AI Made Simple: How to Make AI Approachable and Avoid Costly Mistakes with Kathleen Perley of Rice University | Behind the Numbers

Summary

Behind the Numbers: AI Made Simple – How to Make AI Approachable and Avoid Costly Mistakes

Podcast Information:

• Title: Behind the Numbers: an EMARKETER Podcast
• Host/Author: EMARKETER
• Episode: AI Made Simple: How to Make AI Approachable and Avoid Costly Mistakes with Kathleen Purley of Rice University
• Release Date: July 11, 2025

Hosts and Guests:

• Marcus: Host of the podcast.
• Gajo Sevilla: Senior Analyst covering AI and technology at EMARKETER.
• Kathleen Purley: Professor and AI advisor to the Dean's at Rice Business School, founder and CEO of AI company Demystify.

---

Introduction and Guest Background

The episode opens with Marcus introducing the guests, Gajo Sevilla and Kathleen Purley. Gajo is recognized for his expertise in AI and technology, while Kathleen brings a unique blend of linguistics, entrepreneurship, and academia to the conversation. Kathleen describes herself as "a linguist turned entrepreneur turned academic, focusing on making AI simple and real" ([01:36]). This diverse background positions her well to discuss the practical adoption of AI in business settings.

---

Making AI Approachable

Kathleen Purley emphasizes the importance of demystifying AI to make it approachable for businesses and individuals. Drawing from her personal experience with dyslexia, she understands the necessity of simplifying complex concepts. She uses relatable analogies, such as comparing AI to driving a car:

“If I had to manipulate a car engine to get from my house to campus every day, I will tell you, most of my students don't think I would make it right. But we have a steering wheel, we have a gas pedal, we have gears that we can change based on kind of the road conditions.” ([06:29])

This analogy illustrates how AI can be interacted with through user-friendly interfaces, allowing users to guide AI without needing to understand its intricate workings.

Gajo Sevilla concurs, highlighting the necessity of intuitive user interfaces:

“Maybe demonstrating in simple terms that you choose this model, it could do A, B and C really well...” ([09:56])

Both experts agree that simplifying AI interfaces and providing clear, relatable examples are crucial for broader adoption.

---

Bridging the Knowledge Gap

The discussion shifts to the rapid advancement of AI technology and the widening knowledge gap among users. Gajo points out the relentless pace of innovation and the importance of refining user interfaces to keep up:

“It's the user interface that’s really the only mode of control users have.” ([09:56])

Kathleen adds that understanding AI basics is essential. She uses various analogies, such as comparing AI architectures to medical specialties and cocktail making, to help non-experts grasp foundational concepts:

“Understanding how, like the cause if A, then B, then C, and how all the levers are related are key.” ([10:37])

These explanations help bridge the gap between technological advancements and user comprehension, making AI more accessible.

---

Strategic AI Adoption in Organizations

A significant portion of the conversation centers on how organizations can strategically adopt AI. Kathleen poses critical questions for businesses considering AI integration:

“Are you willing to make a shift or a change? Is this going beyond just experimentation?” ([12:28])

She argues that merely incorporating AI into existing processes without rethinking organizational strategies limits its potential. Instead, businesses should consider how AI can fundamentally transform their operations and offerings.

Gajo further elaborates on the need for businesses to define clear objectives before adopting AI:

“Are they improving processes? Can they use the tools judiciously? And I don't think they spend maybe enough time evaluating before they jump in.” ([14:14])

This strategic approach helps prevent common pitfalls and ensures that AI implementations deliver meaningful value.

---

Leadership and CEO Involvement

The role of leadership, particularly CEOs, is highlighted as a critical factor in successful AI adoption. Kathleen stresses that AI should be seen as a strategic initiative rather than a delegated technical project:

“AI is a strategic play, not a tech play. And I think... CEOs start getting really involved and not see this as a delegated down initiative.” ([17:15])

She cites a McKinsey study, noting that only about 28% of CEOs are actively involved in their organization's AI strategy. This lack of engagement often results in low ROI from AI initiatives, as successful implementations typically require visionary leadership.

Kathleen shares an example of effective leadership with Marie Myers at Hewlett Packard Enterprises, who actively engages in AI strategy development:

“She was like, no, no, I want you to show me how to build it.” ([20:53])

Such hands-on involvement ensures that AI projects align with the organization's long-term goals and deliver sustained value.

---

Insights from Kathleen's Book

Kathleen Purley's latest book, AI Made Simple Results Made Real: An Executive's Guide to Partnering with the Future, provides further insights into AI adoption. One notable chapter explores whether an organization is ready for AI, using an analogy similar to parenting:

“Are you ever really ready for kids? Like, do you have a car that can put a car seat in it? Maybe that's like... but you're never really ready.” ([21:25])

This perspective underscores that while perfect readiness is unattainable, foundational elements such as data governance and strategic vision are essential for effective AI integration.

Another key insight from her book is the profile of successful AI leaders:

“Leaders who are successful within organizations oftentimes don't have an IT background... They have played in the digital marketing space.” ([23:02])

These leaders bring a blend of technical understanding and strategic thinking, enabling them to leverage AI effectively without being bogged down by over-engineering.

---

Common Mistakes in AI Adoption

Both Kathleen and Gajo identify several common mistakes organizations make when adopting AI:

1. Lack of Specific Testing and Optimization:

   • Gajo: “None of these models are really off-the-shelf solutions. They’re pretty raw. You need to forge them and see where they fit within your stack.” ([26:18])

2. Inconsistent Training:

   • Gajo: “People need consistent training on how to use them since they're constantly changing, they're constantly improving.” ([26:18])

3. One-Size-Fits-All Training Programs:

   • Kathleen: “They'll want you to do a four-hour AI intensive literacy... but someone in marketing is going to leverage AI very different than your data scientist.” ([27:28])

4. Neglecting Task-Level Integration:

   • Kathleen: “Breaking that job into tasks... where can AI be supplemented in?” ([28:00])

Avoiding these pitfalls requires tailored approaches to AI implementation, continuous education, and a clear understanding of organizational needs.

---

Final Recommendations

Kathleen concludes with actionable advice for organizations embarking on their AI journey:

• Tailored Training: Develop role-specific training programs that address the unique ways different employees can leverage AI.
• Task-Oriented Integration: Identify specific tasks within roles that AI can enhance, rather than expecting AI to perform entire jobs.
• Leadership Engagement: Ensure that CEOs and senior leaders are actively involved in shaping and guiding AI strategies.
• Strategic Vision: Move beyond mere experimentation and integrate AI into the core strategic vision of the organization to unlock its full potential.

Gajo adds that ongoing testing, optimization, and training are essential to adapt to the rapidly evolving AI landscape.

---

Conclusion

The episode of Behind the Numbers provides a comprehensive exploration of making AI approachable and strategically integrating it into organizations. With insights from Kathleen Purley and Gajo Sevilla, listeners gain valuable perspectives on overcoming barriers to AI adoption, the critical role of leadership, and avoiding common implementation mistakes. By following their guidance, marketers, retailers, and advertisers can navigate the complexities of AI to drive efficiency, innovation, and sustained competitive advantage.

Notable Quotes:

• Kathleen Purley: “AI is a strategic play, not a tech play.” ([17:15])
• Gajo Sevilla: “None of these models are really off-the-shelf solutions.” ([26:18])
• Kathleen Purley: “Breaking that job into tasks... where can AI be supplemented in?” ([28:00])

---

Tune In Next Week: Join Marcus and the team as they discuss “How Weight Loss Drugs Became Popular and the Spaces They Are Reshaping.” Subscribe and follow Behind the Numbers wherever you find podcasts. Leave a rating and review to support the show. Have a happy weekend!

Transcript

[Marcus]

In marketing, everything must work seamlessly or efficiency, speed, and ROI all suffer. That's why Quad is obsessed, maybe too much, with making sure your marketing machine runs smoothly with less friction and smarter integration. Better marketing is built on Quad. See how better gets [e-mail] BuildBetter Foreign hey, gang. It's Friday, July 11th. Kathleen Garjo and listeners, welcome to behind the Numbers, an E marketer video podcast made possible by Quad. I'm Marcus, and joining me for today's conversation, we have two people. Let's meet them. We start with our senior analyst covering everything AI and technology. Based in New York is Gajo Sevilla.

[Gajo Sevilla]

Hey, thanks for having me back.

[Marcus]

Yes, sir, of course, of course. Kathleen, it sounds like we banned him for a while. Okay, that didn't happen. He's always been. Welcome, Gajo. Welcome back to the show. Professor and AI advisor to the Dean's Rice Business School and founder and CEO of AI company Demystify, based in Texas, is Kathleen Purley joining the podcast. Welcome.

[Kathleen Purley]

Thank you so much for having me. Excited to be here.

[Marcus]

Of course, of course, of course. Thank you for joining us. We start with a speed intro anytime we have an external guest on to get to know them a little bit better. So around 60 seconds on the clock, which we pay no attention to. Let's do it. First question. What do you do? In a sentence.

[Kathleen Purley]

So I'm a linguist turned entrepreneur turned academic, and now I focus on making AI simple and real and helping executives and entrepreneurs really navigate this new world and using it to solve problems not as a magic wand, but really rethinking their strategy.

[Marcus]

Okay, you said you're a linguist turned entrepreneur turned ac, and I thought you were gonna say actor, and I was like, oh, no, who are we gonna ask for the show? These are all wrong questions. All right, academic. Thank God. If. Second question. If you could master one skill overnight, what would it be?

[Kathleen Purley]

So I actually thought about this long and hard, and I wanted to say something like, really profound, like the art of storytelling or something like, you know, I will tell you, it's actually doing hair.

[Marcus]

Okay.

[Kathleen Purley]

Your own, mine, my daughter's hair, any hair. Like, the thought of having to use a, like, blow dryer in one hand and a hairbrush in the other hand at the same time is like a foreign language to me. It's physics. I. I went to an all girls Catholic school, K through 12, so it was just like, not needed. And I never developed the skill. Went to college, literally would iron my hair, like, with an ironing board. And now I have a girl who has beautiful curly hair, and I still. I'm, like, at a loss.

[Marcus]

Crew cuts, I think, is what you're looking for. Send them into the military. Yeah, they'll be fine. Gajo, what about you?

[Gajo Sevilla]

Well, for me, I think I've always wanted to be an auto mechanic or a watchmaker, so something that has to do with working with your hands, something AI isn't going to replace anytime soon. So. Yeah. Something analog and just tools and wrenches.

[Marcus]

Why? Was someone in your family that.

[Gajo Sevilla]

No, no, no one was. But then I've always felt that it's an extra skill, really, that more people should be able to do to some extent.

[Marcus]

Yeah.

[Gajo Sevilla]

I mean, they never taught it to us at school, not even as a basic course.

[Kathleen Purley]

So did you change your oil or your tire?

[Gajo Sevilla]

Tires, yes.

[Kathleen Purley]

Oil.

[Gajo Sevilla]

Depends. Could do it, but, you know, I live in an apartment, so I need to find a place to be able to do that. But, yeah, I mean, there's that limitation. Right. You're like, I wish I could fix this. Right.

[Marcus]

So, yeah, that might just be you guys. I've got renting a car right now based on where I live and the oil change. Oil light came on. I took it back and asked him for another one.

[Gajo Sevilla]

That's one way to do it.

[Marcus]

Yeah.

[Gajo Sevilla]

Change the whole car. Yeah, yeah.

[Marcus]

The woman was not impressed.

[Kathleen Purley]

I'm not a pro, by any means, but after I, like, had somewhat of a crisis, I actually got a 1972 international scout that my husband and I have been fixing up in our garage.

[Marcus]

Oh.

[Kathleen Purley]

So we replaced the wiring harness, We've done transmission fluid, replaced the seats.

[Marcus]

What kind of a crisis led you to buy a classic car?

[Kathleen Purley]

It was kind of one of those same things I had just sold and exited my company. I was just gonna be a professor, and so I was like, oh, man, I'm gonna be the coolest professor. If I roll up in, like, a 1972 purple, like, Internet scout that's decked out like, this is so fun. And it's been fine. It is. It's been a. A money pit, but it's been fun.

[Marcus]

Oh, okay. So you meant just like a life, Like a moment crisis?

[Kathleen Purley]

Yeah, it was more of those things, like, what am I going to do with myself?

[Marcus]

Okay. Not like, ah, the house is burned down. Let's buy a classic car.

[Gajo Sevilla]

But by the time you're done, you'll be an expert on that as well.

[Marcus]

Yeah. Nice to get to know, I guess, a little bit better. For today's episode. Today's real topic, how to approach AI The Right way. So talking AI again. And I want to start with this question, because when me and Kathleen were speaking before this episode a week or two ago, one of the ways that she had described adopting AI was making it more approachable, which I thought was an interesting way, because I haven't heard. I've heard AI adoption a lot, but not used. Not had the word approachable use much when it comes to artificial intelligence. So, Kathleen, how do you make AI approachable?

[Kathleen Purley]

I think, you know, so I think we talked about this a little bit, Marcus. So growing up, I had dyslexia, right? So, like, everything was like complete noise to me. And I always had to figure out what was the pattern and identify that and develop analogies to help me understand things. And so I've kind of taken that approach, and it's really about, like, stripping away the smoke and mirrors so that people can start to feel comfortable and understanding it, because I think that is probably the biggest intimidation factor. And since we've been talking about cars, I'll kind of use the car analogy. I'm born and raised in Houston, Texas, so I come from a long line of actually New York transplants to Texas, and oil and gas sector. And so when I told my parents I wanted to be a linguist, they were not super happy because they were like, your options are banking or oil and gas. And, you know, it was November 22nd, and I finally realized. We had a family dinner. My dad turned to me and was like, have you heard this, like, chat GPT something? And I was like. I pushed my sister, who was like, wharton, mba, oil and gas, private equity. I was like. I was like, my time has arrived. We had Alien, that movie that came out where they brought in the linguist. And now I am somewhat relevant again. This is great, but it's really because there hasn't been a ton of advancements in AI from a technical standpoint since 2017. But the reason why it waited till 2022 was that we really saw a UX UI expansion. It became easier for everyone to use. And one of the analogies when I talk about making AI approachable is the idea of a car. If I had to manipulate a car engine to get from my house to campus every day, I will tell you, most of my students don't think I would make it right. But we have a steering wheel, we have a gas pedal, we have gears that we can change based on kind of the road conditions. And the same is true with AI. You're working with an interface. The engine. And the powerful model is the Models that are underlying it, like the O3 Pro model or something of that nature. But you're interacting with a steering wheel, so you're guiding it. You're. You're able to utilize your gas and break as almost like inference speed. So you want it to speed up because you need something quick and you don't need as much chain of thought reasoning. Right. So you could want a fast, quick answer, or do you let up on it so you get the model time to think and then do you switch gears? So, like, do you think about it? So I kind of try to use these analogies. Like, when you think about it, it's just using that analogy and thinking about the fact that AI has become democratized. And so how do we make it approachable? And I think the first step is removing the smoke and mirrors, breaking down the language and using examples that people can relate to.

[Marcus]

Gaja, how possible is that? Because there seems to be a growing knowledge gap between where people are and where the technology is going. And even though people are moving, you're seeing the technology move at such a faster speed that it seems like it's almost going to be impossible for them to catch up and for them to become AI literate, so to speak. So what do you make of making AI approachable? And do you think it's possible to get people, people up to speed or even close to where the technology is today, let alone in the future?

[Gajo Sevilla]

Well, you're right about that. It's. It's super fast. We're seeing innovations almost every week. And, you know, like what Kathleen was saying, it's. It really is the user interface that. That's really the only mode of control users have.

[Marcus]

Right.

[Gajo Sevilla]

So refining that part of it so that it's more intuitive. And I don't mean a dropdown of six different models, that doesn't really help a lot of people. But maybe demonstrating in simple terms that you choose this model, it could do A, B and C really well, but it's going to burn more trees or use more electricity. Just giving the whole equation, I think could help people find, you know, what, what they need.

[Marcus]

Yeah.

[Gajo Sevilla]

And maybe that will let them sort of understand how to use it to, to their advantage.

[Kathleen Purley]

And I think it's not just about how you use it, but also understanding the basics. And I don't, I mean, if you talk to anyone I've ever worked with, I have an analogy for everything, which is sometimes they're absolutely terrible. And not to say that this one won't be, but, you know, I find that even when I try to explain, like mixture of experts, I talk to them about like, hey, do you go. When you have an ailment, do you go to the hospital? And they have a cardiologist, a neurologist, a general practitioner, a GI doctor. Right. Does everyone show up to the room? No, because that's really inefficient. So when you think of an MOE architecture, it's very similar. You have a switch, like a primary care that then identifies what is the best area to send you. And so I think having like, they don't need to understand neural networks at depth. They don't need to understand kind of the ins and outs in terms of the technology nuance. But understanding how, like the cause if A, then B, then C, and how all the levers are related are key. And I think whether it's the MOE example or I use another one, oftentimes being a native Texan margarita. Right. Like making a model similar to making a cocktail. You have recipe cards which are kind of like chain of thought. If you have bad lines or really bad tequila, you're going to have a really bad hangover. It's like data. Right. And so it's understanding the different levers that go into place and providing analogy that people can relate to and be able to understand it from a frame of reference to say, okay, if something goes wrong, I know where the potential breaking points are.

[Marcus]

Yeah. What are some questions that you think good for folks to ask themselves when they're adopting AI or approaching AI?

[Kathleen Purley]

I think. I think the real one is are you willing to make a shift or a change? Like, is this going beyond just experimentation? I think from a business perspective. Right. Are you going to stay in pilot purgatory forever or are you actually going to invest in restructuring your processes so you're not. And it's interesting because I think, Marcus, you and I talked about this and I'd love to get gadget your thoughts on this too. It's like a chicken and egg. Because I've tried it both ways and I still haven't found the sweet spot. Ideally when I work with companies or when I'm working at Rice and talking to businesses and Fortune 500 executives that are coming up, I want them to think about AI and how do they readjust their processes as an organization to unlock the full potential of AI. Not saying this is how what we've always done. Now let's layer AI into that process. Right. They should rethink everything.

[Marcus]

Yeah.

[Kathleen Purley]

But I find that if I go that route first, it's often too intimidating to a lot of leaders or business because they just don't understand the full potential. And so they kind of need to say, okay, what am I doing today? How can use AI to create efficiency? Yeah, but efficiency is like table stakes.

[Marcus]

Right? Right. I liked the way in our conversation, Gajo, Kathleen had said that the question being, how can you change. Sorry, how can AI change what you offer and how you do it as opposed to just improving on what you currently do? And I feel like she was saying like it's hard to get to people to the more seismic changes that it could offer because people just want to tinker first.

[Gajo Sevilla]

Yeah. I think a lot of that is because companies all of a sudden have AI initiatives. They have budgets. Right. And so yeah, they're all in on, on adoption. But then do they ask themselves really what problems are they trying to solve? Right. Or you know, are they improving processes? Can they use the tools judiciously? And I don't think they spend maybe enough time evaluating before they jump in. Right. And that's when you find problems rolling it out. Because if you don't evaluate it properly, you can't really inform your employees about the benefits. Right.

[Marcus]

Yeah. When we were speaking before, we talked about how do you implement it? Do you start with the workers and say, because you just mentioned, you know, what's the, what's the problem you're trying to solve? To me, that seems easier at the worker level. As a worker, I know the pain points in my life, I know what I would like it to solve and then you can kind of implement that. But Kathleen, you were saying, actually it does. There are a lot of benefits to that top down to that CEO that, that senior person having a vision for what they want to achieve and going about it that way. Tell us about the some of the thinking there.

[Kathleen Purley]

Yeah, I think, you know, just given the clients I've worked with in organizations I've worked with, I have a lot of companies who are doing more of an employee up approach, which I think there's some value to letting them experiment kind of get their feet wet. But what I'm seeing is those companies are really focusing on just like that employee thinking about it from their day to day perspective. They're not thinking about how all the silos connect or they are not thinking about how you scale it or measure the success of the AI pilot. I think there's a stat recently from McKenzie like 19% of organizations are actually have KPIs for their AI initiatives. So there's about 80% of them that don't terrifying. And I think that's where it gets to like the real need to have the CEO involved as a visionary. I mean, they're the ones that are looking at what does our future look like as an organization, where are we going strategically? And AI is a strategic play, not a tech play. And I think, you know, I agree with Gajo and I have a framework in my book that I have that talks about, you know, what's the business challenge or opportunity? How do you quantify the value of it? You know, think, is it time saving more revenue generated, better quality? How do you measure that? What does that look like in year one? What does that look like in year three? What is the cost implication? What's the change impact implication? Right. And so I think there's a real strategic way that every organization needs to go about it. But I think it's time that CEOs and C suite start getting really involved and not see this as a delegated down initiative.

[Marcus]

Yeah, yeah. I like that framing of particularly what does this look like in year one, in year three, it seems like people want, even if they're experiments, they expect to reap all of the rewards tomorrow as opposed to a little bit tomorrow and then maybe more in the future. Yeah.

[Gajo Sevilla]

That's why efficiency is usually the easiest, the easiest one. Right. Like, oh, you know, we've, we've reduced headcount by X percent and therefore. And therefore what? I mean, you can do that, but what is the long term effect? Right.

[Marcus]

I mean, yeah, we'll save this much time.

[Gajo Sevilla]

It could be short term boost, but is it strategic in the long term? And do you really retain any value from, from that application?

[Marcus]

Yeah.

[Kathleen Purley]

And does it scale? And the companies that, I mean that company that creates efficiency cuts headcount. Okay. They're going to be successful in the next two to three years. The company that's going to really kill it, the superstars are the companies who are thinking, okay, we've created this efficiency. What do we do with this extra? What can we do now? What's the next thing? What's the next market? What's the next challenge? What's the next value chain thing that we need to rethink? Those are the companies that are going to be disruptive because at a certain point the efficiency game is going to become table stakes.

[Marcus]

Yeah.

[Kathleen Purley]

And you're going to lose that competitive advantage.

[Gajo Sevilla]

Yeah. They need to be steps ahead, like you're saying. Right. It's not just the immediate effects, but also, you know, the, the long term Goals on a number of levels.

[Kathleen Purley]

And I will say, like most C suite and I'd love to hear if you have the same experience. Most of the C suite or CEOs that I work with, they have very much so. Like a. Oh, I like, I use AI for like, you know, pulling out a party trick or a poem randomly or something silly. And you know, the McKinsey study that came out recently shows that like, you know, we have a huge jump in terms of AI adoption. But I think one in five companies are seeing a 5% addition to their bottom line. So there's really low kind of roi, roi. And those companies that had the highest roi, the biggest contributor to success was whether or not the CEO was involved in the AI vision and direction and the counseling of where the organization was going. And Today only about 28% of CEOs are involved in the AI strategy.

[Marcus]

Wow. Wow. And I guess that involvement probably varies across the board from barely to completely.

[Kathleen Purley]

Yeah. And I will say I've worked with a. I don't know if you know Marie Myers at Hewlett Packard Enterprises. She's the CFO there. She's fabulous and she is hands on and she is pushing her team and she's thinking about what are the challenges the organization's facing, how can AI help integrate it, how they can do more with more to meet the future needs. When I say she's in it. I worked with her over coffee recently on building her own AI voice agent to talk to while she was going to and from the car. And she didn't want me to build it for her. She was like, no, no, I want you to show me how to build it. I think those types of leaders are few and far between. Yeah, but you need that.

[Marcus]

Yeah. So you mentioned. So you have a new book. It came out this year.

[Kathleen Purley]

Yeah, it came out June 3rd or something.

[Marcus]

It just came out. Right? Yeah. Hot of the press, so called AI made simple results made real. An executive's guide to partnering with the future. And in that. So you have a chapter. I think it's Five, I thought was an interesting title of the chapter because kind of two questions came from it for me. One was the title is called is your organization ready for AI? And my first question was, how do you know? But then my second question right after that was, isn't everybody's yes and no?

[Kathleen Purley]

So it's one of those things. I feel like it's kind of like the equivalent of kids. Like you're like, are you ever really ready for kids? Like, no. Like Right. Like, do you have a car that can put a car seat in it? Okay. You know, maybe that's like, you know, but you're never really ready. But there are certain aspects and in the book I kind of have like a quick little assessment and it's, you know, I joke. Right. Like as a. I told some of my students this. It's like, you know, do I want to have like, you know, beachbody ready by May? Yeah. But like, do I need to be Victoria's Secret model level or just not scaring children off when I walk on the beach? There's a balance between it. I think there are some foundational things. Do you have data governance? Do you have certain ways that you're thinking about how AI can be applied? Do you have people who can lean in and take the initiative with it and invest in it? Because I think there's a lot of times where I see a lot of start stop where they buy the enterprise license, they roll it out, Kumbaya and no one touches it. And you have a few star people across the board. But even then I am shocked when I talk to them and like, oh, you guys are using deep research, obviously. And they're like, I didn't know that was a thing. They didn't know they could switch the model. And it's because they've kind of been like, let's experiment and bottom up approach with no one kind of taking the initiative. So I think there is some basis. Right. But you don't have to be perfect. Right. You don't have to have.

[Marcus]

You can be a bit more. Yeah, you can be like, if you're gonna have kids, I'm sure having some savings having family nearby to take care of the kids for you when you need, like, that's probably helps. But you can be an amazing parent without that.

[Kathleen Purley]

You can, exactly.

[Marcus]

Yeah. Yep. What was one of the most interesting findings from the book in your opinion?

[Kathleen Purley]

One is that AI is just like SAP implementation. It's going to be painful. It's going to take a while before you see the results. But turn it over to IT and we'll get it done. And what I found is the leaders who are successful within organizations oftentimes don't have an IT background.

[Marcus]

Interesting.

[Kathleen Purley]

So that is probably one of my bigger insights.

[Marcus]

Yeah. So what do they have?

[Kathleen Purley]

So they tend to be, and I say this as a former digital marketer and so I know it sounds like I may be tooting my own horn, but there's some value in people who have played in the digital Marketing space. Like I've written schema, I can kind of hack through some code. I'm used to rolling out a whole social strategy for an organization on vine and then have vine disappear beer six months later and have to throw it all away. I'm used to measuring, I'm used to testing, I'm used to failing, I'm used to educating. Right. That was a big part of my role as a digital marketer in the early 2010s. And so because I also think part of AI implementation is marketing and advertising. Right, yeah. How do you sell it internally? How do you sell it externally? See the bigger picture. So I found some, the people that have been really successful in this role tend to be digital. Like digital marketer backgrounds have a little touch of tech, but not too much that they're hand strapped in terms of they think that you have to have they over engineer it.

[Marcus]

Yeah, yeah.

[Gajo Sevilla]

So they're not mired down by the technology.

[Kathleen Purley]

Right, yeah.

[Gajo Sevilla]

That's good. It kind of sets them for you.

[Kathleen Purley]

And I have like a lovely friend and I know he's in London actually, and he's going to kill me for saying this, he's a Cloudflare engineer and he's fabulous. But sometimes I'm like, I know more than you do on some of this stuff. And it's because he's like, I haven't played in this space like you were doing machine learning algorithms and you know, natural language processing for years and so. And I've always been the type of person that like hack, slash like duct tape stuff together with like APIs to see if they. And I just got a lot more powerful over the last 18 months than I ever been. But you know, I think some of my engineering friends, they're fabulous, but they don't, they don't sometimes see the business opportunity or understand where the pain points are. They don't, they over engineer it or they. I think it's more complicated than it is.

[Marcus]

Yeah. My last question was going to be what should folks not do when it comes to AI adoption or what are some of the biggest mistakes made? You've given me a couple there. Do any more come to mind? Garjo, any jump to mind to you?

[Gajo Sevilla]

I think the biggest mistake really is not to test and optimize AI for specific uses. None of these models are really off the shelf solutions. They're pretty raw. You need to forge them and see where they fit within your stack. And I think they need consistent training and that goes both ways. People need consistent training on how to use them since they're constantly changing, they're constantly improving. So having that training, monitoring, evaluating these models, I think would really help address the needs of companies as they're looking at this ever widening field of solutions. Really?

[Marcus]

Yeah, yeah. The research on that, I remember the one survey and I think it was about half of folks said that their company offered training. But of those folks that where it did offer training, it didn't say they offered consistent training. It might have just been a quick 20 minute. This is what artificial, this is what AI stands for. So yeah, I think that's a really important one. Anything else, Kathleen, to finish us off?

[Kathleen Purley]

I would say on that end, one thing that I see time and time again and I try to push back because I do a lot of this for companies is they'll be like, okay, we have 2,000 employees coming in. They range from this job role to this job role. We want you to do a four hour AI intensive literacy and understanding how it can be used in their day to day. And I'm like, how somebody in marketing is going to leverage AI is very different than how your data scientist or your developer is going to leverage AI. Different guardrails for each and potentially different models and use cases and documentation and prompting. And so I really try to have organizations start to think about it at the, not even at the job level, but breaking that job into tasks. And of those tasks, where can AI be supplemented in? Because I think sometimes people think it's like a magic wand brush. And it's like if it can't do 100% of it at a plus level, then we're not doing it. Whereas you're like, okay, well can it do parts A, B, D and F? And you do C and E And you layer that in from that perspective. And I think changing that mindset in terms of that pattern recognition of breaking things down and then figuring out where the things overlap is I think huge. And it's one of the things I've been pushing at Rice is, you know, companies aren't just, they're not offering a lot of training. And so I think it's a huge, I mean, that's what they're looking for schools and business schools to do is to prep them so that they can think about this when they enter the workforce.

[Marcus]

Fantastic note to end on. Thank you so much to both my guests for hanging out with me today. Thank you. First to Kathleen.

[Kathleen Purley]

Thank you. I'm very excited. This was so much fun.

[Marcus]

Yes, indeed. Thank you to Gaja.

[Gajo Sevilla]

Yes, this was great. Thank you so much.

[Marcus]

Absolutely, absolutely. Thank you to the whole editing crew and to everyone for listening in to behind the Numbers marketing video podcast made possible by Quad. Please do subscribe and follow. Leave a rating and review if the mood takes you. We'll be back on Monday discussing how weight loss drugs became popular and the spaces they are reshaping. Happiest of weekends.

[Kathleen Purley]

Sat.