Kristi Zuhlke: Stop Talking About AI and Start Building

AI is moving too fast for insights professionals to sit on the sidelines. In this episode of the CEO Series, Leonard Murphy sits down with former KnowledgeHound founder and CEO Kristi Zuhlke to discuss why experimentation matters more than perfection, how brands and suppliers should rethink AI strategy, and what happens when researchers stop theorizing about AI and start building with it. From AI-powered workflows to synthetic consumers and the future of decision-making, Kristi shares a practical perspective on where the industry is headed and how professionals can keep pace.

Transcript

Leonard Murphy: Hello everybody. It's Lenny Murphy with another edition of the CEO series. Thanks for taking time out of your day to spend it with myself and my guest. And uh today I am joined it's been like two ships passing in the night. We've been trying to schedule this for for a while. Finally did. Uh but Christy Zulki. Christie, welcome.

Kristi Zuhlke: Thank you for having me. I'm excited to be here.

Leonard Murphy: Uh it's it's good to have you here. Uh so have you kind of do your bio in a minute but but for our audience uh known Christie for for a long time back from uh knowledge hound uh I think it's when we first uh first

Kristi Zuhlke: Sure.

Leonard Murphy: Connected but watching you lately on LinkedIn um of just saying just go build stuff and showing folks how to do that has just been a very cool thing and we we'll get more into that but why don't you give the audience a little more of your background how you got to this place of now You're you're an influencer now, right? You're not just building stuff. You're you're an influencer.

Kristi Zuhlke: I guess so. Yeah.

Leonard Murphy: Yeah.

Kristi Zuhlke: Yeah. An unintentional influencer for sure. Yeah. So, um, a quick background on me.

Leonard Murphy: So.

Kristi Zuhlke: So, I started my career at Proctor and Gamble and Insights and I didn't even know what Insights was when I joined the company. Just someone recruited me in there and um, and I was like, "Oh, that sounds interesting." And, and so I just got sucked into this world. So I don't have a background before that like in market research or anything like that but I was very entrepreneurial-minded and I think that was what someone at Proctor noticed in me and wanted more of that at Proctor and so she brought me into the insights function at PNG and it was amazing like I got to work there had the opportunity to work there for six years just being an absolute sponge on how the company worked how what insights was how it really drove strategy had a huge seat at the table uh And I I just learned a ton. 

Uh but after about six years of corporate, my entrepreneurial spirit spirit got to me and I was like, I got to get out of here and start a tech company because why not, you know, why not leave your nice corporate job and start a tech startup. Crazy. Um so that's when I kind of floated into into tech. So my it was a 10-year journey for me. I started a company called Knowledge Hound and it was really off of this pain point that I had at Proctor where we were spending tens of hundreds of millions of dollars every year in market research data and in particular survey data and we would get that survey data back and it would be unarchable and really it was kind of like stuck in these SPSS files or Excel files and so I was like why don't we create like a search driven analytic ICS platform that could easily search this and visualize it. And so that's what knowledge became. It became the first search driven analytics platform for survey data in our industry. And um that's when I met you, Lenny, at a IEX event, my first one. And I was like, what is this conference? I don't even know. And um and you were just so supportive and um really embracing the the startups that

Leonard Murphy: Yeah.

Kristi Zuhlke: Were entering the space. And so um knowledge hound was a 10-year journey for us. Uh we were acquired by Yuka about three oh maybe not probably like two two and a little bit of years ago. Um, and I have been obsessed with generative AI and I kind of combining this background of insights and serving the Fortune 1000 insights functions and being in one combining that with understanding generative AI and I'm always pushing the boundaries of where tech can take us and um, and so I've just been people have naturally been reaching out to me saying like, "Hey, can you help us think our through our AI strategy, etc." And uh and I've been observing like that they're just talking about it and everyone's like theoretically this or that. And I'm like you guys all you got to do is pick up turn your computer on that button. Just turn it on. Just sign up for a few things and start going at it.

Leonard Murphy: Yeah.

Kristi Zuhlke: But it's hard because everyone has a day job, you know? So I get it. Like it's not it's not as easy as just that. But, you know, it's um you got to do your J day job and then figure out time to to use AI. But yeah, it's been fun sharing with people what I'm

Leonard Murphy: Yeah. And I think I I love that, right?

Kristi Zuhlke: Learning.

Leonard Murphy: And you go on other social media platforms, you see everybody sharing here's the top 10 prompts for this and you know and all of that. But the difference that I think resonates at least with me with what you were doing of look, I'm I'm in your world, right? This is, you know, the world of insights. Um, and uh I get it and we're I'm just showing you how to just do stuff to try.

Kristi Zuhlke: Yeah.

Leonard Murphy: And and I remember vividly what where I went

Kristi Zuhlke: Yeah.

Leonard Murphy: From myself uh about about a year ago really um I was resistant to getting into the weeds on AI for a variety of reasons. Fundamentally was because I was afraid it was going to devalue myself, right? The my use cases because I'm I'm a consultant,

Kristi Zuhlke: Mhm.

Leonard Murphy: Right?

Kristi Zuhlke: Yeah.

Leonard Murphy: The um and circumstances forced me.

Kristi Zuhlke: Yeah. Yep.

Leonard Murphy: There actually there were three projects that all came in right at the same time and they were all due like a couple days and there was no way in hell uh I was going to turn out the money or that I could even deliver unless I use those tools, right? I was just that you know there was I was forced to embrace the

Kristi Zuhlke: Yeah.

Leonard Murphy: Efficiency initially and that was the unlock for me to first kind of recognize that there were ways to create real massive efficiencies in tasks just standard tasks.

Kristi Zuhlke: Yes. Yes.

Leonard Murphy: Um and that started on the journey of but wait now I can just do a whole lot more than that. I I can especially with the agentic you know component over the past few months like no I can make stuff and and we shared beforehand I've been building things over the past

Kristi Zuhlke: Yes.

Leonard Murphy: You know little while um which I would have never been able to do before I never you know for those listening right the things like working on GitHub and render and I'm not a tech guy but I am now right I mean,

Kristi Zuhlke: That's right.

Leonard Murphy: Yeah.

Kristi Zuhlke: Same with

Leonard Murphy: It it Yes. Right. And that and that's the thing.

Kristi Zuhlke: Me.

Leonard Murphy: I've just seen you just doing things and that that sounds devaluing. That's I don't mean it that way. It's not like you were just throwing stuff against the wall.

Kristi Zuhlke: No, no, no.

Leonard Murphy: You were taking real business issues and saying,

Kristi Zuhlke: Yeah.

Leonard Murphy: Look, here is a way to address that and create something new and leveraging these tools for it to become an

Kristi Zuhlke: Mhm.

Leonard Murphy: Augmentation, a superpower. So, uh, uh, anyway, it's not I'm putting words in your mouth, but that's been my perception of watching you. Uh, and is that what it's been like for

Kristi Zuhlke: Yeah. Yeah. I mean, I guess I didn't intentionally,

Leonard Murphy: You?

Kristi Zuhlke: It's helpful actually for you to put words in my mouth on this because I didn't necessarily set out to share with that intention. I was more so I really kind of I had a conversation with someone a mentor of mine in November of last year and and we were just talking about this industry and I was expressing my concern for it

Leonard Murphy: What

Kristi Zuhlke: And his encouragement was he's like Christie there's very few other people better than you to help

Leonard Murphy: The

Kristi Zuhlke: Us lead us through this change and why don't you lean into this and help us with this and I was that was kind of a wakeup call to me. I was like, "Oh, okay." So, instead of like running from it, maybe I should just lean into it. And so, I didn't have really any intention other than helping people um by sharing what I've been sharing. And and it's so it comes very naturally to me. Someone asked the other day like, "How do you come up with the topics? Do you ever feel like you're short on topics to publish on your Substack?" And I was like, "No, never." Like I actually feel like I've got this running list and I'm like which one do I pick?

Leonard Murphy: Yeah.

Kristi Zuhlke: Um and I I got feedback from some folks that I was posting too much like it was hard to keep up on my substack. So I actually paired back a little bit because I didn't want people to feel overwhelmed

Leonard Murphy: Yeah.

Kristi Zuhlke: Either. Um because I was like pumping out like three substacks a week and u people are like whoa I can't keep up with this. So I was like okay I'll do one to two a week so that people don't get too intimidated by it. But um you know it's just it comes very natural to me and what I'm the feedback that I'm getting is that I'm very practical and that's what I'm trying to be is like very it's not like this um highlevel theoretical like this could happen or this could be the way you could do this or or these big grandio statements of like oh you're going to 30% of jobs are going to be cut.

It's like show us how to use this. And I think that's what's really resonating with people. think like a lot of these headlines are very black or white and everyone wants to take this black or white approach. I think AI is very gray like where there is it's not perfect and it's but it's

Leonard Murphy: Yeah.

Kristi Zuhlke: Definitely not useless and so being in this gray area is very hard for people but it is that and it's like the supercharger but it's not perfect.

Leonard Murphy: Right. Yeah. I love that. And I love the the the helpful thing. Right. My kids used to say, "Dad, what do you do for a living?" And I help people. Um, and I think that that spirit of fundamentally of like I'm great. I I' I've found a way to make money from helping people and but that perspective leads us in different directions, right?

Kristi Zuhlke: Okay.

Leonard Murphy: I think that's throughout my career,

Kristi Zuhlke: Okay.

Leonard Murphy: I found that idea of I I want to help. I have an idea. I think is helpful. Um lean into that. Uh and then good things happen.

Kristi Zuhlke: Yep.

Leonard Murphy: Uh overall um yeah that's a and and I think I

Kristi Zuhlke: Yep.

Leonard Murphy: I experienced the same thing of once you do that and then have some discipline around it um of you know okay now let's build some structures and uh etc etc but it's only to enable the helpfulness. It's not to Yeah.

Kristi Zuhlke: That's

Leonard Murphy: Yeah.

Kristi Zuhlke: Right.

Leonard Murphy: That's a it's a cool uh I'm glad to have a kindred spirit in in thinking about things that way. Now, and that said, right now there is so much noise,

Kristi Zuhlke: Yeah.

Leonard Murphy: Right? And I think it if people if they're intimidated by a couple Substack posts a week,

Kristi Zuhlke: Yes.

Leonard Murphy: I mean, like, oh my god, I'm sure that you you're similar to be I'm watching stuff every day and like geez, I can't keep up and that's my job is to keep up with these things.

Kristi Zuhlke: Yes, I know that is. Yes, it is so hard to keep up.

Leonard Murphy: Um,

Kristi Zuhlke: It's so funny. I'll be in a meeting, someone's like, "Have you heard of this tool?" And I'm like, "Nope, never heard of that tool." Like, "Oh, well, it launched yesterday." I'm like, "Okay, well, okay, give me give me a month to catch up.

Leonard Murphy: Right, right, right.

Kristi Zuhlke: I don't

Leonard Murphy: Yeah. Right, right.

Kristi Zuhlke: Know.

Leonard Murphy: Absolutely. I was actually uh was talking to Steve Phillips from Zappy uh earlier today before this and the uh and we both think that we're about to see the greatest boom in entrepreneurs we've ever seen in history, right? There are going to be so many folks that are just creating new things. Um now market dynamics, are they all going to succeed?

Kristi Zuhlke: Yep.

Leonard Murphy: Nope, of course not. Some rise to the top and whatever. But I but I think that we're just going to see a massive explosion of creativity um

Kristi Zuhlke: Mhm.

Leonard Murphy: As people take ideas and now have the tools to be able to execute those ideas and then we get into whether or not it's you know successful or not. But um the and you can't keep up.

Kristi Zuhlke: Yep. Yep.

Leonard Murphy: That's why I increasingly pay attention to fundamentally to what's happening with the LLMs themselves, right? And the with the big tech companies and then try and reverse engineer from there.

Kristi Zuhlke: Mhm.

Leonard Murphy: Of like, okay, that's what their latest product does. Now, here's what that probably means for, you know, our specific sphere of influence and uh and thinking,

Kristi Zuhlke: Yes.

Leonard Murphy: But even then, it's every week. It's just not a, you

Kristi Zuhlke: It's a Yeah. Yeah. It's really hard.

Leonard Murphy: Know,

Kristi Zuhlke: And I think too it's I think that is the problem is that um it's some people want are just trying to you're like, "Oh, I'll just wait till it all settles down and the hype is over and we let everyone work it out." And unfortunately, I think in this technology cycle is very different than other ones. I think we could have said that in blockchain or um crypto is an interesting one but um you know there's the in other tech booms or like when we

Leonard Murphy: Yeah.

Kristi Zuhlke: Went to online surveys,

Leonard Murphy: Yeah.

Kristi Zuhlke: We went to mobile surveys like we could kind of we could let the leaders go and

Leonard Murphy: Right.

Kristi Zuhlke: Figure it out and then be fast followers or whatever. And unfortunately in this situation,

Leonard Murphy: All

Kristi Zuhlke: We can't because you're gonna see too big of a revenue for

Leonard Murphy: Right.

Kristi Zuhlke: Market research suppliers. You're going to see too big of a revenue hit, I think, in the short term to if you don't keep up with it.

Leonard Murphy: Yep.

Kristi Zuhlke: Um, and then for brands, you're your competitors are just I mean, I'm seeing it. It's amazing what I'm seeing of the brands that are ahead. they are really ahead in in their thinking on this and if that's

Leonard Murphy: Yeah.

Kristi Zuhlke: Your competitor you now you are really far behind. Um so I don't think you can just let this cycle settle down because I'm not sure

Leonard Murphy: Yeah.

Kristi Zuhlke: It's going to settle down in some ways whether that's a valuation correction or what whatnot but it's not going to settle down from a the technologies going to go boom and then bust for a bit and then come back. It's here.

Leonard Murphy: A agreed.

Kristi Zuhlke: It's the train is

Leonard Murphy: Agreed. And on the same note,

Kristi Zuhlke: Going.

Leonard Murphy: The uh you know, a lot of folks uh want to standardize on one specific uh model and like no the uh the this I mean that

Kristi Zuhlke: No, no.

Leonard Murphy: That war is far from being won. Um and you have to experiment with all of them.

Kristi Zuhlke: Yeah.

Leonard Murphy: Um and and that takes a different approach.

Kristi Zuhlke: Mhm.

Leonard Murphy: So personally I found the I use perplexity as my orchestration layer because you can use multiple models within perplexity.

Kristi Zuhlke: Yeah. Yeah. Okay.

Leonard Murphy: So it has that all so uh because I was trying to manually I would just test the

Kristi Zuhlke: Yeah. Yeah.

Leonard Murphy: Same you know thing against multiple models. Well that was a pain in the ass right?

Kristi Zuhlke: Yeah.

Leonard Murphy: So uh so instead the uh have shifted towards okay I'm going to

Kristi Zuhlke: Yes.

Leonard Murphy: Use perplexity I can access any of the major models. Um it's my orchestration layer. It's everything's in one place, but I can continue to to assess that I mean obviously Claude builds stuff really well, but Grock is right behind it. um the you know chat's good for for research blah blah blah blah blah right the but to settle on one system um I think is a mistake and that's a real challenge for businesses because we want to

Kristi Zuhlke: Yep.

Leonard Murphy: Standardize you know we want to have one we're we're a Microsoft office we use you know whatever um uh but that that's limiting

Kristi Zuhlke: Yes. Yes.

Leonard Murphy: Because we are just in a different world and the the leaprogging effect uh is unlike anything I've ever seen before uh from a capability standpoint.

Kristi Zuhlke: Yeah.

Leonard Murphy: Like literally this weekend uh uh Grock 4.3 came out and uh beta and uh I used it to build a deck um to see

Kristi Zuhlke: Okay.

Leonard Murphy: As do it.

Kristi Zuhlke: How'd it

Leonard Murphy: It it was flawless from the get-go.

Kristi Zuhlke: Go?

Leonard Murphy: It created a flawless editable PowerPoint deck out of the gate with exactly my specifications. better than better than Claw did, better than uh Notebook LM. Uh uh it was perfect. Uh like the like,

Kristi Zuhlke: Well,

Leonard Murphy: Okay, well, I guess I'm going to use Grock for this now because this is

Kristi Zuhlke: There you go. Yeah, that's the thing. It's like Yeah, exactly. Like I think that's it.

Leonard Murphy: Better.

Kristi Zuhlke: You're exactly right because Okay, last week you would have used I I use Cloud a ton for my presentations, but I used the Gamma plugin, the API into Gamma um because it just makes it it makes decks really well.

Leonard Murphy: Okay.

Kristi Zuhlke: Um but here you go. last week I was doing that, maybe this week I'm going to be on Grock. You know, you just have to be flexible. But that's I think the reality and a lot of people live in is that their companies are um to specific tools. And so that's why I see a lot with the brands that I'm working with is that their IT department has locked it all

Leonard Murphy: Yeah.

Kristi Zuhlke: Down.

Leonard Murphy: Yeah.

Kristi Zuhlke: And it's like all you get is Microsoft Copilot.

Leonard Murphy: Right. Right.

Kristi Zuhlke: And and now you're just like, "Okay, well, I don't blame them for being like throwing your hands up and being like,"Well, I can't, you know, I can't do any of this because I'm handcuffed." I think for those people,

Leonard Murphy: Right.

Kristi Zuhlke: I really encourage them to then start experimenting on your personal stuff so that when the moment comes that, you know,

Leonard Murphy: Yep.

Kristi Zuhlke: You get laid off and you need to go to a new brand that uses any kind of tool, you already know those skills. um that you can use those tools and then be able to come to it and be like, listen, this is what we really need. Um a lot of the work I've been doing has been doing exactly that where we're mapping out business use cases um and then identifying the AI that that can enhance that business um or basically decision-

Leonard Murphy: Mhm. Right.

Kristi Zuhlke: Making that can enhance that decision process um with AI. So, you know, it's but you have to have a business case for it, which is understandable. Uh, but yeah,

Leonard Murphy: 100%

Kristi Zuhlke: A lot of these guys are h handcuffed because their corporations have have done it that way and that's really tough. It's hard to operate in that

Leonard Murphy: Yeah. Yeah. I from a a uh brand side,

Kristi Zuhlke: Environment.

Leonard Murphy: Corporate research standpoint, I I think those challenges are certainly significant. The uh and even with on the supplier side, you and I, we we're both spoken to some of the same suppliers recently. um the uh um and they see it as a good,

Kristi Zuhlke: Okay.

Leonard Murphy: You know, uh peanut butter and chocolate uh scenario to have conversations both of us.

Kristi Zuhlke: Yeah, I'm back. Yeah, that's great. They

Leonard Murphy: The uh yeah but anyway point is and they're not they're they're

Kristi Zuhlke: Said

Leonard Murphy: Not as restricted but they're large enough that they want to have a corporate policy um where the you know the earlier stage companies uh you know we obviously you know both experienced that that the you know you've got much more uh nibbleness and agility to be able to adapt but even if you're building a product now off of one platform, you still get locked in. So there's a lot of navigating these tradeoffs are is challenging across the board. And if that's a fear from an adoption standpoint,

Kristi Zuhlke: Yeah.

Leonard Murphy: It just is where it's just it's where we

Kristi Zuhlke: Yeah. Don't let it hold you back, you know, like you just got to lean into it,

Leonard Murphy: Are.

Kristi Zuhlke: Right? You can't let it hold you back from moving forward because um if it then you're just going to end up in a worse spot, honestly.

Leonard Murphy: Right. So I wonder um so you're you're a mom and the uh I often

Kristi Zuhlke: Mhm.

Leonard Murphy: Think that the core skills of agility, flexibility, um you know uh adaptability, all the you know abilities uh that those are are necessary for success as a parent.

Kristi Zuhlke: Mhm.

Leonard Murphy: Um uh skills and that it actually empowers us maybe in a different

Kristi Zuhlke: Mhm.

Leonard Murphy: Way. um from a professional standpoint to like take those things you you just got to learn how to deal with that as a parent. So to have you experienced that?

Kristi Zuhlke: Yep. Yep.

Leonard Murphy: Has that been uh kind of an

Kristi Zuhlke: Yeah. You know, I don't know. Your kids are older than mine,

Leonard Murphy: enhance?

Kristi Zuhlke: but I think like at six and five of Yeah. They're so um they're requiring me to use different muscles than I've ever had to use

Leonard Murphy: Yeah.

Kristi Zuhlke: Before of um patience and understanding and empathy.

Leonard Murphy: Yeah.

Kristi Zuhlke: Um you know just like trying to be like why is there a meltdown happening right now? I just do not understand um but adaptability certainly

Leonard Murphy: Right.

Kristi Zuhlke: And yeah I think it's similar to this moment that we're in is that you just have to remain flexible and kind of observe what's going on and not let that impact you. It's kind of like, okay, you can look, you can watch, you can and then to be thoughtful and then make a decision versus just kind of having a reaction to it, a reaction of like, oh, this is not happening or that's not valid. You know, like this temper tantrum is not valid at all. Um, or this technology is not valid.

Leonard Murphy: Yeah.

Kristi Zuhlke: That's it's not necessarily always the case. It's like, oh, actually, um, you know, his socks were on backwards and he didn't like the way that felt. And it's like this temper tantrum is valid in his mind.

Leonard Murphy: Yeah.

Kristi Zuhlke: Now I don't know if I'd react that way. Um but you know someone created this technology um for a reason.

Leonard Murphy: Right.

Kristi Zuhlke: They felt a pain point and you know so I think it is a lot of um it's very applicable and some ways I think parenting is even harder though than than professional

Leonard Murphy: Oh yes.

Kristi Zuhlke: Stuff.

Leonard Murphy: And you know Yes. Your kids are younger. mind. Yeah. Wait, wait, wait till there's teenagers. So, this thing Chris, do you ever remember what you were like as a teenager?

Kristi Zuhlke: Oh, I don't I I I'm already saying prayers for myself as a parent knowing what I was

Leonard Murphy: And

Kristi Zuhlke: Like.

Leonard Murphy: There's a reason 13 is considered an unlucky number and I think it's because that's when kids lose their minds. Um the uh um Yeah. So, it just is.

Kristi Zuhlke: Great.

Leonard Murphy: It gets better. They It gets better. um you know 15 16 it's better but boy 12 the 12 12 13 14 especially for daughters I mean it's just experience I have a lot of daughters I've been through this

Kristi Zuhlke: Yeah. Uhhuh. Uhhuh. Uhhuh. Uhhuh.

Leonard Murphy: Uh and it's like like clockwork um to hit that age and it's just like oh my god it's so

Kristi Zuhlke: Yeah.

Leonard Murphy: I'm glad that you're still we bonded over your cuteness as a when you were a baby because one of us may not survive this right I don't know

Kristi Zuhlke: Hold on to that.

Leonard Murphy: We'll But anyway, um so where are you now? Right. It seems seems like you've kind of fallen into uh kind of a consulting type of role. Um uh although my guess is that that's probably, you know, there's probably something else cooking in the background is my guess. Uh in terms of the future, what do you what do you see happening for you as we go forward from here?

Kristi Zuhlke: Yeah. So, I'm I think the road forward is still unclear. I'm allowing myself to kind of go with the wind right now, which is very unlike me. Um, but I'm just kind of letting the market take me where it needs me. And so that means for me I've been most of my time is spent 70% of the time working with brands on how to redo their AI strategy uh for their insights function and then 30% of my time is spent for with market research suppliers really thinking through how they integrate AI into their operations and then also how they talk about it. You know, how do they how do you stay true to the roots that you are with which is meth methodology rigor and um and talking to consumers and all that, but also embracing AI. So that balance is really tricky. Um so that's been really fun work. I've I have become obsessed with, if you couldn't already tell, um with um based on my substack, with the promise of simulating consumers with LLMs and it's been so intriguing to me and I have um some connections to some

Leonard Murphy: Yep.

Kristi Zuhlke: Guys who are doing this at um and they're doing it at very sophisticated levels. Um, and so they've given me this ability to understand at a deeply technical level on how this is actually happening. And that's what I kind of needed because I think there was a ton there's a lot out there

Leonard Murphy: Okay.

Kristi Zuhlke: That's very much like marketing material and um, and there's so many different use cases for it as well. I was really I'm really fascinated by it as to like how do you instead of having to do quant research, how do you how do you use an LLM's power to u come up with business answers and decisions with using that as a methodology. And so that's absolutely been so fascinating for me. And so I'm going to see how far we can push the envelope on that and we'll see where it goes. But, you know, I I wouldn't rule it out by the, you know, end of the year that I've I've leaned all the way into that. But, we'll see. I'm kind of just still figuring out that and and trying to share with everyone what I'm learning, too,

Leonard Murphy: Yep.

Kristi Zuhlke: Because I don't think that is either a black or white um uh solution either. I think there's but I'm going to try to I'm trying my attempt this year is going to

Leonard Murphy: Yeah.

Kristi Zuhlke: Share as much as I can to help people see a little bit more black and white or make decisions around it. uh because I there's a lot there that can be uncovered and I do think there's a lot of also like any big boom there's a lot of people in the market talking about things that a maybe some of them don't really have a great

Leonard Murphy: Heat.

Kristi Zuhlke: Methodology but how do we assess that and how do we dig in deeper is something I've been very curious

Leonard Murphy: Yeah, I agree with you 100%.

Kristi Zuhlke: Out.

Leonard Murphy: Uh actually right before we started was on a call with a a private equity group on a potential acquisition of a company that would you know has has a panel asset and the uh and the conversation was look the existing model is that

Kristi Zuhlke: Mhm.

Leonard Murphy: That that's fine but the you know that's going to go the frequency of asking questions is going to decrease. So uh as more people build out models based on

Kristi Zuhlke: Mhm.

Leonard Murphy: Synthetic which is just you have a data graph and you know there's lots of information and you can model that out effectively and we've been doing that for years been talking about it forever. The AI unlocks that in a different way.

Kristi Zuhlke: Mhm.

Leonard Murphy: Um and that's always where we were headed.

Kristi Zuhlke: Well,

Leonard Murphy: Uh but the value of the data is I I think that's sophistic that thought of the I I

Kristi Zuhlke: Yeah.

Leonard Murphy: Don't see a world where the need to have a steady input of consumer information to feed the models ever goes away. Um and so we thought about synthetic it's almost an interface issue. So right it's it's how we access the information.

Kristi Zuhlke: Kind of. So,

Leonard Murphy: Go ahead.

Kristi Zuhlke: I'm going to challenge you on that a little bit.

Leonard Murphy: All

Kristi Zuhlke: So, the work that I've been doing actually can simulate humans with no

Leonard Murphy: Right.

Kristi Zuhlke: Historical data.

Leonard Murphy: Okay. All right.

Kristi Zuhlke: So,

Leonard Murphy: Well,

Kristi Zuhlke: It's pretty

Leonard Murphy: How how the hell do you do that,

Kristi Zuhlke: Wild.

Leonard Murphy: Christie? So, I I have a hard time conceptualizing

Kristi Zuhlke: Well,

Leonard Murphy: That.

Kristi Zuhlke: I know it's very hard. It's taken me 18 months to kind of get my head wrapped around this. Um, and again, it's not black and white as to like when I say like no historical data, but essentially kind of I think there's two things, you know, two things. There's two ways of going about this and I don't think most I think there's one way many people are going which has some flaws and a better way to go about it which still has I think it everything's perfect so I don't want to like so but it has fewer it has a better way to potentially to work through some of the shortcomings of it but essentially kind of what a lot of people are doing today is that with stimulating humans which is same thing is synthetic. I just don't like the description of it. Um,

Leonard Murphy: Yeah.

Kristi Zuhlke: But they're they're taking they're kind of using cluster analysis essentially. They're kind of going like because these 10 things these 10 people acted this way in the past, therefore this they're going to act this way in the future.

Leonard Murphy: Okay,

Kristi Zuhlke: That's what you know using historical data would requ have you do. Instead, what the guys that I'm working with,

Leonard Murphy: Come.

Kristi Zuhlke: What we're doing is we're actually creating um like neural networks that act like a

Leonard Murphy: Okay.

Kristi Zuhlke: Human that are we're basically using the math of an LLM. So,

Leonard Murphy: But aren't they still trained?

Kristi Zuhlke: We're using the math of an

Leonard Murphy: Sorry. Go ahead, please.

Kristi Zuhlke: Pardon

Leonard Murphy: I I had a question. Isn't it still to train the neural network? Still requires data though, right? the uh where's the math

Kristi Zuhlke: Yes, it used Fair enough.

Leonard Murphy: Just

Kristi Zuhlke: Fair enough. It's um but it's the it's you're more so using the math that um the math of psychology and how a brain works and then

Leonard Murphy: Okay I get that so it's a framework

Kristi Zuhlke: How brains make and how brains make decisions

Leonard Murphy: That a standard framework of decision making so the inputs don't

Kristi Zuhlke: And

Leonard Murphy: Really that that It's not based on the input. It's based on the framework. Is

Kristi Zuhlke: Yes yeah that's a good way of put it and framework Mark the word framework I think is a good consulting

Leonard Murphy: That

Kristi Zuhlke: Word. I think that it's a little bit it's quite deeper than that and I and it's it's more about but I know exactly. Yeah. So yes,

Leonard Murphy: Sure?

Kristi Zuhlke: I think that's a good analogous word of like um that the um you're essentially kind of modeling a human brain and how it might act based on a very specific persona. and and um and so they're and what I'm finding is that AI personas are actually probably more actionable a number of AI personas are more actionable than digital twins. Um, but again, this could turn into like a three-hour long conversation, so I won't even go down that rabbit hole.

Leonard Murphy: What

Kristi Zuhlke: But the um but it's amazing what you can do when you kind of take

Leonard Murphy: The

Kristi Zuhlke: This approach and say actually I'm going to try to model a human brain and not with math and

Leonard Murphy: Heat?

Kristi Zuhlke: How that brain actually crunches information and makes decisions versus hey let me look at a bunch of historical data and go because people like this act in the past they're for going to act like this and um and it's just a fundamentally different way of looking at

Leonard Murphy: Yep.

Kristi Zuhlke: It And I think there's a lot of people sitting in this bucket of kind of what I'd call cluster analysis. And you're right, we have been doing that for years. And it can get us pretty good results. It can get us like 80% accuracy. But if you want like 95% plus accuracy, you almost have to move yourself into this bucket of how do I actually create a human brain out of math.

Leonard Murphy: I get that they and so I'm I'm glad you pushed pushed back. I was making a generalized statement and now that you've you know talked through that then I I would say I I agree and I'm aware of some companies that are doing comparably similar things right um that are more uh more foundational to kind of behavioral science and and psychology and

Kristi Zuhlke: Uhhuh. Uhhuh.

Leonard Murphy: And you know that's the the models um that they built around standard

Kristi Zuhlke: Mhm.

Leonard Murphy: Personas so to speak right the uh uh I see I see those companies incorporating getting uh new attitudinal or behavioral data as well to help give it more ammunition. Um but primarily just to drive the simulations overall.

Kristi Zuhlke: Mhm. Mhm. Mhm.

Leonard Murphy: Um the uh uh but I I I agree and I also see did you

Kristi Zuhlke: Mhm.

Leonard Murphy: See Steve Phillips uh launched his CEO friend platform the uh that's what earlier.

Kristi Zuhlke: It didn't.

Leonard Murphy: Okay. So, so you know Steve from Zappy and Steve uh took the the idea if he wanted to build a

Kristi Zuhlke: Uhhuh.

Leonard Murphy: Virtual uh virtual advisory board for startups.

Kristi Zuhlke: Okay.

Leonard Murphy: So same things built personas of C CFO

Kristi Zuhlke: Yeah. Yeah.

Leonard Murphy: Investor um you know the CMO etc etc. So uh makes that available as basically a licensing platform uh very inexpensive. or startups can access the you know these developed personas of very specific uh functions within uh within a

Kristi Zuhlke: Mhm.

Leonard Murphy: Business to help give them access to information and and you know that you would get if you have an advisory board out of the gate and I think that it's it's all variations on a theme right think about the this idea of I don't like term synthetic either I think it's it's u we need to figure out a better a better approach um But uh oh, and I'm sorry if you're hearing the drill. That's probably we probably need to wrap up in a second.

Kristi Zuhlke: Can't hear it.

Leonard Murphy: We we've had a saga with our our dishwasher being replaced and uh

Kristi Zuhlke: That's the trigger. Oh,

Leonard Murphy: And that's finally happening and yes,

Kristi Zuhlke: Exciting.

Leonard Murphy: Yes. Um anyway, lots more conversation to be had. Where do you see things uh happening for for you uh in the the very short run? Right. You still open to people reaching out and saying,

Kristi Zuhlke: Yeah.

Leonard Murphy: You know, hey,

Kristi Zuhlke: Yeah.

Leonard Murphy: Can we talk and engaging?

Kristi Zuhlke: Yep.

Leonard Murphy: Yeah.

Kristi Zuhlke: Yep. I'm I'm basically like taking anyone's call who wants help on AI in their strategy or

Leonard Murphy: Okay.

Kristi Zuhlke: Implementation. Um, I'm pretty honest, as you know, Lenny, about like what I can and can't do and you know, like I'm if I get something come across my desk that I don't feel like it's a good fit, I'm like, "No." But I'm also letting people I'm letting this market kind of take me where the need is for the first time. Whereas usually I'm used to just being like, "Here's my vision.

Leonard Murphy: Yeah.

Kristi Zuhlke: Let's" and I push it on everyone else because I'm like, "This is the right thing." And this time I'm taking a little bit different approach and being like, "Okay, where do you guys need my help?" and I'm here to help and be helpful. So, um yes, so anyone can reach out and I'm happy to jam on what's helpful. And then also, of course, my substack is free. So, um you can I do gate some some content um on a paid subscription, but for the most part, most of my most my content is free. So, anyone can also find resources there.

Leonard Murphy: Very cool. And isn't Substack great? The uh I love Substack.

Kristi Zuhlke: It's great.

Leonard Murphy: The talk about it.

Kristi Zuhlke: That's great.

Leonard Murphy: They need they need to change the business model though. The I subscribe to like a hundred Substacks.

Kristi Zuhlke: Oh my gosh.

Leonard Murphy: I only pay for like four. Um, right.

Kristi Zuhlke: Okay.

Leonard Murphy: And there there's a whole other conversation I really thought about reaching out to the Substack folks saying, "Look, you could give people a path to monetize audiences that takes away the

Kristi Zuhlke: Agreed.

Leonard Murphy: Subscription thing because it's a real barrier for, you know, uh, for a lot of folks. But anyway, that's a whole other conversation. If you're listening, Substack, reach out. Um, I have an idea." Um, but I love it. I I love Substack so in so many ways. Anyway, uh, Christie, thank you. This has been fantastic. It was good catching up.

Kristi Zuhlke: Yes.

Leonard Murphy: Um, uh, audience, if you are not following Christie on LinkedIn or on Substack, go do that. Um, the, yeah, thoughtful, pragmatic, um, real world. Uh, there's no fluff in what you do and, uh, and really I appreciate that you just kind of get right to the point of, you know, here's the deal. And, uh, uh, I I love it. So, if people aren't following you, they should be.

Kristi Zuhlke: Appreciate you, Lenny. Thanks so much.

Leonard Murphy: Thanks, Christie. All right, that's it for this edition of the CEO series. Uh, thank you, Christie. Thank you to our producers, sponsors, etc., etc. And we will be back soon. Bye-bye.