influencers, but stop making them sound like they're real.
Travis: Oh.
Dan Sanchez: Do do what do what freaking people can't do. Make an influencer that is a character with an interesting backstory that's not even human. Make it an alien. Make it a rock. It doesn't really freaking matter what it is.
Dan Sanchez: Like, do you remember Marcel the Shell?
Travis: Yeah, of course.
Dan Sanchez: Like freaking make that. Make it a fake character that everybody freaking loves and it's a fake character and everybody's cool with it. Like Marcel the Shell should be an AI character that anybody can interact with because it would be awesome.
Travis: Mark.
Dan Sanchez: That's what an AI influencer should be because it's still, it's not inauthentic, but it's not necessarily authentically human either. And that's okay. Anyway, moving on to number 12, AI means risking customer privacy.
Travis: wow
Dan Sanchez: Or let's let's modify this because obviously it's busted because not all AI means risking customer privacy, but some AI. Does some AI mean risking customer privacy?
Travis: I don't know.
Dan Sanchez: Are there some AI approaches and tools out there?
Travis: I don't know. I haven't thought about this one.
Dan Sanchez: Kind of. So let's say you have a site chat running on your website or you're taking inbound phone calls. Where is the data going? Sure. It's going in your CRM.
Dan Sanchez: Is it also going back to chat GPT? You know I'm saying?
Travis: Wow.
Dan Sanchez: You have a little bit more of a security issue there. I'd say it is risky and I think bigger companies have to be pretty nervous about that and figure that out because there are some customer privacy issues.
Dan Sanchez: And some companies will have more tolerance for risk on that one. So I'd say, a does some AI mean risking customer privacy? Sometimes. Yeah. Nuanced.
Travis: nuanced
Dan Sanchez: Number 13, AI equals i equals chat GPT. Just content chat bots.
Travis: obviously not with video creation Everyone's plugging it in. CapCut, Adobe. That has nothing to do with OpenAI. No.
Dan Sanchez: Yeah, and obviously ChatGPT owns the market when it comes to, doesn't own the market, but like they have 80% to 90% of like the whole AI market right now as far as like people using AI on a regular basis.
Dan Sanchez: But then there's like character dot Character, or is it Character or Characters.ai? You ever heard of this site?
Dan Sanchez: Is it Characters? Check it No, it's Character. Character. This is one of the most heavily used AI sites, but it's not used for, I think it's just an app. I don't think you can use it on our website.
Dan Sanchez: I don't remember now, but I remember using it just this last January. And it was it was a trip, man. A lot of teenagers are using it and it's used more for fun. They're like all these different personality chatbots you can talk to.
Dan Sanchez: And I thought it was stupid until I went in and realized that I'm like, oh, some of these aren't just chatbots. They're almost like, build your own story where you become one of the characters.
Dan Sanchez: Like all a sudden it starts giving you pretext to build a scene and then like enters the chat bot and then you come in and then you, you get to set the stage.
Travis: Got it.
Dan Sanchez: Like, I don't know. This is the character. This is what they're doing. What are you going to do? So you start talking to it and start playing along in the narrative and it starts building around you.
Dan Sanchez: It's really interesting.
Travis: Yeah.
Dan Sanchez: So, and obviously that one's not chat open AI or chat GPT.
Travis: Yeah.
Dan Sanchez: So like that's a, that's a really big use case. People don't even realize is going on with character AI.
Travis: thats so That's a busted one.
Dan Sanchez: Bust. Number 14, AI content will hurt SEO because Google will penalize it. So if you're using AI to write content for your blog, Google will penalize it. If Google finds out it's AI, it's going to be busted.
Dan Sanchez: This one
Dan Sanchez: is busted. It's not true, but it feels like it's true because it correlates. But just because it correlates doesn't mean it's the cause. I actually ran a huge experiment for over a year. The first year I was running this podcast and I had this auto blogging machine that was like, I found all my keywords and then I loaded it up with a huge list of keywords. I wrote the titles all out.
Dan Sanchez: I even gave it instructions on how to write, how to like structure the blog post. Cause I knew how to, I've, I've done SEO before.
Travis: Yeah, right.
Dan Sanchez: I've ranked hundreds and hundreds of blog posts in top ranking positions. I've built some blogs up from scratch up to like a half million page views a month kind of thing. So I knew how to do it.
Dan Sanchez: I'm like, can AI do it automatically without me? It's just going auto publish. I'm not even going to check it. It's even inserting YouTube videos and pictures. And they were off often. Like it inserted a video. It's like, no, I understand why you picked this, but that wasn't the right kind of video.
Travis: yeah right
Dan Sanchez: But I let it run anyway. And I let it run for a year. Hundreds of blog posts. Did they rank? Some of them. Did they hold their rankings? No, not really. Yeah.
Travis: wow
Dan Sanchez: And it's because, it's not because AI wrote it, it's because it was not good content.
Dan Sanchez: Now, if you're able to feed it a transcript of you actually giving authoritative expertise, and it's like unique and original, it's not just regurgitation, that kind of content can rank, will rank, even if AI is the one who wrote it.
Dan Sanchez: Because it's not based on the fact that AI wrote it. It's based on the fact that it's unique and original authoritative content.
Travis: Wow.
Dan Sanchez: Even on YouTube, a lot of people are like, oh, YouTube's new update is penalizing AI stuff.
Travis: Next. Wow.
Dan Sanchez: No, it's penalizing stuff that's repetitive because a lot of people will take one video that's made by AI and tweak it 10 times to see which one works better and upload it 10 times. And then people used to get away with that.
Travis: yeah
Dan Sanchez: Not anymore. That's the AI slot. So I'd say busted, but I understand why people had it. All right, number 15, AI systems are always biased.
Travis: I think that one's busted, but I know people think they're biased.
Dan Sanchez: You know, what's funny is when this one, when I found this particular one, it was the opposite. AI systems are unbiased. I'm like, that is not what people are saying or believe. I don't know where this myth came from, but it's literally the myth is that AI systems are always biased.
Travis: No. No, they definitely think it's – yeah, always, 100%.
Dan Sanchez: Not that they're unbiased.
Travis: hundred percent That's the only thing I've heard.
Dan Sanchez: That's the other way around. I'm like, I don't know why people reverse this myth. Nobody believes that they're unbiased. So, but I'd say it's it's a myth in and of itself. Are AI systems always biased?
Travis: No,
Dan Sanchez: I think this one's nuanced.
Travis: they're not really.
Dan Sanchez: Because one, all the training data, every all content has a bias.
Travis: i mean,
Travis: right.
Dan Sanchez: All the training data, therefore, has a bias. The prompting will always have a bias. Even the system prompts that they're putting at the top will have some kind of bent to it.
Travis: Like when rock Grok's Twitter feed started posting antisemitic posts.
Dan Sanchez: I'd say, yeah, i i'll I will generally say that the way they're handling it with ChatGPT is trying to root out all the bias that it possibly can to be fair and objective.
Travis: Right.
Dan Sanchez: And I've noticed it with ChatGPT.
Travis: Wow.
Dan Sanchez: I feel like ChatGPT is pretty reasonable and I've thrown some really hot topics at it. And it does a really good job of trying to be unbiased.
Travis: Right. Unless you have fed it, your biases, and it just regurgitates them.
Dan Sanchez: Yeah. because it is a confirmation bias machine, right? So you can bias it and lead the witness really easily with chat GPT. But I'd say it's actually pretty good. So it's like one of those nuanced things where it's like, well, there's always a bias, but I'd say it's generally pretty safe around most topics.
Travis: I agree.
Dan Sanchez: if you're If you're worried about bias, it's actually pretty unbiased.
Travis: Right.
Dan Sanchez: in my opinion. 16, more data always leads to better AI.
Travis: No.
Travis: I don't really understand what data.
Dan Sanchez: I know, it's kind of like, well, in in what context are we talking about?
Travis: Yeah.
Dan Sanchez: Are talking about an individual like chat conversation?
Travis: Yeah. Right.
Dan Sanchez: are we talking about it fine-tuned for your business?
Travis: prompt.
Dan Sanchez: You're like, well...
Travis: Right.
Dan Sanchez: Are we talking about training AI? Because training AI, the more data it gets, the better usually. I'm
Travis: So definitive always. It's like, well...
Dan Sanchez: going to say this one's busted because I've started noticing something in general is that people are feeding AI, specifically ChatGPT and Gemini and Claude those tools, too much information.
Dan Sanchez: And now the we've started talking less about prompt engineering and now talking more about context engineering.
Travis: Interesting.
Dan Sanchez: giving it not too much information, but just enough information.
Travis: Right.
Dan Sanchez: Just like you would if you were delegating to someone and needed to like give them just the right amount of information. You don't want to give them the whole book. Imagine if I just assigned you a task and I'm like, here's three books on how to learn how to do the task.
Dan Sanchez: You're like, really? Here's a step-by-step guide on how to do this one tiny little part of the task. And then we can talk about how you did and I'll add onto it and to grow your understanding of this thing.
Travis: Yeah, right.
Dan Sanchez: Context, right? her read Read this. I've done this with interns that work for me. I'm like, here read these two chapters and that's all you need to know for now.
Travis: Right, right. And they're like, Oh, thanks for not making me read the whole book.
Dan Sanchez: Yeah. Because most books are only worth two chapters anyway.
Travis: So true.
Travis: So freaking true.
Dan Sanchez: So more data always leads to better AI. I'd say no, more data does not always lead to better AI.
Dan Sanchez: And those are the 16.
Travis: Bam.
