Your Opinion About AI Has an Expiration Date: Sean Campbell

Jordan:0:00
Welcome back to the [High Ticket AI Systems] podcast. Everyone, please welcome the winner and only Mr. Sean Campbell. Sean, how you doing today, my friend?

Sean:0:08
Doing great. Thanks for having me on, man.

Jordan:0:10
You're very welcome. Now for those of us that don't know the brand or the company name just yet, can you please explain to us at a very high level, basically what you guys do and who you guys serve best?

Sean:0:21
Yeah, so Cascade Insights has been around for about 20 years. Historically, we've done a lot of work with the tech sector. And we really serve two markets. So for the last several years now, we've been helping organizations adopt AI meaningfully. And at the same time, if you're building AI into your solution, you're trying to integrate AI platforms into an existing solution. We do a lot of research studies to help you kind of understand markets and buyers and that kind of thing. And those two things kind of work together to some extent. And then the other thing I briefly mentioned is in addition to running Cascade, I've always had a joy of teaching and doing that kind of thing. And so I've had a couple of stints at a university level. So currently I'm teaching in George Fox's business school as a assistant professor of artificial intelligence. being there basically refers to me somewhat jokingly but real at the same time, like you're the AI guy for the department.

Jordan:1:21
No, actually though, you actually are. That's pretty cool, Do you feel like organizations, not organizations, institutions are going to start adopting like AI more so in regards to curriculums and like bringing experts like you into their institutions to actually teach this to students?

Sean:1:40
Well, I sure hope so. mean, one of the things I really like about what Fox is doing is they, and I should say this varies a bit by department, as you might expect. Like certain departments are going to feel more excited about AI and others maybe not. But the business school is really pretty forward in that. Like we're reinventing the whole MBA curriculum too that's going to be AI centered. But you're right. There's a lot of like real concern. and resistance. And to be fair, it's a very different problem to solve at the collegiate level versus, let's say, a kindergarten classroom. Because along that continuum, you've got scenarios to keep it brief. At the kindergarten level, well, they can't give away all of their thinking to AI. What do we fill a kindergartner's or grammar school student's head with when AI makes access to knowledge kind of ubiquitous?

And I know somebody probably heard that went, but it hallucinates. But let's assume it's giving mostly right answers just for the sake of argument for right now, right? You know, like, what do we train on? What do we teach on? Right? At the collegiate level, it's a different question in a way. It's like, how do we make sure that what the student is paying is actually worth it? How do we make sure that they're getting the right kind of learning and the right kind of just skills? And I'd say one last thing on that is that I think a lot of universities, not that I can speak for all of them, but I think a lot of them are going, There's a lot of soft skills training that we need to bring back. Like we might've been training hard skills training, but if raw knowledge and raw like thinking power is going to be an AI thing to some extent, you know, then it's about relating to people, communicating empathy, all those kinds of things. And traditionally, technically, unless you're in a program that really focuses on that, you know, you're not going to get a lot of that. You're going to get a lot of like domain specific technical knowledge, which isn't the same thing. So yeah, lots of changes going on.

Jordan:3:32
It's exciting stuff, would you agree?

Sean:3:34
100%. I mean, I think it's also frightening, though. I mean, to be honest, if you think about a college environment, there are a lot of people that their lives are invested in the idea that I know things. And to some extent, I mean, to be honest, I know things that others don't, and my job is to provide them that knowledge. And so AI kind of breaks that down a little bit. Now, for some people, like me, like I say, I get out of jail free card because I'm the guy teaching AI.

My challenge is just keeping up with it, right? Because it's thundering ahead at such a great pace, right? But I don't have to deal with that same existential problem the way some other people do. So yeah, but to me, I think it's a win because it helps people make smarter decisions as a general rule. I think it levels the playing field. this is maybe a little bit of a diversion, but I just want to throw this in really fast. I think it's also a great equity level.

Like, for example, you know, we know that some people write really well and we know that some people might communicate really well verbally and everybody knows somebody who doesn't do either one or both of those things well. But those people have stories to tell, right? They have things that they're willing to share and like AI can make that happen, right? Because AI can take some thoughts that you have and build a really wonderful written piece. And I got a story that I could share on that later if you want, but I think that's...

That equity leveling thing is something is going to be really interesting. But of course, that presumes people have access to the tools at an appropriate price, which is a whole other topic.

Jordan:5:13
On the topic of staying ahead of the curve with AI, mean, with your background, with your profession and you teaching at a very collegiate level, you have to. So the way I stay ahead of AI is I just watch people on YouTube, right? But I imagine you and I are in different ballgames when it comes to staying ahead. So from your experience and again, with your profession, how do you personally stay ahead of the

Sean:5:39
Well, it's interesting because I have people ask me this all the time, like all the time. Students ask me this, other professors ask me this, employees that work for me ask this. And the funny thing is there's a part of it that is just raw data gathering that you don't really get away from because as great as the AI tools are, right? And I have custom apps I build that'll go churn through my inbox and the 50 or so, no lie, like, sub stacks I subscribe to that are AI driven. And I've got tools that'll go through podcasts and generate. mind maps and transcripts and all this kind of stuff. And I can chuck that all in a blender and basically get a synthesis. But what I find is sometimes, no matter how tuned it is, there's going to be kind of things you kind of miss, particularly maybe like an editorial perspective or a way of looking at AI use that's a little bit unique. So a lot of those tools would do a really good job of like grabbing like what got launched, what got released, you know, so I don't miss that. And then the other thing I would say, especially with AI tools is constant. experimentation because channeling stuff I've heard other people say, so this isn't like an idea unique to me by any stretch is that, you know, your, your opinion about AI should have like a milk like expiration date on it, you know, because if your opinion of AI is grounded in AI three months ago, that can be very, very different depending on the linear. And then worse yet, it breaks down by task, you know, How many people thought ChatGPT ImageGen was a poor also ran compared to Nano, Banana, and Gemini for the better part of the last six months or more? And then ChatGPT comes out and they're like, well, we're in the running again. And those kinds of things happen all the time. it's just really good situational awareness that I think you need to have. So some of it's just raw reading stuff in your inbox, watching YouTube, podcasts. Some of it's building tools to kind of make that easier.

But the space is moving so fast, there's always a new voice out there that's worth listening to too. And that's another piece of it. So I don't know if there's an easy way. There's ways that make it easier, but I don't know if there's an easy way.

Jordan:7:38
For sure. And let me see if I can articulate this correctly. So essentially you have custom AI agents that I guess are attached to specific sub stacks channels, YouTube channels that you follow whenever a new episode gets released or a new thread gets released on sub stack. I have no idea how that works by the way, but roll with me. You throw it in a blender, you kind of spit that information. Okay, Sean, here's the latest updates, X, Y, blah, blah, blah. Is that kind of how that works?

Sean:7:45
Yep.

Sean:7:58
Yep. Yep.

Sean:8:05
Yeah, but I still need to sometimes hand read through some of the stuff because somebody's really unique perspective, the AI engine might not quite grab no matter how much I've tuned it. And of course, then I go look at what it didn't grab and I try to train it to do a better job. But I think there's just some of that. And some of it just is the soak time of, anyway, you get the general idea. That's basically how it plays out.

Jordan:8:30
For sure. Thank you, my friend. And on the topic of the two aspects or the two pillars of your business, one serves organizations adopting AI and the other side is actually building AI products. So how do you stay credible with the adopter side when vendors are also writing checks for research that if it goes their way, makes their product look good?

Sean:8:52
Well, and I totally get why you would say that because like a typical, the good news is it's not us, but I can make that really concrete. Like there's a huge part of the research market that's the analyst market, you know, where they're basically paid to go write public studies that make somebody's products, you know, ultimately look good or in rare cases bad because they don't line up right on, let's say the Gartner Magic Quadrant.

Now that's not to say we don't occasionally, a limited percentage of time, do thought leadership research where somebody has us go write a paper, but the lion's share of our work is actually for internal use. So our research is used by the companies internally to make their products better. So there's really no friction there. And clearly, whatever we've learned from like an NDA standpoint about, let's say a new product launch or something that's coming that we're doing research on, well, that doesn't make its way into anything we educate. Obviously, that would be problematic.

But there is this nice ability to kind of understand AI from both perspectives, the people that are creating it, and at the same time, the people that are adopting it. And frankly, the adoption training work also helps us on the research side, because we'll talk to, I don't know, the college, the department, students, organizations that I train, and all the rest. And all of sudden, we're like, that next survey we write.

I think there's this thing happening in the market that we need to make sure people pay attention to and actually address. So it's just to me, it's a really good cycle. But you're right. If we were in a different style of research, the more pay to play kind of thing, then there'd be much more of a potential conflict for the way it's played out.

Jordan:10:31
Gotcha. And this might be a loaded question, but from your experience with all of the AI products that you've built, what are like the top three most valuable and will have the biggest impact on your organization? If you could even answer that.

Sean:10:44
Well, when you say built, guess, let me, think what you probably meant was more like that's out there. You know what I mean? Like the stuff that's out there. if I was to say, cause if I said like what client software we've done research on, I we've done it in narrow spaces and broad spaces. So I can tell you like, well, there's the CRM product that I really like or whatever, but that might not resonate with anybody because maybe they're not looking for a CRM.

But I would say right now, if I had the bag of 100 nickels that I was gonna go spend it on stuff, I would spend more money on Claude than anything else. There's a million reasons for that. My students feel the same way. Most of my students come to class and they think of Chachiputte, because it's like the Kleenex of AI right now, right? And so they're like, that's what I know. And then they get to working with Claude and they're like, this thinks better, there's more capabilities. Then I show them co-work, then I show them kind of how Claude code can actually write some really meaningful code. And again, like I said earlier, know, ChachiPT is catching up, you know, and then, you know, but on the, and on the visual side, I'd still say it's probably Gemini, but visual is an interesting one. Cause I think, not to go too long on this, but like on the visual creation side, I see people kind of cheapen their efforts pretty early. You know, they, they go into ChachiPT or Gemini and they make an image they kind of like, and they're done. But you know, there's not only those platforms, but there's things like mid journey and ideogram and a whole bunch of like. you know, really narrow models for like video gen and animation and all this stuff. And I think the problem really is like, it's just, if you don't know the space of working as a creative, if you've never even like dabbled as a photographer, you know, like in my house, I self-phoned it by the family. take better photos than anybody else. And that may just be because I always have the newest camera. Like maybe that's it. Like that's the only real reason, right? But, But I feel like I've got enough aesthetic sense that I can drive those tools a little bit like a photographer or an illustrator would, even though I've never had those jobs. Anyway, there's interesting things when it comes to AI use and kind of tools. But I would say that side of it, definitely a big fan of perplexity when it comes to any kind of research or pulling stuff together. It's kind of what we wish Google was. Like just this morning, I was at the gym. My son's looking for an F-150. So if anybody has a good

Sean:12:57
used F-150 that he'll feel pretty, email me. But anyway, the point is, it was one of those classic, okay, so he wants to tow his race car trailer with his spec Miata in it, which has a weight, but then is that the tongue hitch weight? Don't worry, I'm not gonna go longer. But there's all this math, anybody who's towed knows, and it's like, I could go to Google, but then I'd go to 72 calculators. And I just sat there with perplexity, and it just kind of gave me the synthesis I wanted. from all that open source Intel basically. And I'm like, son, here's a PDF. This is exactly what you need to do. yeah, anyway, so those are just some things that stand out.

Jordan:13:39
It's so funny because everyone, whenever you're first dibbling and dabbling in AI, of course you're going to use chat GPT. Like you said, it's the Kleenex of it. But once you really start playing with cloud code, you're like, dude, chat GPT kind of sucks.

Sean:13:53
No, no, it does. And it's funny because my students even like will literally call in the first couple of weeks. I noticed this start to happen about a year ago. They'll be like, they'll say, well, I use chat. And that always means chat. Like they've shorted AI to chat, right? Go figure. Right. And then they're a little this is okay to say, because I think the students I've had would know where I'm coming from. If anybody listens, they're a little frustrated that I'm asking them to buy a Claude license because, you know, every college student you know, money matters, right? And I'm like, no, no, no, promise. I promise you'll, and by the end of the semester, everybody's like, Chad, I don't use Chad anymore. I use Claude, man, you know? And it's like, feel like I should get like a little like accelerant for all the Claude licenses I've probably like co-sold, but like, no, I think it's just legitimately a really good model. But here's the thing, I'd keep it real. And I think you probably feel the same way too. It may not always be so, right?

Jordan:14:26
We're

Sean:14:51
I mean, like that's the thing that like, I'm always like, don't declare a winner. Like, you know, your IT department wants to declare a winner because they just want one model because it makes it easy for them. As the individual professional, do not declare a winner. Don't, you know, and I'm like, and then people will be like, well, it's 20 bucks and it's 20 bucks. I'm like, dude, like if you actually use quad for the 20 bucks a month and you use it every day, the amount of time saved. you will be able to sit at Starbucks and do no work. It's not that it's the cost of Starbucks. You were able to go to Starbucks and do no work while Claude works on the laptop you left in your car. You know what I mean? But it's a hard thing to get through to people because, and this maybe takes us somewhere else for a minute, but just really briefly, I think people struggle to work with AI platforms to some extent because it's such a different interaction model. I mean, to me, I found that the people who are the best without any training, if you just drop people in front of an AI and say, go, somebody who's really good at delegating to humans is almost invariably really good at delegating to AI because all the same skills apply, right? Like I set good context, I give good directions, I ask what I want for output. Like these are things we hope good bosses do when they talk to us, right? And they say, go do this. They don't go, go do this and stop. You know, they bring all this other stuff to the table, right? And the people will walk up to an AI and they treat it like, Google that's more verbose, like it's not nearly as good for them, right? And then they get frustrated. so, so anyway, I just, that's, that's kind of where I'm at, you know, with a little bit, but.

Jordan:16:29
I think it makes it a fantastic experience and journey from the consumer standpoint, because you have all these AI companies, OpenAI, Anthropic, whoever the hell owns Proplexity, really competing to become number one. What that really means is they are working day and night, 24 hours a day, or 24-7 to make the best model. So we're always getting the best improvements, the latest model, the best of the best.

Sean:16:57
They're always looking left and right and they're like and by the way just quick thing don't we all feel sorry for the guy that owns perplexity? Cuz like everybody else knows who you're like the entropic guy You know who he is the chat GPT guy if we took a poll and we're like, who's the guy that runs perplexity? They're like I Don't know He's probably got like a huge ego problem all day long. He's like, come on marketing team. I want my name out there, know, anyway, sorry I just couldn't resist that

Jordan:17:15
no.

Jordan:17:22
But, but I, no, you're totally fine. I personally never used Proplexity before. So can you kind of, you said you like using it for a research purpose. This is probably off topic, but it was fine.

Sean:17:32
Yeah, yeah. mean, the short version about perplexity is, like I said, it's like what you wish Google was. Like the way I went to a conference talk a number of years ago, this long before AI, and the guy I think framed Google the best way ever, even though it was way before AI, said going to Google is like going up to a small window and saying, I would like to learn about penguins. And a guy on the other side holds up a penguin and says, this penguin?

And then he walks over and he says, this penguin? And he opens up another one. And you have to go to each penguin and inquire of it if it's what you want. Now, again, to Google's credit, that link algorithm that they developed, mean, I'm old enough, I'm 55, that I used Google when it had a beta sticker on it. I remember that day on the home page. So it was revolutionary. It's why we don't use Yahoo anymore and a million other things. So all that's legit.

But what we kind of wanted is something in between, right? Something that I could say like, well, you know, we're all back to that F-150 example. I could use anything, but since we just brought it up, like there's a million websites that have like tow ratings and combos and what the size of the cab impacts the total towable weight that the truck can pull and all this stuff. And it's buried in PDFs on Ford sites. Well, I can go click through all those and eventually get some synthesis of the ones that I might want to buy on the used market. Or I can give more of like, a direction, a delegation to perplexity. And it will surface all that data for me in exactly the same way, but do it in a way that's almost like a research package that comes back to me. And perplexity computer, perplexity computer, which is something they released not that long ago, is actually like that on steroids. Like you can do some pretty amazing, like long running research, like activities and stuff. So I think it's not as heralded in the toolbox, but I...

I don't turn to it as much as Claude, let's say, but when I feel like my problem is more go collect data and package it up for me, I will tend to go to Perplexity to go do that.

Jordan:19:34
Understood, understood. Okay, cool. We can definitely stay there for a minute, but let's move on here. So the sales schema episode you did was titled 15 years of inconvenient B2B truths. What's the one inconvenient truth about B2B buyer behavior in 2026 that vendors keep refusing to hear even when your research puts it directly in front of them?

Sean:19:56
would say the thing that has been so consistent year after year, and clearly there's multiple things that we end up bumping into. Because when you do research, you see patterns. I'd say the biggest thing is they aim their marketing too high. We will have people come to us and say, I want to go do a research study on CFOs. And I go, why do you want to do research study on CFOs? Because we want to convince CFOs they should buy our widget. I go, OK, what role does a CFO have in the buying funnel?

Well, the important one, they say to me, almost like I should know this. And I'm like, well, of course the CFO is important. How much time are you going to spend with the CFO? And they're like, huh? Like, no, no, I mean, like in the buyer's journey, like how much time do you spend with the CFO? They're like, the end? I'm like, yeah, the end. That's important. said, so when you go do your marketing, do you think the CFO is going to see your marketing? They're like, yes. Like, okay, let's try this again. So at the beginning of the buyer's journey, who looks at the marketing? The CFO? No. Who?

I'm like, okay, so somebody about three levels down from the CFO. Yeah, kind of makes sense. Why don't we build a study on them? Well, I don't want to build a study on them because somebody says the CF, you you see where this goes. This loop is like constant, constant. And I think, I know, I know that part of the resistance initially, when you talk to research firms, they think we're trying to make the study more executable because clearly getting a bunch of CFOs to go commit to a study is more expensive, hard, timely, whatever, right?

And they hear me saying, let's go to the director of finance and just making up a title, right? For, you know, lower down. And they're like, you're just trying to make it easy. You can't get to CFOs. like, no, no, no, no, no, no, no, no, no, no. Like if your problem is early funnel, like they're not reading it, you know, and, they'll build whole marketing portals. I've seen this. They'll build like whole studies and marketing portals, like the CFO and whatever. And I'm not saying all of that's wasted, but I'm like, they're We will bring that up and study after study and there will be some guy, and by the way, I have a lot of respect for B2B sales teams. I do. I know that's a super hard job. It's way more intellectually challenging than people give it credit for. So, so everybody hear this from that perspective for me. It's almost always a sales guy that tilts the apple cart then and says like, but no, the meeting with the CFO is super important. And I'm like, I'm not denying that. I'm just asking you where it is in the journey.

Sean:22:17
And what are you trying to do? If you were trying to have me do a research study on how to have a meaningful sales conversation with a CFO, well, that would be different then. You know what I mean? But I feel like that's the biggest, it's honestly the biggest hang up. if it sounds like it frustrates me, it's more just because we sometimes will do studies and the client gets to decide ultimately what they want to target, right? I mean, as every service firm owner knows at some point, or they should, if their ego is not too big. you are in service to your clients. So at some point they ultimately get to weigh in and decide, right? And so like, you know, we've done these studies where I'm like, I really, really don't think you need to spend all that money to survey those executives. think you should go down and they're like, we need executives. And I'm like, okay, I've said my piece. We're going to go do it. I don't think it's really going to help you as much as you think, but anyway, so that'd be my biggest thing.

Jordan:23:03
Right.

Jordan:23:08
Now you're good. With all the research you have conducted over the last few years, what has been the most surprising to you? Of course, on topic, I've got you out here.

Sean:23:18
gosh, I mean that's hard. What's the most surprising thing?

Sean:23:25
The most surprising. Okay. I don't know if this is, this is probably the easiest to say, I guess. I mean, I'm not trying to find the easiest, but going on recency of what pops into my head first is probably the safest thing. the people don't act on the truth. I mean, I had a student once when I was teaching an MBA program, like 14 years ago, who said, and I stopped cause my kids were young and I wanted to coach baseball. And then I ended up back at college now, but like this, this young, yeah, this young kid said, do you get paid?

Jordan:23:37
For sure.

Jordan:23:50
Love baseball.

Sean:23:55
if they don't act on your decision, what you gave them. And I'm like, Padawan, no, there's no way that would work because people decide against their self-interest all day long. I mean, that's probably the biggest surprise to me is we will put data, and I'm not saying if somebody hears that, they're like, well, certainly you don't know everything. Of course I don't, but I'm talking about like a pie chart that's like 97.3, you know, like it is absolutely ironclad that this should not be done.

And we hear later like, well, Charlie the executive decided that we still need to do it. And it's the three. And I'm like, really? I'm like, why did we go through all this? I mean, honestly, that's the biggest thing, is that they decide against their self-interest in violent ways, even though the data is just clear cut. And not everybody's that way.

Clearly, like lots of people act on like, you the news that they hear, but I think that's the part that just befuddles me to this day, owning a research firm.

Sean:25:02
I would like to, but it's probably hard to give one that would feel deep because like, like, like I either have to like hide who the client is if I give you a detailed example, which then makes it hard because of NDA, or I have to like not tell you a lot about what happened and I named the client name. So that one I think is harder for me to dig into, but I totally get why you'd ask.

Jordan:25:24
For sure, for sure. Okay, cool. Moving on here. The cascade philosophy is built around fluid intelligence. What's that? Repeat. The cascade philosophy is built around fluid intelligence practitioners who also do the research instead of pure researchers reporting from the cheap seats. What does the pure researcher model break down most painfully? And was there a specific engagement that pushed you toward that thesis?

Sean:25:53
Well, I think the biggest thing, yeah, yeah, I think there's a number of things that to get at what I think you're getting at that like, like how does AI change, you know, the role of a researcher? And I think I could put a finger on analysis and I could talk for a long time about how like, you know, quantitative analysis has made more efficient. Like Claude is amazing at dealing with, you know, quantitative data. But the one that I think is less heralded, but I think it's probably had the biggest impact that clients might

Jordan:25:53
If you need me.

Sean:26:23
feel even more than they know. And I've seen it play out now to know that it's definitely happening is visuals. Why do we not have visuals in the deck? Because it takes time. mean, that's why slide decks have a bunch of bullets, right? You go build a presentation circa five years ago, and you know you need great images. So you run the substack, right? And you hope to grab a bunch of stuff that nobody else has grabbed so your presentation doesn't look like everybody else.

Right. And you grab some images and you get it all set up and you put it up there and you put some text on it and you run it and you're like, woo, it looks great. And then you go present the guy next to you has the same header image, but that's a separate point. And like you go through that. And then, but it took time. Right. And so you got some slides in the deck that are still just bullets, right. Or hero images, right. Or whatever hero tax. We have the issue in spades with research, right. Because, you know, you've got all this data to communicate, right. All this context, because back a little bit to what I was talking about earlier, right. Like.

For people to make a decision that is contrary to what they currently believe, they have to have a lot of justification. So we have to put lots of justification into a deck, which can make it look boring if you're not careful because it's a lot of data. But everybody knows a visual communicates more. But the reason decks are boring up until you had an excuse for it, I should say, up until six months ago, is that it took a lot of time. Build a visual. You don't have that excuse anymore. So we have been just building some banger decks and output that is just impacting people viscerally more because it's an image and it gets in their gut more. And it's not any less truthful. It's just, grabs people's attention more. It's the same truth packaged more interestingly, which is fine. And the best concrete example I can give you that is we had a client, again, this goes back to like, can tell you a lot about the story, but I can't say the client name or I can do the opposite.

Jordan:28:08
in the next

Sean:28:19
So I won't say the client name, they make software people use a lot of. so somebody on the research team said, my team is so tired of Dex, the stakeholders don't look at Dex anymore, that I want you to go build basically a graphic novel of the findings, like anime style. And we did. Coolest thing we ever built. The biggest downside is...

Because in our line of work, can't post it anywhere, obviously, because it's all like the client's data. Like I would love to, but for that one client, I can show them it over and over and over again, all the other people we could work with, because I'm like, look at this really cool thing we did. The idea of building like a comic book style, Japanese anime of all these research findings, you know, I would have easily put on it like a big line item of cost. Like I would have, you know, and then the client would have said, give me the boring deck.

Like that's better, I'll take the boring deck. I'll take the bullets, you know what I mean? And we'll be fine. And now I can sometimes, like for no change in cost, build this other thing that really changes people. And to finish this off, like I tell people on my team like all the time, and I feel this, like, you know, we are in the business of helping people make a better decision. So either to feel really good about the one they were about to make or to go, no, we need to make a different one.

So anything we do that makes people interact better with what we shipped is a win, right? And so to me, that's the biggest thing. And then you take one other click. I gotta say this really fast, because you probably know this. We're at the point now that even the concept of building a deck will go away. When I closed off the semester this semester with the students, Quad Design had released about a week or so before we were wrapping the semester, right? Maybe two weeks.

And I had this little riff with them where I was like, look, between notebook LM making decks that are now slightly editable. mean, some people might quibble. I think they're slightly editable the way they are now. And then you got Claude design and you got all this idea that we even like think of deck production as a separate thing might not even exist in two years. And that, that alone is crazy to think of. Cause if you've been in business long enough, like, like you feel like half your life has been building decks at some point, you know?

Jordan:30:38
Especially at the executive, you know, C-suite level, 100%. I kind of want to pivot if you don't mind. This is just a person of interest of mine. So there's a lot of people out there that are generally lazy. And now that we have AI at our fingertips, it can learn anything, it can think for us more importantly. How do you recommend students in college, high school, and even younger?

Sean:30:41
Right, right.

Jordan:31:03
actually utilize AI while still growing their their brain simply put.

Jordan:31:11
Is that kind of?

Sean:31:11
It's such a great question because the answer today won't be the answer in six months or a year. Something I just read earlier today from a guy, Ethan Molluck, that I really appreciate. He's an academic and he said, tend to want to believe that we'll be left to decide everything. Deciding will be the human thing. I think that's still mostly true.

But he said, you know, we got to face the fact that AIs are really good at giving us suggestions and strategy. And at some point, well, doesn't that put them right on the verge of deciding? You know what I mean? what's, you know, and if you're, if you're Musk and for the record, I'm not a Musk hater or fan boy. I mean, I do like the idea we get to go to space in a meaningful way. So hate me for that or not. But like, but yeah, so separately, you know, I mean, go figure. I'm a tech guy, so I probably would like rockets. You know what I mean? At the end of the day. But, but, um,

Jordan:32:01
Yeah, dude.

Sean:32:10
But Musk is, you you talk to Musk and he's like, well, no, no, the solution to this is like, we don't have to worry about that. Cause like everybody's, you know, production is going to drop to zero cost. Everybody's going to be fine. You know? Okay. So, so the point is it's, it's, it's a moving target, but what I walk students through is a little bit of a framework I use called rise, which is basically says that in, the current state of AI use, which is all I can meaningfully teach them on, right. Is routine tasks in your job. should just go away to the AI. Whatever is routine, whatever is like boxable and whatever, do not like tell yourself that is something you're going to spend a lot of time learning and figuring out. mean, there used to be a guy in every office that knew like weird janky Excel formulas, like, know, index match and VLOOKUP and you know, and he was proud of it. And you always went over to his desk and you were like, Hey Bob, like, you know VLOOKUP. You know what I mean? Can you teach me it? Okay. Well that, we don't need that guy anymore. Right. I mean, we've, we, so, So there's routine and then there's like investigative, which is a bit like the perplexity thing. Like it used to be kind of like, like an art to go get all those toe ratings and all that stuff and do it faster than somebody else, right? Research penguins faster than anybody else. And then there's like strategy, right? And that's where Malik is going, well, we used to think this was just human. Maybe not, because anybody's worked with Claude long enough goes, it can be pretty strategic if you really like push it, right?

And then the last part of that framework is engagement. And so what I've been telling students is focus on human to human engagement. Like in my class, they do a huge debate with each other on a whole host of AI topics, UBI, whether we should have it or not, ethical considerations. It's a huge list. They get to pick. And what frightens me is when I've polled my students, how often any of them have engaged in a public debate in front of their classmates.

You structured, not just yelling at each other, right? At best, one of my students each semester has ever engaged in a debate.

Sean:34:17
And think about that, right? Like what's left to do when AI does everything? Debate. Like that's what we get left to do, right? For better or worse, right? That's what we're left to do, right? I mean, cause we, you know, and yet nobody knows how to do it. And I'm like, they're not training it. And so like, it's, I really love that the students go through it cause like it's frightening for them too. And the same thing happens like you'd expect, right? They get up there and they're like, and I also, by the way,

Jordan:34:27
Yeah.

Sean:34:45
The last couple semesters, basically had them argue the opposite side they believe on after they've argued the side they believe on. And they all do the same thing. Like anybody who's ever done this in real life with any topic, they get done with the second one where they flip sides and they write in their reflection at the end of it. I don't feel as sure about my original position anymore. I don't.

I'm not, and every once in a while there's a kid equally rightly by the way, who's like, now that I argued the other side, I'm pretty sure. I'm like, good, good. That's what civics should be. You know what I mean? Like, so, so what am I trying to say? I think it's all the soft skills stuff. think, I think you, and the other thing I tell them is, you know, we used to say you were T shaped as a skill, right? You, you built this one 10,000 hour thing. You learned all about it.

You got paid for that your whole career. Your goal was to make the part of your T longer than everybody else. So you were super genius on that. And that was your way. And as long as you picked a good T shape, you were set. I'm like BW shaped, shaped, Z shaped, whatever, know, like add more skills, but those skills are AI enhanced, right? If you were really good at verbal and you sucked at writing, well, you don't have that excuse anymore.

You can be great at writing. AI can fill that gap, right? But you need to feel comfortable using it. So anyway, that's kind of the, but I'd sum it all up by it's incredibly a moving target because I'll even have these conversations and something will launch in the middle of the semester. And then we will wrestle with it. And we will be like, well, I guess AI can do that now. Now what do we do with that? Right?

Jordan:36:27
That is the big question, know, because if you as a human being always rely on AI to answer for you, to think for you, wouldn't that like, I don't know, decrease your cognitive ability over time? Do you think?

Sean:36:39
Well, 100%, 100%, which is why I'm like, leave that E for you. Like even if the R and the I and the S all get taken up, you know what I mean? Like the ability to engage with another human that you have to develop. You absolutely have to develop. And the truth is for better, unfortunately, most college education, like I talked about before, and I think this is even equally true of primary, although I should be careful here because I've never taught K through 12, but my recollection of it.

Jordan:36:49
I see.

Sean:37:06
as long ago as it was, and even what I saw my kids go through, is that it's very hard skills. It's like, let's teach long division. Let's go teach blah, blah, blah, blah, right? And it's not about how do I relate to my classmate and have a good civics debate about something, and how do I deal with really thorny issues? That's a lot harder to teach. I think the last thing on that is, why is it hard to teach? Because it's actually hard to grade.

Jordan:37:27
Mm.

Sean:37:34
I mean, this isn't an educational professor podcast, but it's super hard to grade. It's way easier to grade. Did you get these long division problems right? Than it is to like, did you have a good debate? And anybody who's tried to grade debate knows that that's or grade public speaking or any of these human centered skills. It's a very different classroom to run, but I think it's essential.

Jordan:37:36
It's interesting because it's high.

Jordan:37:57
It's not black or white like science or math. It's yes or no. It's not.

Sean:38:01
No, no, there's a whole lot of maybe in the middle, right? And there's a bit of like, you know, you know, and there's a bit of like, you know, like, like back in the day when I was getting my master's degree, I was teaching public speaking classes at Purdue to help pay for it. And, you know, I walked out of that experience saying, no one teaches public speaking who doesn't grade on a curve, because there will be somebody who will walk up and we'll just nail it. You know, I can tell you story after story, but there's somebody else who gave a really, really good speech, but they were like shaking like a leaf.

You know what I mean? And it's like, or if you coach you sports, it's the same thing, right? You know, like, like there's kids that walk on the playing field and they like turn a double play and you're like, I didn't even tell them how to do that. I don't even, you know, and then there's another kid that the ball bounces off his chest every day. Right. And you're like, you're like, this isn't inherently my coaching ability. There's just innate ability sometimes that we all have. And that comes into play more, I think with these soft skill things, but that's not an excuse for, for developing it. Right. Like. not developing it. It's just, I think it's just harder. It's harder to deal with.

Jordan:39:06
would agree with you there for sure. I do want to possibly end on this question here. So what do business leaders need to think about when it comes to adopting AI across their entire team?

Sean:39:22
There's a couple things. So I'll try to package them up pretty tightly. would say, One, you have to herald failure as much as success because most business environments aren't happy with, know, I tried this thing for an hour and it didn't do what I expected, but AI demands that. You have to explore it. You have to, and you, and you know, you can do that with like a weekly Slack channel where people share their successes and failures, but as a boss, don't go in there and just like, you know, praise the people who were successful.

You need to say, that was a really good try. What did you learn from it? Because people are scared about it. think the second thing, just directly related to that point, people are just scared, They are. You need to really be clear and honest about what you're thinking about AI. You can't just walk around and be like, oh, we're never going to fire anybody.

Well, no, that doesn't really work because if everybody else uses AI and they get the production cost down to a third of what it used to be and you all can't find a new market and you can't go find new revenue right away, what are you going to do? You can't keep everybody around just for giggles, right? And I think the other thing is, and this is a big one, but it might run people a little wrong. Don't let IT lead the rollout.

Because IT was great and will be great at discrete point solutions. You know, the next Excel, the next SharePoint, the next whatever. Something that radically changes every task your team works on, that is absolutely the domain of the business leader of that team. So if you lead an HR team, you should be the one.

Sean:41:10
advising your team on what tools you should use from an AI stack. Or if you don't know how to do it, well, then you got to go find somebody internally like a champion or externally to help you. But to go to the IT department who wants to say, what's the one tool we should use in perpetuity for AI use with the cheapest set of licenses that is the most secure? Like that's not the way to win this game, you know, ultimately. And that doesn't mean to ignore security, but I think you can, you can be very secure if you're smart about it. But this like fusion of like, I'm changing my entire team's job from the routine to the strategy to whatever, IT can't help you with that. That's the domain of the person running that business, right? If you're a small business, it's on your shoulders as the key entrepreneur. If it's a marketing team, it's the CMO's job to come up to speed on that. And the part that I feel empathetic about is I wish somebody was giving you whole extra two days a week to just worry about it, but they won't. That's just life right now.

You know, some companies have more elasticity than others and they have the time to maybe give people like a freebie day to go learn AI and some won't. But that's what I would say. Don't leave it in the hands of IT. I don't mean ignore IT, but do not leave it in the hands of IT. That will slow everything up, 100%.

Jordan:42:24
If you are in that position where your job is possibly at risk, not possibly, you your job is at risk. What do you say to that person? How do you help them? I guess not worried, not living in fear all the time, maybe sort of adopting AI so they can maybe start a side hustle, start their own business with AI.

Sean:42:41
Yeah, well, let me, I don't know if this, yeah, yeah, there's so much to say there, I'm happy to say more on this, but I'll give you this example of a conversation I had with a student once and see if this gets to where you're going. I think it does, because I think of it a lot when people ask me something like that is the student walked up to me and let's just say the student was invested in writing, liked writing, and was really bothered by, I shouldn't say really bothered, but somewhat bothered by this guest lecture I gave in another class where I was like, look at all these amazing things that are happening to research and AI. And the student popped up and said, are you okay with all this? And I was like, well, one, I'm not really in charge of all this, but like, you know, it's fine. said, but what do you mean? And the individual said, well, you know, kind of after a couple of minutes, I was like, well, I really like writing. I don't like what AI is doing to writing. AI writing is horrible. And I was like, okay, okay, hold on a minute. I said, so. you get into writing the string prepositions together? And she's like, no, I was not. After a while, like, and then, and then I went up to the board and I drew a vertical rectangle and I said, so all this stuff at the bottom, you're just gonna have to give that up. Like string and prepositions together, like writing an amazing sentence. You might even have to give that up. I said, but why do you write? Isn't it to actually like impact people? And I said, so here's the deal. If the top of your box is dotted, that it goes on to infinity, you probably still can keep that job. If the top of the box is solid and you're just going to bump into it, I go get another job. I'm like, that, and cause it just, that's what I, and I still think that holds true. I think that holds true for a lot of jobs, right? A buddy of mine, daughter got married, married a kid that was going to be a medical transcriptionist. Well, I think we know those jobs are probably dead, right? But that's a box that's got a top hard line in it. There's no dotted line at the top of that box.

Jordan:44:25
next.

Sean:44:39
Right? And so that's what I would tell people to do. And if that means you need to change careers and change jobs.

I feel bad about that, do, but far be it from me to not tell you the truth, right? That's what you gotta go do, maybe, if that's the kind of job you got.

Jordan:44:56
We people like you to actually speak the truth, regardless if it hurts or not. One last question for you, Sean. What is the single best piece of advice you've ever heard?

Sean:45:05
Yeah.

Sean:45:10
Read stuff you disagree with.

Jordan:45:13
Read stuff you disagree with.

Sean:45:13
Hey, read stuff you disagree with. No, mean, because I honestly think it's the best piece of advice I've followed in a way because because it's a bit like debate. didn't I didn't I've answered that question. I've gotten it a couple of times over the years. It's a great question. I would have brought it up even if I didn't bring up debate because I think I think you end up either reinforcing your position in a more meaningful way. Right. Or you end up maybe changing your position.

Right. And most people don't really read stuff they disagree with. They kind of engage with it at a distance, right? They give it a frown on Facebook. You know what I mean? But they didn't actually read the post. You know what I mean? They didn't read it end end. And heaven forbid if they actually clicked through the links in the post and read some more, right? You know, like that's, that doesn't happen. And I think we're worse off for it. You know, and again, maybe your position stays the same. That's great. I got no problem with that. None at all. But I think you got to read more stuff you disagree with.

Jordan:46:11
My favorite teacher in high school always said this. said, unless you fully grasped both sides of the argument, you are not able to actually have an opinion on either side because you don't fully understand the context and what both parties are trying to articulate. So whenever we had to write our essays in high school, we had to first write about the opposing side and their beliefs and then our counter argument on what we think. And I think, mean, damn, I can't believe I brought that up, but

Sean:46:34
There you go. Right, right.

Jordan:46:41
Yeah, that makes a whole lot of sense, right? And I still agree. You're not allowed to have an opinion if you only know one side of the argument, dude. Like, no, you have to understand both. And I'm sure you can elaborate and we're running out of time here. Sean, you're the freaking man. Thank you so much for hopping on and sharing all the insights about it. Where can they find you? Yeah, where could they find you if they wanted to explore?

Sean:46:57
Hey, no problem. Thanks for having me on,

Sean:47:04
Well, so you can reach out to me at Sean, know, s-e-a-n at CascadeInsights.com and just go check out CascadeInsights.com. And then I've also got a newsletter on LinkedIn people want to check out called the Human Side of AI that I share a bunch of thoughts there too. But any of those places are fine, man.

Jordan:47:23
Thanks again, Sean. Audience, we'll talk to you all soon.