Latest / Sleeping Barber - A Marketing Podcast / SBP 229: The PostPod - Lessons from Pats McDonald. AI Won't Decide the Future of Marketing. We Will.

Transcript

SBP 229: The PostPod - Lessons from Pats McDonald. AI Won't Decide the Future of Marketing. We Will. · Sleeping Barber - A Marketing Podcast

• 0:00
  SPEAKER_00: Welcome to the Sleeping Barber Podcast.
• 0:03
  SPEAKER_01: We have to stop blaming AI for things that haven't really happened yet.
• 0:06
  SPEAKER_01: AI is gonna take our jobs.
• 0:07
  SPEAKER_01: AI is gonna commoditize creativity.
• 0:09
  SPEAKER_01: AI is going to make strategy worse.
• 0:12
  SPEAKER_01: But AI isn't sitting in the boardroom making any of those decisions.
• 0:16
  SPEAKER_01: We are.
• 0:17
  SPEAKER_00: Welcome to the Sleeping Barber Podcast, a place for business leaders to get the best and most credible information on marketing, strategy, and innovation.
• 0:26
  SPEAKER_00: Your hosts, Mark Binkley and Vasili Storos, share their experiences as they gather insights on the world's leading experts.
• 0:34
  SPEAKER_00: Now, on with the job.
• 0:36
  SPEAKER_01: This is weird.
• 0:39
  SPEAKER_01: We're in the same city.
• 0:41
  SPEAKER_01: At least.
• 0:42
  SPEAKER_01: Finally.
• 0:43
  SPEAKER_01: Same city.
• 0:43
  SPEAKER_01: It's been a while.
• 0:45
  SPEAKER_01: It's been a minute.
• 0:46
  SPEAKER_02: This is the post pod.
• 0:49
  SPEAKER_02: Yeah.
• 0:50
  SPEAKER_02: Pat's McDonald, Chief Strategy Officer at Dent Suit, which we recorded in Cannes.
• 0:55
  SPEAKER_02: We did.
• 0:55
  SPEAKER_02: You just remind me, which is about 50 days ago.
• 0:58
  SPEAKER_02: Yeah.
• 1:01
  SPEAKER_02: Totally.
• 1:02
  SPEAKER_02: It was a great like a great time and great to have you back from Greece.
• 1:06
  SPEAKER_02: And uh you do.
• 1:07
  SPEAKER_02: And also a great conversation with Pat, I thought.
• 1:11
  SPEAKER_01: Yeah, you know what it really comes down to is I think we have to stop blaming AI for things that haven't really happened yet.
• 1:17
  SPEAKER_01: At least this is one of the big themes that I think I got out of that conversation with Pat's.
• 1:21
  SPEAKER_01: Like AI is gonna take our jobs, AI is gonna commoditize creativity, AI is going to make strategy worse.
• 1:28
  SPEAKER_01: But AI isn't sitting in the boardroom making any of those decisions.
• 1:32
  SPEAKER_01: We are.
• 1:32
  SPEAKER_01: People are.
• 1:33
  SPEAKER_01: Oh yeah.
• 1:34
  unknown: Yeah.
• 1:37
  SPEAKER_02: You know, I was just gonna say it's that she had used the term agency, which I kind of like I love the term, and I forgot the true meaning of it.
• 1:46
  SPEAKER_02: Like we have the choice available to us to make our own decisions.
• 1:52
  SPEAKER_02: Absolutely.
• 1:53
  SPEAKER_02: Not just the agency that we hire for media or creative.
• 2:00
  SPEAKER_02: Yeah.
• 2:02
  SPEAKER_01: Yeah, it's like we I know we've talked about this.
• 2:04
  SPEAKER_01: Our peers talk about this, the industry talks about this.
• 2:07
  SPEAKER_01: It's like, are we going to use AI and is it gonna eliminate those junior roles?
• 2:11
  SPEAKER_01: Or if you the flip side of that is are we gonna make junior people better instead?
• 2:16
  SPEAKER_01: Yeah.
• 2:17
  SPEAKER_01: Are we gonna use it to generate average work or use it to identify average work?
• 2:21
  SPEAKER_01: It's just these little things, like it's just a reframe of it.
• 2:24
  SPEAKER_01: Like it's very easy to look at like the negative, the the one that sells the clicks, the one that sells, you know, the uh I guess enlicits more of that urgency.
• 2:34
  SPEAKER_01: But you know, at the end of the day, those aren't really technology questions, those are they're really leadership questions that we need to be having and understanding inside our respective organizations.
• 2:45
  SPEAKER_01: So yeah, I don't know.
• 2:47
  SPEAKER_01: Is there one place I know you talked about agency kind of being, do you want to do you want to start there?
• 2:52
  SPEAKER_02: Yeah, sure.
• 2:53
  SPEAKER_02: I mean, I I think really well, you you brought it up a little bit earlier, like you know the idea that AI is gonna take all of our jobs, or some jobs more than others, and I do see that happening to some degree.
• 3:09
  SPEAKER_02: Like since the recording, there's been a lot so many changes in AI.
• 3:14
  SPEAKER_02: Like I use Claude the most, probably Gemini the second most, and then everything else is a distant third.
• 3:21
  SPEAKER_02: But the the things that you can do in Claude, there it seems to be updating almost daily, it feels like.
• 3:28
  SPEAKER_02: And so there are things I could do now that I couldn't do then.
• 3:33
  SPEAKER_02: And the and like today's version of Claude is gonna be the worst it's gonna be.
• 3:39
  SPEAKER_02: You know what I mean?
• 3:40
  SPEAKER_02: Going forward into the future.
• 3:42
  SPEAKER_02: So it's just gonna keep getting better.
• 3:44
  SPEAKER_02: I built a website page uh in Claude in between meetings and stuff one day before I went on vacation.
• 3:53
  SPEAKER_02: So I could see that becoming a thing where like from an agency, like hiring an outside agency to build websites might become a thing of the past.
• 4:06
  SPEAKER_02: Like it's harder, it's and it's a really simple website page.
• 4:10
  SPEAKER_02: It's just one page, it's not like an amazing thing.
• 4:13
  SPEAKER_02: It's you know what I mean?
• 4:14
  SPEAKER_02: So there's um but it will become easier and easier and easier.
• 4:18
  SPEAKER_02: The the friction between how websites used to get built and having to know all the things like with you know HTML coding and all that kind of thing.
• 4:28
  SPEAKER_02: That goes away.
• 4:30
  SPEAKER_02: Yes.
• 4:31
  SPEAKER_02: So I think where there were barriers before in in building a website as an example, it's not just websites, but building a website now becomes something that's that's gone that anybody can now do.
• 4:44
  SPEAKER_02: And so there's sure I could see there becoming an opportunity for new jobs though to replace the old jobs because there's still a lot of like people that don't get how to get into AI and use cloud code as an example.
• 4:59
  SPEAKER_01: Yeah.
• 4:59
  SPEAKER_02: So there's more of that kind of opportunity, and maybe less of the website building opportunities.
• 5:04
  SPEAKER_02: It's not necessarily a zero sum, like I win, that company wins and I lose.
• 5:11
  SPEAKER_02: Yeah.
• 5:12
  SPEAKER_02: Or vice versa.
• 5:13
  SPEAKER_01: Yeah, no, and I I think that's a very great uh it's a great example.
• 5:17
  SPEAKER_01: The the small nuance, so like you and I don't actually build websites.
• 5:21
  SPEAKER_01: So we can we have a limited framing on how we're gonna use, say, Claude Code or Chat GBT code or whatever, all these different things to help us build a website who where someone who specializes in building websites, they could probably their output would be a lot better than for us.
• 5:38
  SPEAKER_01: So there's still Totally.
• 5:39
  SPEAKER_01: There's still a role there for those that are specialists in those areas.
• 5:43
  SPEAKER_01: Now, the thing is the cost of entry or the barit entry has been essentially removed.
• 5:48
  SPEAKER_01: So in the same vein, we will have even more marketers.
• 5:53
  SPEAKER_01: You know, all they have to do is put a query in in Chat GPT or one of the LLMs, and they'll have a frame full-framed opinion on a topic as an example.
• 6:02
  SPEAKER_01: Yeah.
• 6:02
  SPEAKER_01: But everyone that's listened to this podcast, we could probably prompt those LLMs further to get a better, more thoughtful response to a topic that may be permeating in an organization.
• 6:13
  SPEAKER_01: So I think that's where specialization still matters, but the cost of entry or the barrier to entry right now is almost negated completely.
• 6:25
  SPEAKER_02: Yeah.
• 6:25
  SPEAKER_02: Yeah, totally.
• 6:27
  SPEAKER_02: Um what was one of the like big takeaways that you got from that conversation, or maybe something that you've been thinking about since then?
• 6:35
  SPEAKER_01: I think one one that kind of stuck out to me as well was she made a point around design for differences, not the sameness.
• 6:42
  SPEAKER_01: So because LLMs understand category conventions exceptionally well, uh one of their best uses may be identifying the average so humans can deliberately move away from it.
• 6:54
  SPEAKER_01: We've talked about this.
• 6:55
  SPEAKER_01: Like, you know, every the LLMs push you to that mean.
• 6:58
  SPEAKER_01: Totally.
• 6:58
  SPEAKER_02: Yeah, yeah.
• 6:59
  SPEAKER_01: So I think that's constantly, I think, playing in my in my in the back of my mind.
• 7:05
  SPEAKER_01: So and it really there is that criticism that, you know, AI creates the sameness.
• 7:11
  SPEAKER_01: Perhaps the problem isn't AI, it's how we're prompting and using it as well.
• 7:16
  SPEAKER_01: So like it's also a little bit on us too, as prompt engineers, if you want to call it that.
• 7:22
  SPEAKER_01: So, you know, maybe we have to be asking AI, you know, what would everyone else do before asking what we should do as an example?
• 7:29
  SPEAKER_01: It's a it's a small little thing, but for me, like because we're all learning as we're going around this, I think the way we were probably prompting to a year ago is a lot different than the way I know I'm I prompt a lot differently now than I did originally.
• 7:45
  SPEAKER_01: And it just depends where you are on that adoption curve.
• 7:48
  SPEAKER_01: So I think just being mindful of that as well, especially when you know everyone's quick to kind of criticize that AI is just pushing us to the mean.
• 7:55
  SPEAKER_01: It's probably because we're just asking for the mean.
• 7:58
  SPEAKER_02: Yeah, that's true.
• 7:58
  SPEAKER_02: And I I think that's it's a good point because it it is the mean for the most part, which is fine.
• 8:05
  SPEAKER_02: And yeah, if you know that, right?
• 8:07
  SPEAKER_02: Like rather than going for the absolute truth, just knowing that there's a whole range of truths out there, and it's gonna give you the the most popular answer, but you can it knows all the other stuff too, but it's stuff that's happened in the past almost always.
• 8:22
  SPEAKER_02: Yeah.
• 8:23
  SPEAKER_02: So it's not necessarily making net new ideas up, which is part of like I think where if you understand that you can go from across the spectrum, go from and ask it for opinions on unpopular to very popular, yeah, and then kind of everything in between.
• 8:42
  SPEAKER_02: There's something there like that that is rather than looking for absolute truth, you can you can scan it or use it to scan what's the right like the spectrum.
• 8:53
  SPEAKER_02: Yeah.
• 8:54
  SPEAKER_02: Right?
• 8:54
  SPEAKER_02: Like, and then then from there then you can get all kinds of different opinions.
• 8:58
  SPEAKER_02: And I think that's there's something helpful there for sure, to s help you see your own blind spots, but also help like anticipate how others have done things and benchmark against that.
• 9:09
  SPEAKER_02: So I don't want to think the sameness is necessarily a bad thing, just as long as you don't think the sameness is the reality of everybody.
• 9:19
  SPEAKER_01: I love I love what you just said there.
• 9:21
  SPEAKER_01: I couldn't agree more.
• 9:23
  SPEAKER_01: Well, thanks for and with that, no, I'm just kidding.
• 9:26
  SPEAKER_02: What else what else stood up for you?
• 9:27
  SPEAKER_02: Well, there's the and we've talked to reference this a few times.
• 9:30
  SPEAKER_02: I've referenced the strategist.
• 9:32
  SPEAKER_02: Strategish.
• 9:34
  SPEAKER_02: Wow.
• 9:34
  SPEAKER_02: I can't say strategy-ish.
• 9:37
  SPEAKER_02: Strategy-ish.
• 9:38
  SPEAKER_02: Oh, yeah, there it is.
• 9:39
  SPEAKER_02: Yeah.
• 9:40
  SPEAKER_02: Yeah.
• 9:40
  SPEAKER_02: And I've seen that more and more.
• 9:42
  SPEAKER_02: Like, this is one that I think is really challenging for all of us.
• 9:48
  SPEAKER_02: It it so the idea that the AI can do a lot of strategy is is true.
• 9:54
  SPEAKER_02: And I mean, really anything, it can do a lot of those things very quickly.
• 9:59
  SPEAKER_02: But if there's no friction, if there's not like choices that you're making, like it just turns this, like it becomes AI slop.
• 10:06
  SPEAKER_02: Yeah.
• 10:07
  SPEAKER_02: And so the answer becomes fast, and it sounds plausible, and so you just, you know, yeah, you agree with it, but it's there's nothing really substantial and tangential, and you may not have the challenge part where go where you go, Well, we're gonna make a left-hand turn instead of a right-hand turn because of this, this, and this.
• 10:29
  SPEAKER_02: I think there's the the thing that I find problematic where or where I have a lot of concerns, and I've seen it even in my own use.
• 10:38
  SPEAKER_02: As soon as timelines become a pressure point, it's easy just to go and say, Yeah, it's close enough.
• 10:47
  SPEAKER_01: It's close enough.
• 10:48
  SPEAKER_02: It's okay.
• 10:49
  SPEAKER_02: Yeah, it's okay.
• 10:50
  SPEAKER_02: I'll just do instead of like the real fundamental, like struggling with an answer.
• 10:57
  unknown: Totally.
• 10:58
  SPEAKER_02: Struggling with the question to find an answer and think of the best possible answer and look at this option and this option and this I think it can do that quickly.
• 11:08
  SPEAKER_02: But it's one of those ones where like I don't know, I I feel like the the time pressure to get things done faster becomes more more of a risk than ever.
• 11:19
  SPEAKER_02: That it you just pass off good enough for you know, good enough.
• 11:24
  SPEAKER_01: You're on fire today, man.
• 11:26
  SPEAKER_01: Yes, absolutely.
• 11:28
  SPEAKER_01: I've seen I'm sorry, and I'm gonna raise my hand here virtually.
• 11:32
  SPEAKER_01: I've I've done the same thing though.
• 11:33
  SPEAKER_01: You know, when time when time when you're time bound and you need something in the next five minutes, yeah, you're not using that critical thought process behind it where you're kind of digging in and saying, okay, how how does this track?
• 11:48
  SPEAKER_01: Well, actually, we could probably pressure this a little bit more in this point.
• 11:51
  SPEAKER_01: And I think that becomes, again, the beauty of AI, because it can get you something really quick that can pass, right?
• 11:58
  SPEAKER_01: Yeah.
• 11:58
  SPEAKER_01: At the same time, though, do we overrely on that and just you know what, it will always just be good enough.
• 12:04
  SPEAKER_01: And I can just, you know, continue to move forward in some of the other things that I'm working on.
• 12:08
  SPEAKER_01: Like that's a decision, that trade-off that you have to make individually for what you're working on, what you what you're trying to tell or support or whatever the case may be.
• 12:18
  SPEAKER_01: And I think, again, at the end of the day, that's our I don't know, dare I use like cross to bear.
• 12:25
  SPEAKER_01: Like inside us, we know exactly what we've done to get to that point.
• 12:29
  SPEAKER_01: So yeah.
• 12:31
  SPEAKER_02: I know, and I like I I think some of this is so much about just our own process.
• 12:37
  SPEAKER_02: I don't like I don't know why this is coming up in my head, but I have this quote from Gary V in my head that 95% of the shit you do doesn't matter.
• 12:47
  SPEAKER_02: But if that's the case, then it's kind of like, okay, well, what 5% am I gonna give a shit about?
• 12:53
  SPEAKER_02: Yeah.
• 12:54
  SPEAKER_02: Right?
• 12:54
  SPEAKER_02: And so then I can use the 95% like AI for 95% of the shit that nobody gives a shit about.
• 13:00
  SPEAKER_02: I don't know.
• 13:01
  SPEAKER_02: It just it's it's a bit of a it's a bit of a wormhole or rabbit hole because he can I just see it happening where the accuracy and the slop can become so invasive in just everything that's because not everybody's gonna read all the stuff anyway, or or you know, put the thought like you know, any into into critiquing anything that we everything that we do.
• 13:30
  SPEAKER_02: And so then, well, what are the parts that really matter and what are the parts that am I okay with just having less quality and and but it's tricky.
• 13:40
  SPEAKER_02: Like I think that will become important going forward in AI.
• 13:44
  SPEAKER_02: Yeah that the quality and credibility part of how we use it is gonna become more important than ever before.
• 13:52
  SPEAKER_01: Totally.
• 13:56
  SPEAKER_01: The I wonder if there's just kind of going off of like your strategy-ish the role of strategists of the future.
• 14:07
  SPEAKER_01: Pat's talked about like it becomes a little bit broader.
• 14:10
  SPEAKER_01: So she's mentioned, you know, strategically increasing, increasingly requires customer understanding plus algorithms, creators, community, comms, and channels while still knowing when specialists are required.
• 14:21
  SPEAKER_01: And this challenges the extreme specialization of modern marketing organizations.
• 14:26
  SPEAKER_01: And again, this is another thing that we've talked about.
• 14:28
  SPEAKER_01: Like, think about a lot of the in-house teams that have really specific roles for specific people.
• 14:35
  SPEAKER_01: Is that even required anymore?
• 14:37
  SPEAKER_01: Do we move to more of a generalist style operating model that, you know, because of AI can help adoption, you know, seeing things a little bit more holistically.
• 14:49
  SPEAKER_01: Yeah.
• 14:50
  SPEAKER_01: Is it is the specialization even a thing that we or people should be focusing on?
• 14:55
  SPEAKER_01: I don't know.
• 14:58
  SPEAKER_02: Well, I've like I've written a lot about this and thought about it with that pie-shaped article.
• 15:04
  SPEAKER_02: Yeah.
• 15:05
  SPEAKER_02: So the the concept is that like rather than a T-shaped person, you need a pie-shaped person.
• 15:10
  SPEAKER_02: And I'm kind of rethinking that in a way, like you almost need to cross hatch both ways.
• 15:15
  SPEAKER_02: But the pi meaning pie like 3.14, the symbol for that.
• 15:19
  SPEAKER_02: So the across the top, broad but narrow depth of understanding across across a broad range of topics, and then deep understanding in in an area of specialization, but then also a paired deep area of specialization in in how to use AI.
• 15:36
  SPEAKER_02: And I think some of the skills there on the AI part are really around critical thinking components and like the taste making kind of idea.
• 15:47
  SPEAKER_02: If if you're letting AI use become the thing that is the both broad breadth and depth, where you don't have depth, I think you're running into trouble.
• 16:00
  SPEAKER_01: Yeah.
• 16:01
  SPEAKER_02: Because it it will start giving you answers that you don't have the the like the own knowledge, your own knowledge of to critique whether or not it's a good answer.
• 16:12
  SPEAKER_01: So then it becomes you don't have questions for the answers it's giving you.
• 16:16
  SPEAKER_02: Yeah.
• 16:18
  SPEAKER_02: Yeah.
• 16:18
  SPEAKER_02: And so like recently, the just actually actually this weekend, we were all we had an issue with a boat that we were on trying to def like diagnose what the problem was.
• 16:29
  SPEAKER_02: AI was giving us all kinds of answers, and none of us are experts in this like field.
• 16:35
  SPEAKER_02: And as it turns out, like the first few answers, you know, would always say things like, Whoa, that's a great question.
• 16:42
  SPEAKER_02: Of course I should have thought of that.
• 16:43
  SPEAKER_02: You know, those kinds of things.
• 16:45
  SPEAKER_02: But it gives you answers very confidently and precisely around what the exact questions are, and or what the exact answers are.
• 16:54
  SPEAKER_02: And then you ask it a slightly different way or ask it to check on this, and it's like, oh, I should have thought of that.
• 17:00
  SPEAKER_02: Well, in that case, this is the right, and so it just constantly needs that kind of correction.
• 17:07
  SPEAKER_01: Yeah.
• 17:08
  SPEAKER_01: It's funny because that's such a great example.
• 17:11
  SPEAKER_01: I'm helping one of my buddies.
• 17:13
  SPEAKER_01: Uh, he's built an app.
• 17:14
  SPEAKER_01: It's called Running Gram.
• 17:15
  SPEAKER_01: Shout out to Alex behind this, and we're talking about it, and I had this idea in my head.
• 17:22
  SPEAKER_01: I'm like, just give me some time.
• 17:23
  SPEAKER_01: And I literally took like 10-15 minutes, and I because he wants to redesign, he wants to kind of, you know, do maybe even rebrand the the actual app itself, etc.
• 17:34
  SPEAKER_01: And uh I I just spent 20 minutes to 30 minutes with a with an LLM, and I got all the ideas out of my head and presented it to him.
• 17:44
  SPEAKER_01: And it's like, here's here's kind of what this is not final, you don't look at this as final.
• 17:48
  SPEAKER_01: Like, but to give you a conceptualization of what's in my head right now, this is what I think you need to do.
• 17:55
  SPEAKER_01: And for him, it was it's and it's not like he doesn't use AI, he never used it that way.
• 18:02
  SPEAKER_01: Right?
• 18:03
  SPEAKER_01: So this is an example where, again, we're two different people.
• 18:06
  SPEAKER_01: He's a programmer, I'm coming obviously from the commercial slash marketing side, and it's kind of opened up a whole new world.
• 18:13
  SPEAKER_01: And then he started having his own question.
• 18:14
  SPEAKER_01: I'm like, well, this is where you can play with it to get it to where you want potentially, right?
• 18:19
  SPEAKER_01: Because like at the end of the day, it's your decision.
• 18:21
  SPEAKER_01: I'm being like a person, like, you know, almost like two degrees removed from this, and I'm just I'm just giving you my input based on what I know.
• 18:30
  SPEAKER_01: And I just love that I was able to get to that conceptualization so quick.
• 18:36
  unknown: Right.
• 18:36
  SPEAKER_02: So that's the part you said that too before, where it like it removes the barriers.
• 18:40
  SPEAKER_02: And I think that's the part where it would like if we can't figure out how to make money off of this stuff and like make it productive for ourselves, where it's just always about fear.
• 18:52
  SPEAKER_02: I think the fear side is so limiting because it's so good at like making your ideas become a reality.
• 18:59
  SPEAKER_01: Totally.
• 19:00
  SPEAKER_02: Totally.
• 19:02
  unknown: Yeah.
• 19:03
  SPEAKER_02: With the gastrics of like, yeah, it still can be really shit, like the inshidification of everything.
• 19:11
  SPEAKER_01: Totally.
• 19:11
  SPEAKER_01: But imagine like it like because we had some delivers like some UI examples and things like that.
• 19:17
  SPEAKER_01: Well, you just take that to a UI engineer and say, listen, I want something like this.
• 19:22
  SPEAKER_01: Totally.
• 19:23
  SPEAKER_01: Yeah, yeah.
• 19:24
  SPEAKER_01: Yeah.
• 19:24
  SPEAKER_01: That that makes then the guy that's working on the guy or girl that's working on, you know, uh the UX side, it gives them a point of reference.
• 19:33
  SPEAKER_01: Of course, they're gonna build it.
• 19:35
  SPEAKER_01: Now he won't build it.
• 19:36
  SPEAKER_01: Yeah.
• 19:38
  SPEAKER_01: Or when you're thinking about a new logo, hey, here's here's like a new logo idea, then give it to someone that actually professionally does logos to bring it to life for you.
• 19:46
  SPEAKER_01: Totally.
• 19:46
  SPEAKER_01: Yeah.
• 19:47
  SPEAKER_01: I I love that.
• 19:49
  SPEAKER_01: I love that.
• 19:50
  SPEAKER_01: I really do.
• 19:51
  unknown: Yeah.
• 19:52
  SPEAKER_01: Take the idea in my head and make it a thing.
• 19:55
  SPEAKER_02: Yeah, it kind of gets you past that, you know, I'll know it when I see it.
• 19:59
  SPEAKER_02: Just come back with something else.
• 20:01
  SPEAKER_01: Yeah.
• 20:01
  SPEAKER_02: You know what I mean?
• 20:02
  SPEAKER_02: You get you get to remove that obstacle of like, I don't know how to make this thing.
• 20:08
  SPEAKER_02: Totally.
• 20:10
  SPEAKER_02: Uh, but I have an idea for what I'm is in my head.
• 20:12
  SPEAKER_02: And and like it to bridge that gap, I think it's super useful.
• 20:16
  SPEAKER_01: And to tie this all back to like Pat's conversation, it's why human agency is the leadership.
• 20:22
  SPEAKER_01: Uh human agency is still the thing that matters in all of this.
• 20:26
  SPEAKER_01: Totally.
• 20:29
  SPEAKER_02: I had a qu uh one thing I wanted to ask you about too, with regards to her conversation.
• 20:34
  SPEAKER_02: Like, there's that we had talked before about strategic drift.
• 20:37
  SPEAKER_02: And yeah, and like AI is a like separately, there's also a conversation I've had with some of the people in the Calgary Marketing Association about AI and about the skills we need for the future.
• 20:54
  SPEAKER_02: And everybody agreed that the AI AI in general is not a skill, it's a tool.
• 21:02
  SPEAKER_02: So then what are the what are the skills that we need and how do we prevent like the idea of strategy-ish or strategic drift?
• 21:13
  SPEAKER_02: Like what have you thought much about that in terms of like just where AI goes and how it supports or how do we keep it in the rails so that we're not letting it make decisions, you know, that idea of agency.
• 21:29
  SPEAKER_01: Yeah.
• 21:30
  SPEAKER_01: I want to say I'm kind of in it right now.
• 21:32
  SPEAKER_01: And the the reason being is, well, I think like you, we have many different frameworks in our heads constantly for different things.
• 21:39
  SPEAKER_01: Like they're constantly, you know, coming up.
• 21:42
  SPEAKER_01: The one thing that I'm actively trying to work on is build a strategic foundation that is informed by commercial strategy that helps then build a marketing strategy that then, because for those that don't know, like I lead the media kind of division on for TELUS, so all things media, that then helps inform, say, The media strategy.
• 22:02
  SPEAKER_01: You shouldn't build the media strategy first and then try to retrofit moving up.
• 22:06
  SPEAKER_01: You know, how do then how does it play into your marketing strategy?
• 22:08
  SPEAKER_01: Or then how does it play into your commercial strategy?
• 22:11
  SPEAKER_01: Has to be the opposite.
• 22:12
  SPEAKER_01: And one thing I've noticed as I've try I'm trying to provide more rigor for the respective teams in each of those areas is when I'm either prompting AI to help ask like basically because I'll get questions from these stakeholders, and then I'm going, then I go back to the LLM.
• 22:28
  SPEAKER_01: Okay, how about this?
• 22:29
  SPEAKER_01: How do we think about this?
• 22:30
  SPEAKER_01: And adding it as a part of the framework, how do we acknowledge this nuance in this industry, et cetera?
• 22:35
  SPEAKER_01: That it naturally wants to take it in a different direction where it's not what we're trying to do here.
• 22:42
  SPEAKER_01: Because, like, you know, we have the playing to win framework that I know you and I both love.
• 22:46
  SPEAKER_01: How do we use that as a deep root?
• 22:48
  SPEAKER_01: How do we make sure the category entry points become part of that, you know, the strategic rigor?
• 22:55
  SPEAKER_01: But I find that when you start prompting, sometimes the LLM itself tries to build a whole different ecosystem around what it is that is answering that one question you just put in front of it.
• 23:08
  SPEAKER_01: Yeah.
• 23:08
  SPEAKER_01: Where it's like, no, no, no, no.
• 23:10
  SPEAKER_01: I need, I need to come back.
• 23:11
  SPEAKER_01: So again, it's an example of someone that maybe doesn't have some of those frameworks in their head, could very easily just go down this path where it becomes, you know, strategy adjacent or call it whatever you will, where it's like, you know, you and I, people listening to this podcast, will have that rigor to say, no, let's pull it back because we're trying to make sure it works within this framework.
• 23:36
  SPEAKER_01: So you do see it.
• 23:37
  SPEAKER_01: You do see, and it's it's hard because, like you're like you said, the ants, the ants are so confident, right?
• 23:43
  SPEAKER_01: And they'll they'll it it will manufacture a use case that actually makes sense in that setting, but then it's actually a departure from what you're actually trying to solve for.
• 23:53
  unknown: Yeah.
• 23:54
  SPEAKER_01: I don't know if I answered your question or not, though.
• 23:56
  SPEAKER_02: No, I I think he did.
• 23:57
  SPEAKER_02: I I mean I I the way I'm kind of thinking it just occurred to me that as you were describing that and thinking about that pie shape thing, the reason why I was thinking about the like the cross hatch now instead of it just being a pie shape, is because the the way I describe it is is that the the broad T, the top bar.
• 24:16
  SPEAKER_02: There's equally as important like for me as a specialist to have knowledge of my area of expertise and then also knowledge of AI across the top is like and like my general knowledge, but also from an organization's perspective, yeah, their cultural way of thinking or their way of doing things, that kind of thing.
• 24:37
  SPEAKER_02: So, as an example, we talked about you just talked about frameworks.
• 24:41
  SPEAKER_02: If if the straight uh if the playing to win strategy model is the framework you want, then put that in as the model, and then everybody can use it.
• 24:53
  SPEAKER_01: Yeah.
• 24:54
  SPEAKER_02: Right?
• 24:54
  SPEAKER_02: And and I've done that with different clients.
• 24:57
  SPEAKER_02: But it's got me thinking, like, I almost never use AI's chat.
• 25:02
  SPEAKER_02: It like it just occurred to me actually projects.
• 25:06
  SPEAKER_01: Yeah.
• 25:06
  SPEAKER_02: Yeah.
• 25:07
  SPEAKER_02: Like I was looking at mattresses, we're looking at buying a mattress.
• 25:11
  SPEAKER_02: So I was looking at mattresses and and I was using the chat.
• 25:14
  SPEAKER_02: I'm like, oh, that's weird.
• 25:15
  SPEAKER_02: I almost never use just regular chat.
• 25:18
  SPEAKER_02: And it occurred to me like it just I I think that's an important to kind of kind of piggyback off of what you're talking about, the strategic drift part.
• 25:27
  SPEAKER_02: I think that's the leadership part.
• 25:29
  SPEAKER_01: Yeah.
• 25:30
  SPEAKER_02: Where if we bake in our philosophies, our fundamentals, our rules of engagement, whatever you want to call that, our ways of working, the cultural nuances, and you bake that into the AI, there's a training onboarding program there that you need for the AI just as much as you need it for the people.
• 25:50
  SPEAKER_01: Yeah.
• 25:51
  SPEAKER_02: And I think that's part of the way that what it as you were describing, that's what I was hearing in my head is like, I do think there's an element of onboarding that is really important for all of us to think about.
• 26:04
  SPEAKER_02: And it could change from one manager to the next or even one department to the next.
• 26:10
  SPEAKER_02: Totally.
• 26:12
  SPEAKER_02: But I think that that's something that's really important for us to start to really uh gain some level of agency over top of the Yeah.
• 26:27
  SPEAKER_01: I agree.
• 26:28
  SPEAKER_01: Stop being so agreeable.
• 26:29
  SPEAKER_01: I'm sorry.
• 26:31
  SPEAKER_01: It's the first time we've been we've been sitting in the same virtual room in quite some time, so I'm just being nice.
• 26:37
  SPEAKER_01: So Patz, if you're listening, thanks again for spending your time with us at Cannes.
• 26:42
  SPEAKER_01: We genuinely really appreciate the conversation, and thanks to you listeners, for keeping us in your ears.
• 26:48
  SPEAKER_01: Stay sharp, everyone.