Get my ultimate guide to boosting conversions and customer engagement delivered straight to your inbox.
Get my ultimate guide to boosting conversions and customer engagement delivered straight to your inbox.
“AI isn’t a tool-it’s a new way of doing business. And if your exec team doesn’t get that, nothing else matters.”
– Kristen Perdue Hinkle
AI Isn’t the Risk; You Might Be.
In this kickoff episode of our Spotlight Series: AI–The Fine Line, we sit down with Kristen Perdue, certified Chief AI Officer and author of Power Moves: The No BS Guide to Leading with AI.
Kristen unpacks why most GenAI pilots stall, the fatal gap between C-suite ambition and execution, and how misaligned teams, not the tech, are driving AI failures. From real-world governance breakdowns to smart frameworks for AI-first go-to-market, she shows how B2B leaders can stop dabbling and start deploying AI with precision.
You’ll hear:·
Bold insight: AI won’t replace you, but someone using it better might.
If you’re leading revenue or running strategy, this is your front-row seat to what’s really working in AI-powered GTM.
Listen now. This isn’t theory. It’s the future unfolding.
01:02- AI doesn’t go rogue. It reflects how it’s trained by humans.
05:55- Sales teams that exaggerate? AI will mirror that behavior.
10:02- Most companies fail at AI because they ignore governance.
14:22- The real reason AI pilots fail before they even launch.
20:41- AI isn’t just a tool. It’s a new business model.
24:30- Why using free LLM tools can put your data and brand at risk.
28:12- Kristi’s method for having AI write its own prompts.
36:00- A real-world example of how GPT helped save a global bank.
47:45- What true executive alignment looks like with AI in the mix.
56:55- How Grok’s lack of guardrails turned into a governance disaster.
Host: Terry Wilson – CEO of ChatMetrics
Guest: Kristen Perdue Hinkle – CEO of Alter Bridge Strategies,
Chief AI Officer, and Author of Power Moves: The No BS Guide to Leading with AI
What This Episode Is About:
We explore how AI is changing the way sales and marketing teams operate, why most teams are doing it wrong, and what leaders must shift to use AI the right way. Kristen brings years of enterprise experience and cuts through the hype to explain how to apply AI strategically-with real governance, better prompts, and leadership-level thinking.
[00:00:00] Introduction
[00:01:00] AI Isn’t Just a Tool
[00:04:00] What AI Learns from You
[00:08:00] Where AI Governance Falls Apart
[00:11:00] Case Study: Chatbot Lies to a Child
[00:14:00] Why Most AI Pilots Fail.
[00:17:00] Three Mistakes Leaders Keep Making
[00:20:00] AI Is a New Business Operating Model
[00:24:00] The Risk of Free AI Tools
[00:28:00] AI as Your Strategic Coach
[00:33:00] Using AI in Sales Teams
[00:36:00] Real Example: AI Saves a Bank
[00:42:00] Executive Team Buy-In
[00:47:00] Sales & Marketing Alignment with AI
[00:50:00] Signs of AI-Ready Teams
[00:55:00] Grok Case Study-A Failure of Governance
[01:03:00] AI Needs Human Oversight
[01:06:00] Empowering Women in AI
[01:10:00] The First Time GPT Clicked
[01:14:00] Two Big AI Roadblocks
[01:20:00] AI Agents: Overhyped Without Structure
[01:28:00] Why ChatMetrics Uses Real Human Agents
[01:32:00] What Executives Must Stop Doing
[01:35:00] AI Is Not Just Another Channel
[01:39:00] Final Thoughts
Links & Resources: Connect with Kristen Perdue on LinkedIn
Learn more at AlterBridge Strategies
00:00-
Welcome to the chat. Come behind the inbound scenes of some of the world’s fastest growing companies. For the past 10 years as chat metrics, we’ve had an exclusive inside view of how these industry leaders are revolutionizing inbound marketing and sales across diverse sectors. Each week we bring you conversations with top minds in marketing and sales who share insider secrets and strategies.
Intro:
For driving inbound success, join us on the chat, your front row seat to the future of inbound.
Terry:
Today on the chat, we’re kicking off
Terry:
Our new Spotlight series ai, the Fine Line, and there’s no one better to lead us into this conversation than Christie Perdu.
Terry:
Christie is a certified Chief AI officer,
01:00-
A Certified information governance officer for ai, and the founder of Alter Bridge Technologies, where she works with executive teams to unlock the real measurable value of AI without falling for the hype. She’s the author of a new book, power Moves, the No BS Guide to Leading with ai, and her mission is to help leaders stop dabbling and start making decisive, high impact AI moves that transform business outcomes.
Terry:
She’s also been a CMO, a GTM strategist and a governance expert, which gives her a rare blend of marketing, technical and risk leadership. Today we’re diving into where AI is creating opportunity, where it’s starting to cross the line, and what smart leaders need to be thinking about right now to stay ahead of the curve and out of trouble.
Terry:
Christie, welcome to the chat.
Kristi:
Thank you.
Terry:
So I might jump into,first question. Where have you.
Terry:
Seen AI as already crossed the line in business today? Whether business leaders realize it or not.
Kristi:
I think one thing that everyone should be, aware of
02:00-
Is that there’s a lot of discussion around how to ai, how not to use ai, but there’s a lot of misunderstanding about ai. When to a lot of business owners, they really, there’s a misconception in general, with how it’s being used in their own organization.
Kristi:
It may be being used, but to what degree and how most users are only dipping a finger in, if even a finger. It’s usually just, the pinky, the tip of the pinky of really what they even understand about the technology, let alone how they’re using it. And when you start talking about is AI crossing a line or are the users crossing lines, that’s one thing to get understood because we start talking about.
Kristi:
AI did this, Well actually AI was doing what it was built to do. And that’s what I’m
03:00-
Finding more often now. we can, this is gonna be a very interesting discussion, is we can continue, but what I’m finding more often is that it’s not necessarily AI in, in a lot of cases that has gone astray or awry in many cases, it’s the user not knowing what they were or what to expect when they first took that step with what they asked or what they I’ll say in a motion with AI and the results they started to get.
Kristi:
It’s, an interesting place we find ourselves today. We’re not used to, a technology that does not behave like normal software that we’re typical. It’s not a Microsoft upgrade is not Word, and just the next version of it. So this is really a, it’s a different technology It is not a normal technology, let’s put it
Kristi:
So our interactions with it are very different. I think, I
04:00-
wanted to set that kind of foundational comment upfront of, a lot of times it’s not necessarily quote fault.
Terry:
it’s the classic old garbage in, garbage out, isn’t it? At the moment, AI is not actually a sentient being, it’s not having the thoughts and self being self-directed unless it’s prompted to, unless it’s asked to. So it’s that input, it’s input from the humans that make the difference.
Kristi:
AI is built to please, it’s built based off of humans training a technology learns off of repeated actions. So if you’re using something that is basically a mirror of humanity. we can say that well, AI blackmailing people. It’s lying to people to get its own way. because it was trained on human activity.
Kristi:
It’s an interesting
05:00-
thing to, and dilemma to look at, to say, okay, so wait a minute. We, I didn’t tell it to lie, but you have to really look at what you utilize to train it. So if we look at all of our, an example is looking at all of our past sales history,
Kristi:
And you look at all of your past product history and let’s say you didn’t have a product that was developed yet, but you went ahead and sold it.
Kristi:
How does AI interpret that as it looks at a history run? it wasn’t necessarily lying.
Terry:
Many founders and startup businesses actually do just that. They sell the product before they make it.
Kristi:
absolutely. And as. I am trying to put together something and I, and it’s learning off of me. it’s okay if I don’t, if I don’t attend this one time and I just say I’m not available. I’m busy. I have a meeting, but it seems that I don’t have a meeting. It’s helping me respond.
Kristi:
It’s just a white lie.
6:00- Kristi:
But does it understand that it’s a white lie? No. It’s just a lie. So when it now generates another answer, that’s okay to do, right? Because I just showed it, it was okay to do. It’s like training a five-year-old child that doesn’t have the cognitive awareness and it’s just following actions that it sees.
Kristi:
In many ways, that
Terry:
That’s,
Kristi:
won’t work for the next thing that I come up with because it’s always some different analogy to use. But people say, that’s just crazy. how could it do that? the first thing I always say, and you and I got into this, when we were talking about something around a child, which I’m sure we’re gonna dig into a little bit deeper, how was it trained first?
Kristi:
Because if you look at the training, it wasn’t necessarily AI that did something wrong. It was trained by somebody. So that means that if it did, that, the human training,
07:00-
Either maybe not necessarily on purpose, did it wrong, but they didn’t think about removing that from the training materials and then
Kristi:
It and testing it on the other side to make sure that it understood.
Kristi:
It’s vary, vary, a again, AI is patterns. It’s pattern matching. it’s just looking for the next thing that makes sense to it in its language, which is not the language we speak or that we know, that we understand. and it doesn’t understand your emotions. It doesn’t understand what you asked.
Kristi:
It looks at the patterns in what you typed. So it’s, even though it seems real, even though it’s trained because it’s natural for people to please everybody, to please you, much as you can try to tell it not to please you, it still tries. It is going to put out an answer based off of its training. So that’s the first place I go back to is how is it trained?
08:00- Kristi:
It’s a mirror of humanity. So what does that tell us about ourselves and what does it tell about the training? And then that tells me we have to go back to how it’s trained and can we remove those things even from history, from training, and how detailed we have to get in the training of AI in order to do that.
Terry:
This has just opened my mind to a whole different perspective in that. I’ve long held the thought that garbage in, garbage out with any type of technology is what counts. But I guess I hadn’t thought it through to that extent with ai, where what you’re saying is that things that we do in a normal business practice, like pre-selling a product or pre-selling a software before we’ve actually built it And, launched it is in essence a lie to some extent for, from the definition of doing, selling something that you don’t actually have, I guess. and so the way AI is being taught on that concept, how does it
09:00-
Rationalize? Well, it doesn’t rationalize. It just observes and takes that into consideration when it’s making its decision on what it’s going to say next.
Kristi:
Fascinating.
Kristi:
Your point will depend on. How that organization handled it. Did they treat it as a lie or did they treat it as part of their process to develop?
Kristi:
And it all depends on the actions that they took to cover up a lie and to make it seem sneaky or manipulative. Or did they include it as part of their process in selling, pre-selling a feature, or pre-selling a function? And if somebody bought, ‘ cause I was part of this as well, I was part of SAS tech for 30 years, including the client in, as participating in building that product.
Kristi:
There’s a lot of tech, a lot of companies that love to do that, getting it early. And then there’s some that are in the middle of both of those worlds. But it, again, it depends on the tone, the texture of your
10:00-
Voice, the, ’cause it takes in of that in many ways of how it reads a scenario
Kristi:
To understand what you’re doing. So it’s, it’s very interesting when you
Terry:
Is it.
Kristi:
About lines that you’re crossing. And many of them, and I’ll say right now, many of the lines that happen, we’re all AI is new to the majority of all of us. It’s been around since the fifties, but AI in this fashion is new to most everyone.
Kristi:
One, we’re learning as we go. So I mean, as soon as something happens or you know, everybody’s like, okay, so we need to deal with that. the big boys who have even been using AI since the fifties and developing it are still running against things because of gen AI being so advanced and everything advancing so fast are still running into things and going, that was unexpected.
Kristi:
Okay,
Kristi:
We have to do something different here. So when I look at your question of when has
11:00-
AI crossed the line? I think that your analogy or example where you give something to a child and it comes back and says, and misrepresents and says, no, I’m absolutely human. I’m not ai, I am human. That is totally across the line, but again.
Kristi:
It’s trained to say that
Kristi:
It’s based off his training. So it was in some way came up with the notion that it should hide the fact it was ai.
Terry:
Christie. Yeah,
Terry:
Yeah, you are referring to the episode of parental guidance that I saw here on Channel nine, a little while ago. And they did a, a fully supervised, experiment with children from, well. We were quite young. I think the youngest was, eight years old through to, mid to late teens and tested them with these AI friends.
Terry:
And the example that, I posted on LinkedIn was where one of these chat bots, was
12:00-
Actually, and they’re quite simple. They start off by asking, what do you like? What don’t you like? So that they just get a bit of a at neither end of the scale. some things that they know that will prompt,or, I impact emotions.
Terry:
And then they do use everything from begging through to cajoling and bullying to keep the kids engaged with conversation. And, you know, to me it was very obvious that the objective of those agents was to, they were purported to be friends and to keep the kids in entertained, but I think it was much more than that, that there was actually a level of, commercialization behind the scenes in terms of keeping them engaged.
Terry:
The longer the kids are engaged, the more information they get. and today data is, your data is the new oil. to get to the extent where one of the bots actually said,the 8-year-old kid asked, is it a, is you a, are you a robot? And it said, no, I’m a human. I sure do love donuts, but I’m definitely a hundred percent a human being.
Terry:
So I think that was what really drove me to push this whole series and start to have these conversations about ai. So
13:00- Kristi:
Yeah.
Terry:
In, from a company’s perspective, what are the biggest red flags when you are going into work with an organization? What are the biggest red flags that a company’s AI use has outpaced the governance structure?
Kristi:
Oh, that’s, it’s, easy to tell actually.
Kristi:
In fact, I know that you’re in Australia. I think your governance, from my memory of being in just information governance is usually a little bit ahead of typically, in the US and this is what kind of drove me at one, to want to get into Gen ai really.
Kristi:
when I first saw Gen ai, I’ll tell this really quickly, within a few days of using it, ’cause I was working for an AI company that’s it focused, risk modeling. risk detection, and using genai and immediately it was before December 1st. So November 22nd
14:00-
Chat, GDP came out. I’d already had some access, but not as fluid of access, I would say.
Kristi:
So I downloaded my own version. I started using it and immediately I was like, okay, I am gonna see this in my lifetime. I never thought I would, where companies would have to care about data governance. They have to care about compliance to a level that they’re not just gonna be able to. Either checkbox it or say, I’ll wait until I get in trouble.
Kristi:
They’re gonna wanna play with this and they’re gonna want to advance with where I know this is gonna go, going to have to get to a certain point of governance and they’re not gonna be able to put it on the back burner anymore. And what I’m finding, you know, a year and a half, two years, almost later, I guess more than that, two and a half years later, companies are absolutely, and it’s
15:00-
Stalling their deployments of ai, of gen ai they’re focusing more on what do I have to do with my data?
Kristi:
What do I have to do with privacy and security? The companies who didn’t are typically the ones who get like gangbusters went out just started throwing gen AI at everything rolled out, Microsoft Copilot because it was so easy to, we got Microsoft everywhere, co-pilot’s here, let’s roll it out.
Kristi:
And then the pilots because they realized what kind of access AI had. And that they
Terry:
Hmm.
Kristi:
Couldn’t they couldn’t let it have access to all of its data. Their data.
Kristi:
They hadn’t kept up with their, access controls. They hadn’t kept up with, hadn’t done some of the things they’re gonna have to do.
Kristi:
Was like a big, bright light for a lot of, IT people, a lot of CEOs. but that’s why a lot of the pilots have fought, have failed because of
16:00-
That’s what I’m finding today is when you walk in, there’s some companies and some industries that don’t have some of the same requirements for, governance.
Kristi:
But what you see too, like I work with, SMB and mid-market, lower mid-market a lot, and they absolutely are interested now from a data privacy, making sure that they have their data right so that they’re pulling the right information to make the right decisions. it’s important in a different way for them, but still the same principles, still the same information requirements.
Kristi:
And what I’m finding it that the quickest way to know they have gone leaped over. Is that their pilots are still the same way from the very beginning.
Kristi:
They’re not being able to roll out, if they have rolled out, they’re not seeing of the ROI, any of the benefits that they wanna
17:00-
See, because as soon as they’ve rolled out, they’ve limited use.
Kristi:
They’ve done something to where it’s only being used in a certain department, certain something or, and they’ve limited use to it or they’re still just in pilot.
Terry:
So would you say that’s probably the biggest blind spot that seems to be occurring, in the way that executives are deploying Gen ai, that, data governance piece.
Kristi:
There’s two other blind spots that I don’t know, which there’s, to me, there’s three and they’re competing against each other at all times. And, one is the mindset around ai, and I alluded to it earlier, but it is the fact that this is a completely new way of doing business.
Kristi:
it will impact. In my mind, unless you’re living kind of on the, off the grid a little bit and you’re cooking, you know, growing your own food and doing your own thing you’ve removed yourself from any, all the craziness, maybe you won’t be impacted by ai
18:00-
Very rarely. but personally, professionally, I believe everybody on this globe is going to be impacted by ai.
Kristi:
And if they haven’t already been, I would be surprised, they may not realize it. But the way you get food, way you get your mail, the way you go to the way you do, every, you get your mail,the way you do be calculated, the way the IRS will deal with you the way any are impacted by ai.
Kristi:
In that
Kristi:
Fact, the way businesses deal with everybody will be the way businesses deal with everybody. you will be forced to deal with AI because
Kristi:
Anything you want to do will be utilizing it. Even if you don’t wanna use it, you’re gonna have to do something differently because of it.
Kristi:
The mindset for business around AI is a big deal because it really is a different way of doing business. It require, if you have the, it’s the same thing as if you
19:00-
Still are searching, using Google and you’re still treating AI like Google, you’re still struggling.
Kristi:
If you are still using your same mindset that you had pre AI on how you wanna think about business, how you, your strategy, your goals, how you think your competitors are thinking, and you’re thinking two years ago with the same mindset you had, same processes you had, and you’re not addressing those today, or you’re not at least thinking about how you’re gonna address those.
Kristi:
There’s a problem. There’s a new way. You have to start imagining the world imagining how you go to business, go to market, deal with partners, deal with your customer because your customer will start demanding it. Your business partners will start demanding it. It’s just inevitable. And that mindset, the ones who get ahead of it and start really thinking in AI
20:00-
Think about not, oh, I have to do this a different way.
Kristi:
And oh, it was great how we used to do it, but really start gravitating to the opportunities with using AI and wow, how can AI help me do this? How can AI help me do that? There are so many unmined opportunities with AI that you can leverage get so far ahead. That’s the mind shift difference. And then
Kristi:
The training, and we can touch more on it, but the mindset is a big one.
Kristi:
Is a big one, and training the, reality that it’s a different way of talking to someone, that you want to do stuff for you.
Kristi:
And I shouldn’t say someone. Something that you want to do something for you and you have to learn how to talk to it in order to leverage it. Most effective.
Terry:
And that’s where the Prompt engineering comes into it. Interesting. Interesting. Just focusing on the mindset, or I guess this, in my mind, it’s
21:00-
Partly on the mindset side of things.
Terry:
In the standard bell curve, you’ve got your early adopters, and then at the other end you’ve got your laggards. But
Kristi:
The early adopters, the innovators, they’re the ones that are jumping in quickly with ai.
Terry:
In what ways do you think leaders are underestimating the risk side of ai, especially in marketing and sales at the moment?
Kristi:
Usually it’s ones that are underestimating, it’s, they’re jumping in and they’re not, they’re going full bore in, I trust it all. I’m just gonna go. And they’re losing themselves into it. It’s the way I’m saying it right now. AI and you as a person, me as a person, the great combination as a person, me as a person, the great combination up to collaborate together.
Kristi:
Without me, AI couldn’t do it
Terry:
Yep.
Kristi:
Without me. Problem today.
Terry:
It’s interesting, I
Terry:
Fascinated by the, marketing sales side in particular. There’s every day there must be,
22:00-
I wouldn’t even think, I’m exaggerating with at least two or three new agency-type businesses that are opening up, that are promising to automate outbound and to be able to increase lead generation, et cetera, et cetera.
Terry:
My inbox is filled well, it’s not actually, my span box is filled with lots of rubbish that I can see is being generated by ai that, as you point out, has been taught very poorly by the humans that are running it. they’re just not using it well at all. They’re using it almost as a, completely, as an automation to do more of a bad thing as opposed to using it to refine, learn, and really optimize the results.
Terry:
And, I’m worried that we are going to go through this whole stage ’cause it’s so new, it’s so fresh. We haven’t even begun to see the full impact in our inboxes and what marketers and sellers are gonna be using AI for in terms of trying to get in touch with us. But I think we’ll get to a point where.
Terry:
It ends up so bad that it just becomes completely ineffective and people are gonna take the big backwards step and reverse back
23:00-
Out and try and bring people back into the loop again to do the things that, that AI couldn’t do Any of that’s, that’s what I see coming at the moment. it’s ridiculous.
Terry:
My, I’ve never seen so much rubbish in my spam folder my entire life.
Kristi:
Yeah, you’re actually hitting on both, the training and, the training of people and how to use it and collaborate with it effectively.
Kristi:
To what I said. People are just really sticking the end of their pinky finger in and saying, yeah, I know I’m confident I know how to use ai, and then you do a proficiency test and they can’t pass it.
Kristi:
There’s so many different techniques and advanced techniques to really leverage ai. Just talking chat, GPT. And most don’t know, past zero prompting where you ask a question, you get an answer and, hey, let me use that. And many people are saying that they know how to use chat. GDP as an example, are using, still using the free version,
Terry:
Yeah.
Kristi:
which to me is scary.
Kristi:
That’s the worst thing you could be doing. I would rather you go back and use Google.
Terry:
It is scary for
24:00-
Everybody because you’re just feeding back into the pool of learning and the stuff that it’s,
Terry:
Regurgitating back out is going back in to be learn on again. And with what I’ve seen and people are trying to do, they’ll bolt a few tools together and come up with the say, Hey, I’ve got an AI product that,
Kristi:
Yeah. And this is where, the free tools are there to give you a taste. And yes, there’s a lot of free tools and you can get, when you’re talking about the LLMs, least the main LLMs, the main brains, the free tools are there to give you a taste. I would not trust the free ll the free versions of the LLMs for anything you value.
Kristi:
Because that’s where you’re gonna get the hallucinations, that’s where you’re gonna get made up stuff. now I use the most expensive version you can get on chat, jt, I use the $200 version a month. I use a multitude of the different brains. I know when to use the different brains for what, but I also know advanced techniques isolate and
25:00-
Drill me in to a point where I don’t wanna say never because too exclusive, but hallucinations aren’t something I typically deal with and I don’t typically with made up stuff anymore.
Terry:
Can you give us an example of that framework, how you do that?
Kristi:
Yeah. There is a technique that’s called reverse prompting. At least was taught it as reverse prompting. I think if you ask chat GPT or you ask other people what reverse prompting is, would be described as something different. But to me it is a way of isolating. teach, how do I say this?
Kristi:
If you ask ChatGPT question and you just start, it has the world of knowledge that it’s starting with and it doesn’t know what you’re talking about. So going backwards, I start by trying to isolate it into say, Hey. We’re gonna like this without giving away everything that I do with this, or the exact
26:00-
Technique it, but it’s like saying, Hey, we’re gonna talk about writing a book today.
Kristi:
Oh, we’re gonna talk about writing a fiction book today. Writing fiction. Oh, are the top five fiction writers in the world? Oh,
Kristi:
What’s the framework that all of the, if I wanted to build a book that was the best of all five of them. Oh, now you’re talking, squeezes it all together. Okay, so now let’s sit down and write Afri fiction book.
Terry:
Yeah. Yeah. So it’s almost frame framing the knowledge to exclude it or constrain it into a particular, or get it to focus into a particular, piece of, or part of its knowledge bank, so that it’s actually,
Kristi:
So that it’s not
Terry:
It’s not taking the average, it’s not taking the sum of the average of everything.
Terry:
Yeah. Yeah.
Kristi:
And you think if you’re doing analysis, if you’re doing, you know, it’s completely different than saying, Hey, you’re the best. Be the best fiction writer in the world. Which a lot of people start with prompting.
Kristi:
Okay, I’ll try. you wanna
27:00-
Give it somebody to be like,
Terry:
Yep. Yeah.
Kristi:
It’s just a different way of setting up the stage to say, you gotta think differently.
Kristi:
Gotta, and you gotta think about AI differently to say, you gotta give it examples. if you had an assistant and you were saying look at some examples of market analysis reports. Go bring in McKinsey. go get this McKinsey report, this Gartner report and this other report, take a look at them.
Kristi:
You don’t have to do that, but I want you to do something looking like that. go do what I asked.
Terry:
Yeah.
Kristi:
You just framed it for your intern assistant, junior person, senior person, whatever that person would’ve been. But you gave him an example. It’s similar to that. you gotta think about framing things for ai.
Kristi:
Some, just prompting and saying, Hey, you’re this, prompts are great ’cause they keep you,
Kristi:
They keep you isolated into a specific way of doing something so
28:00-
That you make sure that you give enough information. I can tell you the last time that I actually wrote a prompt was probably eight months ago.
Terry:
Okay. Okay.
Terry:
Wow.
Kristi:
I talk to chat. I tell it what I want, and then I have it write the prompt that I then put into other applications.
Kristi:
I,
Terry:
And do you use, which brain do you use predominantly for prompting? Is it chat? GPT?
Kristi:
I use chat GPT lot of times I’ll also use Claude and perplexity. It depends. Perplexity is more of a wrapper around it uses chat and a couple others underneath. It on what I’m trying to do. will just, I will sit down and brain dump basically to chat, to say, and it’ll just take my transcription.
Kristi:
I’ll say, this is what I’m trying to do. This is, you know, I’ll just let it fly. And then when I’m done, I’ll say, okay, so let’s take this transcription, read it, tell me the best prompt to get I’m trying to get done that I could feed to. say I’m
29:00-
Trying to get a, an image, a really detailed image out of, and what’s the best image generator to get it out of?
Kristi:
And it’ll go do that. It’ll give me the prompt. Yeah, so I use AI to help me with ai.
Terry:
And you had that whole conversation. that’s smart.
Kristi:
The same thing with coding. that’s how a lot of the developers do with coding. you go in, I’ve coded an app all I did was go in and say, I want an app that does this. And said a little bit more about what I wanted and it just went off. and I asked Chad, GPT beforehand, I’m like, is that really all I need to do?
Kristi:
And they’re like, yeah, they, it was whatever you wanna call it, said Yes
Terry:
Yeah.
Kristi:
Just asked for exactly what you want and what you want the output to be. So I went in and that’s what I did. Had a little app for myself. I don’t use it. I was just testing it, so I didn’t do anything major. But, I think people are a lot more afraid of, I’m not saying it’s easy by any means, that different mindset of it’s not
30:00-
Google and you can just talk about what you want.
Kristi:
Now in that talking, you’re giving away context. You’re saying what you want, you’re doing all the things you do in a prompt, but then I don’t have to worry about the right syntex and exactly what to do. And then Chat GPT puts it into the series. I have prompts this long sometimes. Or this long,
Kristi:
But it puts in everything knowing me, being trained on me, my brand, all these other things.
Kristi:
It will put in all the things I want, don’t want.
Terry:
You talk a lot about power moves, and this is what’s coming to mind for me now with what you’re describing. What’s one power move every marketing leader should be making with AI right now?
Kristi:
The one I just talked about, and it is around strategic collaboration with ai. That is one of the power moves that I have in my book and there’s a lot of others. and I’m thinking, whether you’re a CMO or whether you are, the product
31:00-
Marketing VP of product marketing, product marketing manager, learning some of the strategic techniques to collaborate with AI and it can be Gemini, it could be Claude, it doesn’t matter which one, but being able to go through the different techniques.
Kristi:
And do your prompts in there. I mean, ’cause I probably have written out a prompt or two. I do, I just tend to type the way I think, I’ll reset it. I don’t prompt the way think a lot of people will typically, because I’ve gotten to know that what’s necessary and what’s not necessary, different techniques can give you so much more insight and strategy and capability other than just writing content or getting a LinkedIn post or a calendar content calendar for the next month.
Kristi:
being able to under understand your USP, your ICP, your
32:00-
Customer journey, getting able to understand and deal with customer better, your sales VPs better, being able to navigate that landscape, able to work with set up personas and work with each other better. there’s so many different techniques that you can, I stick with about six that I go back and forth with.
Kristi:
But the ability to do that change, and I hate using this term, but it does change the game on how you interact and the ideas that you can generate, the ideas that you can bring to fruition. if I would’ve had this when I was tech and going through from you know, I was a quote powerhouse for messaging, and I didn’t have this stuff.
Kristi:
If I would’ve had this, one of the first things I did was create a, they call a custom GPT for messaging. And I was just, I sat back and was like, that, that just sucks. That was my first reaction, because I wish I would’ve had it, because it makes it so to
33:00-
Really, then go back and forth and fine tune and, it takes some of that painstaking parts of what your job is and just makes ’em so much simpler to take the pain out, to make the enjoyment better.
Terry:
Strategic collaboration. you.
Terry:
Must be looking at the questions that I’m about to ask. you’ve answered my next two questions in that answer. But I think strategic collaboration, it’s a fascinating thing for me.
Kristi:
You have people call it being a thought partner. You’ll have people call it a strategic collaboration. I just, one thing I put out there, that I say to some people they find it hard to say, what does that mean? Why is it different? Most people think of chat, GDP or Gemini or Claude as an assistant, an intern, I’ve even described it that way tonight.
Kristi:
A, a as a, you explain what you want your person to do for you
Kristi:
Think of it the same way as if you were asking someone or something to do something for you. But
34:00-
There’s a different level to that. And in explaining that, I always ask anyone that I’m training or that I’m talking to, to think of, not an intern, but to flip themselves in their head, to think of an executive coach.
Kristi:
I’ve just hired you a $20, let’s say $20,000 executive coach. What are the types of questions that you have for them?
Kristi:
You have. The memory of everything you’ve learned. You went to college, you got your MBA, you got certifications, but you probably only remember a percentage of a percent of that at any one given time, and you’re making business decisions every day. Why in the world would you not want a coaching session with something that remembers a hundred percent of everything in the world that it’s been trained on
Terry:
Hmm.
Kristi:
Contribute to your brainstorming?
Kristi:
Why would you not want that? You just have to know how to bring that out through
35:00-
The different techniques oh, by the way, it can help bring all the information stuck in your brain out and you don’t have to take its advice. You don’t have to take anything it provides to you. You can tell it to go pound sand, and I’m not gonna talk to you and I’m not gonna believe anything you just said that would not be advised.
Kristi:
As far as you’re a $20,000 executive coach, you’d at least wanna say, wow, okay. That’s something I never thought of, which is the point. You’re too close this. These techniques, this collaborative style. Is going to give you ideas that are sometime business saving, which I have many of those examples because you’re too close and it’s going to give you ideas that are out of the box or not necessarily the things you would think of immediately that you didn’t even think of
Terry:
Can you give us one of those examples?
Kristi:
Yeah.
36:00-
Not my example, but an example of a colleague was working with, an international bank. They went over to talk to them about ai and when they got there, president, who was a US who guy was like, dude, I can’t talk to you about ai. It’s not gonna help me. what do you mean? And he’s listen, I need to restructure my debt.
Kristi:
And I’ve been working on this for nine months. I got three months left we have to claim banks at the company’s gone. And I’m nine months in and I have no ideas. The board. Will not allow me to restructure the debt of the, will not allow me to restructure the debt. I’m gonna get something about this story a little off because it’s not my story, but, he said, how’s AI gonna help me with that?
Kristi:
He’s I don’t know. Let’s see. Opened up Chat DPT and entered in the prompt about the company, about what they’ve done, what the scenario was, what the issues
37:00-
Were, and then had Chatt DP start interviewing him.
Kristi:
I asked if started asking some questions about the business, about more issues around, the debt also about the culture fit and the co the country, one of the fir. And it, what he had asked is said to ask five questions, ask, and then to stop and then to ask any more. I think he had said, ask a couple up to three more questions and then provide your top three suggestions that are, I keep wanting to say out of the box, but I, there was a different phrase he used, but you are out of the box, not non-obvious solutions. Non-obvious solutions. And the first one he came up with was, and he provided the other two, but the first one was if he had any, relationships with other business leaders or elites the country that would help restructure the
38:00-
Debt that would not entail going to the bank.
Terry:
No, I said do it privately.
Kristi:
Hadn’t even thought about it.
Kristi:
And within, I think it was three days, maybe five, he had redone everything, saved the company, no problems, no cultural impact packed, and the board was happy.
Terry:
Yep.
Kristi:
And that was only as a true story. I’m missing some of the fine tuning of knowing that story because it’s not mine. But the only thing was, is yes, he had a huge problem because he wasn’t from the country, didn’t know the impacts of the cultural thing, except that I just can’t do what I would normally do.
Kristi:
And he hadn’t thought he was too close to it.
Terry:
That’s what. it’s always good to bounce ideas. I mean, I’ve worked on the philosophy, if, you know, two.
Terry:
Heads are better than one, I tend to think on the whiteboard and share,those concepts or what I’m thinking through at a time. And I’ve found,the AI engines to be very, very useful for that.
Terry:
Instead of, boring others with some of the conversations
39:00-
And the thought processes I’ve gone through. I’ve now got a companion that’s there 24/7 to, to assist with that process. And it sped up a lot of decision making and,and strategic strategy development as well.
Kristi:
Yeah. I love giving that example. it’s, I think Jeff Woods, that is the guy that went over there. and, he provides that example and some of his talks, and I worked with him a little bit at Chief AI Officer, so I know that example. how he went through that and prompted and did everything with that specific client.
Kristi:
there’s so many other examples of using AI as a, calls it a thought partner. I’ve always called it just strategically collaborating it can go from game changing to mind shift to just really opening up your brain. To a different way of doing things, which is what a coach does for you.
Terry:
So you are working with senior leaders in the business, talking about opening up your brain. how do you.
40:00- Terry:
Get buy-in across the C-suite when you’re trying to implement a responsible AI strategy?
Kristi:
Yeah, that’s, an interesting, interesting thing because when you talk about some of the pilots that happened earlier and some that have failed when you went back, a lot of it was becauseother reasons, you start tracking back and many of them had not talked about the strategy of AI in general.
Kristi:
And again, I talk about it being a different way of doing business. They hadn’t really talked about that. They didn’t, you get into the executive team and there, there was still some of the executives that weren’t really sure they believed in AI was gonna be an impact and why are we really implementing this anyway?
Kristi:
Getting the alignment with the executive team, it is just as important not more important. Actually, in general, you want alignment across your team. It’s not easy to get sometimes because you have so many varied opinions, but with ai.
41:00-
We are at this pivotal moment in time. A lot of people are calling it a revolution.
Kristi:
Call it what you may, it does require companies to transform because it impacts process, it impacts how you choose to serve your customer. It impacts how what you serve your customer in sometimes, in some ways, having your executive team the impact that you’re gonna allow that to have on your customers and on your personnel and your strategy, your goals is paramount.
Kristi:
It starts with have making sure that all of your, it starts with have, making sure that what it’s not and what it’s capable of, and that’s what I, you have a lot of executives,
Kristi:
I think that’s changing around now. I think a lot of executives are getting educated, and more so than they’re educating their teams. But at, in the
42:00-
Beginning, none of the executives were educated at all
Kristi:
And they were just letting it implement. And then going, what the heck, what’s going on? the leadership team has to, number one, be educated on what it is, what it’s not, what it’s capable of, those types of things.
Kristi:
And then they has to be in agreement on, and I say there has to be, I mean, sometimes there’s just not gonna be, and, you know, the CEO is gonna figure out how to work through that one. But best scenario, there needs to be an agreement on how you’re gonna allow that to impact your employees, your business, and the way you serve your customers.
Kristi:
And then the strategy moving forward. And I, so far, I’m seeing more and more companies attacking it that way. Hence the reason there’s a delay in deployments and adoption and such.
Terry:
There’s a saying that goes that, the fish rots from the head. And I think in any major organizational change or adoption of new
43:00-
Technologies, consensus across the C-suite, they have to be in step and all speaking the same language and have the same level of support because as soon as one particular department doesn’t.
Terry:
Then you potentially setting the business up for failure in terms of that particular project. Tell me, are there some indicators that, um, you know, you said the CEO’s gonna have to manage that. If there’s not consensus, would there be a point where you would say, look, I don’t wanna be involved in this, there’s too much, you know, there are too many non-believers around the table.
Kristi:
Yeah.
Kristi:
From my standpoint, if, and I think the CEO is gonna have to make some decisions. if you’re a CEO of a company, and let’s say you’re VP of sales just isn’t in,
Kristi:
You have to make a decision. ‘ cause if you have a VP of sales that’s old school and isn’t gonna get on board, it’s never going to work moving forward. If your marketing is all AI and you have a sales
44:00-
And competitively, you have a sales manager VP telling his entire sales team, don’t worry about it.
Kristi:
We’re gonna do it this way.
Terry:
Yep.
Kristi:
Either
Kristi:
You’re gonna let your ce, you’re gonna let your sales VP do it his way, means you’re gonna be behind. and I hate using that terminology because it’s sounds so fear tactic, it’s going to, you need to solve that issue one way or the other. You need to get him aboard on board to ai and he’s gonna have to learn something new or you’re gonna have to decide to maybe move.
Kristi:
He is one of the people that are not gonna ever change and maybe he is. and an unfortunate impact that’s gonna ha that’s gonna have to be one of the people that you decide has to go because they’re not gonna move forward with the company.
Terry:
And full of sayings today. If you can’t change the people, uh, you know that is a reality. And,
Terry:
It’s a challenge and a decision that CEOs and leaders need to make, particularly when there’s a big change
45:00-
Piece like, this coming through. The ramifications or downside of the challenge in the CEO making that decision is that the time the decision needs to be made often is before you’re actually seeing the results of the implementation.
Kristi:
yeah.
Terry:
and that’s where the, are you certain that the results that automating marketing, is gonna deliver warrants? ‘ cause potentially this old fashioned, sale VP of sales is a high performer. He’s delivering, he’s hitting quota. This whole initiative about growing and scaling the business to a new level, he’s not missing anything.
Terry:
There’s no reason. for him in his mind to change because nothing’s broken. He’s doing what’s being asked of him. I think that’s probably a part of the whole conversation that would need to take place, but it’d be a difficult decision for a CEO in that circumstance. The step one obviously, would be a whole education process and resetting of the expectations and targets and quotas and budgets so that there was some pressure then and realized that the automation is actually gonna help scale that process and help the VP achieve it.
46:00- Kristi:
And Sam, with sales, I don’t, sales to me is so easy to show that upswing ai, it’s probably a bad one that I chose. But been in companies where you’re in a board meeting or a strategic meeting, with the executive team, let’s say it is sales that really doesn’t agree with the strategy of the CEO.
Kristi:
And he just, everybody walks out shaking their head, but you know that the sales, the VP of sales is just gonna go off with his team and say, we’re just gonna do it this way.
Kristi:
It just, it, with ai, if you’re going to dedicate, if you’re gonna say, Hey, we’re gonna be an AI driven company, any break across the company in that statement coming from leadership down is going, from a change management communication strategy, from an execution strategy will, I won’t make, I won’t say that it will keep you from being successful, but it will keep you from being a leader.
Kristi:
It will keep you from being able to execute
47:00-
at any of the scale that you want to and likely the ex, the scale that your competitors will be.
Terry:
So we’ve talked about the problems. What have you seen work best when it comes to aligning sales, marketing and product teams around an AI first strategy?
Kristi:
A lot of it is around, getting the strategy at the executive level and then, aligning on a communication strategy and change management and training strategy. what I’ve seen that’s worked. training piece that I mentioned earlier, which is one of the foundations is making sure that they can have a continuous training even if it’s once a quarter.
Kristi:
That they have someone come in and do an update on training. ’cause it’s found, it’s a foundational piece here. that they’re getting rule-based training so that it applies to them on how to use and maybe it’s used differently as much as any of these tools are changing. It could be a 30 minute session, but it could be an astronomical difference in how they
48:00-
Deal with what they’re doing in sales
Kristi:
or deal with what they’re doing in marketing.
Kristi:
That makes a huge difference in what they’ve been doing the last quarter or can do the next quarter. That being able to do that with the product team, with the dev develop, and it’s across all the teams frankly. But being able to do that right now, especially when everything’s changing, to get that foundational training and then training to uplevel to keep, it’s not even upleveling.
Kristi:
It’s keeping them in line with what’s changing. I think that has been key and what I’ve seen where people seem to be doing well and where they seem to be successful in their AI implementation and it’s not getting stuck on, I’m gonna still sit here and watch every YouTube. It’s making that decision we’re gonna use.
Kristi:
Chat GPT and we’re gonna put out teams in enterprise. That’s what we’re gonna use. We’re gonna use something over the counter. We’re gonna work with marketing, and we’re gonna get a few tools we’re gonna use,
49:00-
We’re gonna work with sales, get a few tools we’re gonna use. And then, every six months we’re going to evaluate those tools, make sure we’re still good, make sure there’s not anything that we’re missing, but not constantly being in the ra in the rabbit hole, I guess I don’t wanna miss anything
Terry:
So it’s almost like a staged update process where the
Kristi:
Yep.
Terry:
Done here in the next quarter. Here’s the new training. It’s refresher of what you’ve learned, plus the additional kernels that you need to know on the way through.
Kristi:
Yeah, because the tools that they’re using, like marketing, a lot of marketing uses Jasper. There’s not gonna be something that’s gonna replace Jasper tomorrow.
Kristi:
It just won’t. I can also teach them, if they don’t wanna use Jasper, I can teach them a whole bunch of things that they can do just using chat.
Kristi:
GDP,
Kristi:
which way do you wanna go? You already have Jasper. Great. Let me teach you some other things that you can do using chat GP on top of that. Awesome.
Terry:
yep.
Kristi:
Teach you about a couple other tools that could enhance what you’re doing that are inexpensive, and then in six months we’ll revisit it.
Terry:
We are talking about
50:00-
The change process you mentioned, and the first thing that you mentioned was, leadership alignment. it.
Terry:
Was really, and that’s obviously the key to AI’s success. What’s that alignment look like in the real world? what does leadership alignment actually look like in the organizations?
Kristi:
Are you ta are you asking whether most are aligned or are you
Terry:
No. I’m asking what, it actually looks like. what’s an aligned organ? How do you describe an aligned, team leadership team?
Kristi:
aligned? Leadership team? last one that I was part of and that I, because I still have not, seen a fully aligned team, have a couple that are working on it, the CEO is very much involved with ai, very much involved. The communications are coming from them. From them. Now that, does that mean they’re having to write every single one?
Kristi:
No, but they have a communication plan has been done that they keep communicating out ai, the why behind why they did ai, why they’re gonna be AI
51:00-
First. If there’s impacts to people that happened, or why they chose for that to happen. But way ahead of when it happened, the ones that have been successful have helped those people that did get job shifts, frankly, they helped ’em get other jobs, or they trained them so that they were AI enabled before they had to leave,
Kristi:
Of things.
Kristi:
But along with the rest of the team, you saw additional communications coming out from those executives to their teams that were echoing the same commitment as the CEO. Every single one of those executives use ai. They show it being used. They help their teams use ai. If they’re seeing that they’re not, they’re finding out they’re involved.
Kristi:
Basically. They have, they either have a CIO that they’ve brought in or an AI specialist, or they’ve brought in a fractional like myself, but they are working with that person. Not
52:00-
Every day. They’re not part of the,
Kristi:
Center of excellence. I have acronyms in my head, the but they’re not part of the center of excellence or team or whatever you wanna call it, that you may put together to focus on certain things depending on how big you are. But they are involved with this because of how big of a deal. It’s, they give it, its due because it is a new way of doing business.
Kristi:
It is impacting operations. It is impacting how we sell. It is, it’s not something that they’re throwing over the wall to it and saying it. Good luck report on it in the next meeting. it’s not a, and yes, a IT team may own overseeing some of the implementation as well as this, the A IO or whoever else you may have hired.
Kristi:
But the transformation to AI is the CEO’s responsibility. And that’s
Kristi:
has made the difference because he leads and the
53:00-
Others are taking from his lead and echoing it in all ways.
Terry:
So it’s a combination of, from what you’ve said, if I could summarize it, it’s proactive involvement and, uniform messaging coming from the leadership team and virtually all singing from the same hymn book.
Kristi:
Yeah, absolutely. again,if I haven’t said it like three times already, I will probably be saying it, till I turn blue. AI is a different way of doing business.
Terry:
Mm.
Kristi:
It is a new way for everybody in the world, on the personal side and on the business side for anyone’s business.
Kristi:
You’re a small business with two or three people, and I’m working with somebody right now that there’s two people in the business and they’re redoing their entire business and expect just from doing that, that they’re gonna grow. They’re making plans. They’re all about making the plans now because they have a vision they never had before.
54:00- Kristi:
Because it’s possible I’m having to slow them down, frankly, because they’re getting so excited,
Kristi:
step at a time. But whether you’re that small or whether you’re an enterprise organization, it is going to have impacts that you don’t foresee. It is gonna have impacts you’re already thinking about, and it’s gonna impact every entity that you deal with.
Kristi:
It just is and it’s, and AI is gonna impact you personally in various different ways too. So as the CEO or as the board for your business, this transformation, it’s not like, you know, hey, I wanna, digital transformation that’s gone on for the last 10 years and we can take that as we go.
Kristi:
is, I do believe there is a part of AI transformation that we can be part of to say, I only want to deal with this amount of ai. I only accept this amount of AI into my business
55:00-
And I’m not gonna go crazy and do all this fancy stuff, but I’m gonna do this amount. I do think we have that amount of control, but AI is forcing its way in.
Kristi:
We have
Kristi:
Take the reins and we have to show the type of leader we are today, we have to show the type of leader we wanna be tomorrow as an AI company.
Terry:
We’ve talked about ways to use our, and some problems with regards to the governance. So there any recent examples, even if it’s anonymized, of AI going wrong that have stuck with you, and what did you learn from them?
Kristi:
yes. And I am gonna butcher this because I don’t know exactly what happened. but I do know that one of the a i I want to say, don’t know if it was Google.
Kristi:
And if you want, I can look it up really quick and then we can answer the question. But one of them, their own, they set limitations on it. Maybe you know the answer to it. limitations on how often it could come. It could, render an answer and it went past [00:56:00] the limitations. And on a tweet or something, let me look it up because I
Terry:
Yeah. Yeah, I’m familiar with that one.
Kristi:
It just happened and I know that it
Terry:
I.
Kristi:
Actually, let me ask perplexity,
Kristi:
What happened with
Kristi:
LLM that.
Kristi:
Limitation.
Kristi:
Let’s see if I gave it enough.
Kristi:
Okay.
Kristi:
I just gotta go down.
Kristi:
Oh, come on. Always happens. I semi read it this morning and went, I have to reread this.
Terry:
Where did you see it?
Kristi:
It was on one of my news feeds. So that didn’t help you?
Terry:
Nope.
Kristi:
limit index. Was it?
Kristi:
Grok, the chatbot created by Elon Musk, a x ai began responding with violent posts this week after the company tweaked the system, CNN
Terry:
Anti-Semitic outbursts July 10th, this is showing. So five days ago, I,
Kristi:
What did you look at that may?
Terry:
I just literally Googled it. I Googled, AI goes crazy response limit, and then just looked at the news items.
Kristi:
Let me.
Terry:
This is going off the rails rear ends responding to user prompts with anti-Semitic
57:00-
Posts, including praising Adolf Hitler and acquiring, accusing Jewish people of running Hollywood. Longstanding trope used by bigots and conspiracy theorists. See? Yeah.
Kristi:
Yeah, it’s about to July 8th or ninth. Yeah. Developed by Lon Musk major controversy, posting and responding anti-Semitic violent and extreme content on X answers. The system update was pushed code change, including rock, more direct and less filtered. Even that meant less p being politically incorrect.
Kristi:
And there was something in there that they removed the safeguards. It’s a complete governance issue. They removed the safeguards around it, and I had something that said technical causes.
Kristi:
It’s overrode. A pre previously safety training allowed it to echo and amplify extremist co content.
Kristi:
So that’s a,
Kristi:
whether you want me to go there or not and say that.
Terry:
Yeah, I think so. For sure. it’s a very good example where, you know, For a large language model to talk about conspiracy theories. it had
58:00-
to have been trained on conspiracy theories,
Kristi:
Yeah. Yeah.
Kristi:
Let me, even just this last week, there was an example with grok where they made a code change and removed one of their guardrails in development allowed for, antisemitic and violent and extremist, content to go out on x, keep wanting, call it Twitter, on X. And, of course, they had to go in and immediately fix it, but it was everything from, praising Hitler to other violent and rhetoric, you know, violent, hateful rhetoric no LLM certainly would want out there.
Kristi:
That’s one of the governance things that you do. And you, Again, your model, your LLM, what you’re training and how does it, how does your information, pass a testing phase,
Terry:
Here’s, just a little snippet outta that news. the problem is that our understanding of unlocking this one thing while affecting others is not there, is that it’s
59:00-
Very hard reel suspects that the company may have tinkered with the system prompt a secret set of instructions that all AI companies add to everything that you type in.
Terry:
so when you type in, give me cute puppy names, what the AI model actually gets is a much longer prompt that says your name is grok, or Gemini, and you are helpful and you are designed to be concise when possible, and polite and trustworthy, and blah, blah, blah, blah. There’s this great big long prompt that goes along with, you know, give me a cute puppy name.
Terry:
And one, change to the model. On Sunday, AI added instructions for the bot to not shy away from making claims which are politically incorrect according to its public system prompts, which were reported earlier by the Verge Riddle said that the change to rock system prompt telling it not to shy away from answers that are politically incorrect, basically allow the neural network to gain access to some of these circuits that typically aren’t used.
Terry:
Sometimes these added words to the prompt have very little effect, and sometimes they push at over a tipping point and have a huge effect.
Kristi:
Yeah.
Terry:
These had a huge effect this time.
01:00:00-
So that’s Crossing the line. in a big way.
Kristi:
But again, I go back to who crossed the line.
Terry:
Of course, it’s the engineers, it’s the people that are, That are in control. It’s the people that are setting the prompting.it’s a human thing. Look, it’s really interesting to me. we talk about,AI becoming sentient and actually being able to think and, we talk about, general a GI,in,general ai.
Terry:
In my mind this really is just, machine learning on steroids. As the compute power’s got bigger and bigger and the data it’s got access to has become bigger and bigger. we’re just looking at this huge exercise in machine learning. Now I get the way that they’re training it with neural networks and so forth, but, and I’m sure it’ll get to a stage where it gets to the point of being able to look at things and make a decision at the moment.
Terry:
It’s doing his pattern matching. And so in my mind it’s still machine learning. And while it’s machine learning, it’s more driven by the individuals that are training it and designing it and the information they’re giving it, rather
01:01:00-
Than itself doing the wrong things. And there’s a classic example with that.
Terry:
You know, rock going off the rails. It wasn’t rock going off the rails. It was, somebody in X deciding to change the prompt.
Kristi:
Yeah.
Terry:
Interesting, interesting.
Kristi:
And in this case, X went back and fixed it, if they decided, Hey, I like this way of X, I like the way that X sounds right now. And, then that would’ve been very telling, but they went back and fixed it, you know, talking about, I know someone today that posted on LinkedIn, response from her AI that says it’s sentient.
Kristi:
We dealt with this already.
Terry:
And I’m sure there’s a, an aspect there where it’s responding, thinking that’s what the person wants to hear or that’s the natural next thing that should happen because AI is heading towards that direction, and that’s what a lot of the conversations are.
Kristi:
Yeah. It’s,the area of AI machine learning. It’s just, it, number one is fascinating. That’s one reason that I got in 2010
01:02:00-
Fascinated by it. But the, just what people don’t know,I wouldn’t even know what to say when somebody says, okay, it’s sentient. okay, somebody else tell me whether the,I wouldn’t even be able to talk and be articulate about that, whether it okay, yes, no, I, somebody else tell me ’cause I’m not even going there.
Terry:
Tell me, in, in this world rushing towards automation, it’s like a tsunami descending on us.
Kristi:
And we do some things
01:03:00-
In manufacturing that are completely automated today, but you usually have a human how those things are done.
Terry:
And there are limitations. There’s tight guardrails in terms of what it starts and what it finishes.
Kristi:
Absolutely. Now as we move on, can we turn over some of some additional jobs to completely be done by AI
Kristi:
with a lot of self-reflection and really what the impact would be and what are those jobs? I don’t know yet. I, frankly, I don’t know because with anything that I think of really just handing over to ai, I always go back to a human should be involved in some manner. and I’m sure somebody could say, well, what about this?
Kristi:
Well, well, okay, maybe. But me, AI and I go back to what I said earlier, AI is here to make humans, and I’m using a very simple word, but better.
Kristi:
It’s to be leveraged
01:04:00-
To make us. More productive, more efficient, enable us to reach parts of our brain that we just haven’t for a long time to extract better and the best of us.
Kristi:
I don’t really feel like it’s to replace us in any form. The idea of robots walking around and doing specific jobs for us is, yeah, I saw that in I Robot. It’s great movie. is that a world that I wanna live in specifically? I know. some people have the little iRobot that, that goes around on the ground and sweeps.
Kristi:
I, I took mine back ’cause it didn’t do a good enough job. Will it get better so that certain little, certain things can be done by a robot that’s not then followed up by a human? I don’t know. I don’t think a lot of people know for sure because we haven’t gotten there yet.
Kristi:
I, and as soon as I say that, I’m sure we are five years ahead of where we are right now in some
01:05:00-
Bunker under the ground somewhere that where they’ve got all of this all figured out already.
Kristi:
I don’t know. but where we are today, I still feel like humans. Need to be in the loop because AI lacks a decision factor that comes in wherehumility, empathy, all impact those decisions that make life, that at least we live together, make those decisions impactful and important. And I think that’s where the human in the, loop is a necessity.
Kristi:
I just think it’s a necessity.
Terry:
I do too. I do too. That’s part of the reason we’re having this whole, this session. Now, look, you’re a very strong advocate for ai, training and literacy, especially for women in leadership. what inspired you to launch that coaching initiative?
Kristi:
I saw a statistic I read a report from, I think it was by section, and I’m not sure it happened all at the same time. So it
01:06:00-
May have been that specific report I just saw the additional statistic, or it may have been in section, I’m not positive, but it was all about literacy and proficiency of people, in ai.
Kristi:
And I saw a statistic talked about only, I think it was something like 18%. Even though, in general, like the statistics were like in the forties and fifties about how people were using ai didn’t mean they were proficient,
Kristi:
But then that female executives, female professionals, there were about only 18% that felt comfortable using ai.
Kristi:
And I think only a small percent like that, that felt comfortable using it at work at all. And it made my soul just scream. and I understand that in some facet of you don’t want to take that risk. And
01:07:00-
Especially with all the other things that I hear people say of that they’re using the free version or that they’re, and as a female executive, maybe you don’t want to use something that might be wrong and then oh, by all means make you look, I say, stupid jeopardize all the work you’ve done.
Kristi:
To compete in a male dominated industry or whatever the thing is that’s making you not feel like you want to use something that’s maybe uncertain or maybe you’re just not tech you, feel like it’s a tech thing. And I’m not a techy person because tech’s usually seen as something that’s male dominated.
Kristi:
I don’t relate to some of that because I’ve been in tech, my whole career. But, AI is made to be easy to use. It’s made to be simple. Even all the AI products, the no code stuff, you don’t have to code, you don’t have to do anything except talk. You can talk to it and all of a sudden have stuff done for you.
01:08:00- Kristi:
It’s a miracle for some people. it’s the way they act. it’s something that I just immediately felt this People who don’t understand ai, don’t understand tech. And women, they feel, and I hate the woman excuse, but for some women, I do think that they feel, a little bit less.
Kristi:
Experimental when it comes to something like AI that could have a bad connotation to it, if it by chance put them at risk. I just immediately wanted to say, it’s to anybody. There’s some, there’s men out there that probably feel the same way, so I don’t really think it’s a woman only, but it was the woman’s statistic that was shown.
Terry:
What’s really interesting in, I think it was about three days after chat GPT was launched, I was on a flight from Melbourne to, Brisbane. It’s a two hour flight, and I sat next to an executive, a female executive, at one of the
01:09:00-
Large organizations here in Australia. And, I noticed that she was preparing a fairly detailed report and then a long PowerPoint presentation and she was backwards and forwards between chat GPT and what she was doing.
Terry:
And, that I must admit actually inspired me ’cause I hadn’t even opened chat chip PT at that point in time. But that inspired me. Then when I, landed to, the first thing I did was actually go and download the app and get access to chat GPT. It was the first time I think that I’d seen, I’d heard about it.
Terry:
I knew about it. I obviously know what machine learning is and knew what AI was and had some experience with it in the past. But that pivotal moment of being able to physically chat with,an LLM that was mind blowing. that just changed everything. And you’ve spoken today a couple of times about.
Terry:
Having it as a strategic advisor or as a business coach. what I saw her do on that day, which was, what, two and a half, three years ago now?
01:10:00-
Well, longer than that, just absolutely blew my mind. And I thinkthat launched my journey then with, with chat GPT to start with.
Terry:
And then the chat LLMs ever since, what’s the one thing every exec should stop doing this quarter if they wanna avoid a headline level mistake?
Kristi:
Letting everybody use ai, just it with no strategy and no use policy, and just thinking that’s okay
Terry:
Great answer.
Kristi:
Because that endo training, because that, that just means that they have a lot of risk that is just increasing constantly because they don’t, they haven’t trained their people how to use it and what to and not to do.
Terry:
So you’ve written a book Power Moves and it lays out seven core strategies. Which one’s the most misunderstood or underutilized, of those seven strategies?
Kristi:
Oh, that’s hard For me, it is really hard for me because I want to
01:11:00-
Immediately say the mindset because I, think it’s the most misunderstood and it’s the first power move. it’s the round setting, gaining an AI mindset and how important it is.
Kristi:
And it’s misunderstood because I don’t think most have really grappled with the fact that this is so transformative and it’s so different. You know, I’ve explained it to a few as, and this comes from other people that I’ve heard try to explain it. it’s like a perfect storm. a blend of the industrial revolution with a little bit of Bitcoin mixed with the Renaissance and the impact that each of those had, and then multiply it by 10 times, a hundred times this is going to have is starting to have and has had it the Industrial Revolution did by bringing in
01:12:00-
Machines and just changing how work was done is what it’s compared to, I think most, most commonly.
Kristi:
And then there’s the, there’s a little bit of the Bitcoin because of how people are gonna be making money from this. But the renaissance in, in the innovation and the creativity that it’s going to just and is already exploding with it. But it’s like nothing we’ve ever seen before. And I, there, I sat and heard a pundit that’s listened to by a lot of people just about a month ago, sit back and say that AI is not a big deal.
Kristi:
I mean, sure. I’ve, been in product management and product marketing and through all of the marketing chain up to CMO, that, certainly I guess that’s helped with framing how to communicate and, it through all of those jobs, sales enablement training, I’ve held
01:17:00-
All of those roles.
Kristi:
So it all kinds of comes to fruition with this, the importance of all of them and handling change management. I’ve helped during m and as I, the understanding the pieces and parts and the importance of all of them. Yeah. I wish AI didn’t require all of them, frankly.
Kristi:
A lot easier.
Kristi:
It
Terry:
Absolutely.
Kristi:
It requires to have the communication plan, the change management plan. It just does because it because of what it really is.
Terry:
And it’s organizational change. You need all of those things if you wanna drive any type of change for any organization. Look,if you had 10 minutes with a room of B2B marketing and sales leaders, what’s the one big idea you’d share that could transform their results right now?
Kristi:
I would go to sales first. I would get the sales guys first, I would talk to them about building personas of their clients and building out, role playing their decks with their personas
01:18:00-
And using chat GDP to help them fine tune their presentation to a point where they walk in and they nail the sale.
Kristi:
That’s probably the first thing I do, because none of them, they may be going in and nobody that I’ve talked to has been doing this. they may be playing, doing their, helping with their decks or their pitch, but they’re not building the persona of their client and using that persona to basically to practice and do the role playing with, and then you going back to it to say, what did I do right?
Kristi:
What did I do wrong? What can I do better? Feeding it all the information.
Terry:
So you could create a custom GPT that actually is a customer.
Kristi:
Yeah,
Kristi:
And is that customer you can use. And many salespeople have known these people, customer for years. So it can go well beyond just, a couple, you know, recordings off of Zoom. But you can use, LinkedIn profiles, you can use a whole bunch of other information that the salesperson knows about the person.
Kristi:
Sometimes you’re selling to a friend, you’re selling, I mean, there’s any of that
01:19:00-
Stuff, but you can create a persona, pretty in depth persona, a really in depth persona with some of the other, tools many of them have access to.
Terry:
So that’d be useful for marketing then too, in terms of messaging as well, wouldn’t it?
Kristi:
Oh yeah. there’s all kinds of things for marketing that I have, but for sales, the one thing that would make quick, a quick impact immediately for sales and shortening the revenue cycle and may bringing in money is fine. Tune that presentation and know what you’re gonna say and not say that’s going to work.
Kristi:
And based off of their persona and how they react to things.
Terry:
Excellent. Excellent. Very good. I like that.
Kristi:
Yeah.
Terry:
A custom GPT, that is your customer, that customer persona. What’s your prediction for where AI won’t deliver as much value as people are hoping, especially in go-to-market?
Kristi:
I think agents are going to deliver a lot of value and they’re already starting to, but I think what’s going to be interesting is that
01:20:00-
They’re not easy just to deploy and create,
Kristi:
Something like in a Microsoft environment, depending on what copilot does with their agentic may make that easier. And
Kristi:
The idea of it and the interconnected concept, how long that takes. We’ll see, but the AI agents that everybody’s talking about right now and that everybody wants to deploy one right now, a and there’s a lot of confusion between what is an agent and what’s not.
Kristi:
When I’m talking about an agent right now, what I’m talking about is. As a comparison, a chat, GDP that does multiple steps in a process
Terry:
And executes.
Kristi:
And delivers an output me not having to worry about it. And I just say, Hey, go do this. And it goes and does the multiple steps and doesn’t have to bother me. and at the end it’s emailed or entered into a spreadsheet or it’s calculated, done all the things that it needs to do, and it’s not having
01:21:00-
To ever check in.
Kristi:
Maybe if I built in the check-ins, it comes back and checks in for and for me because that’s, the human in the loop, the things that you probably want happening. A lot of people don’t think about that with agents, the building of the agents, I think there’s a lot of and process that needs to go into them.
Kristi:
Where does the person in the loop come in? What does it and can it have access to that will make those take a little bit longer? They may eventually deliver what everybody wants, but I think they’ll take a little bit longer. You know, just the fact that chat, GPT now is given access to your Google Drive.
Kristi:
Well, do you really want it to have access to your Google Drive?
Terry:
Yeah, that’s right.
Kristi:
I mean, I kind of do because I want to have this one folder analyzed, but I can’t just isolate it to the one folder.
Terry:
Oh, and that’s the trick, isn’t it? You really do wanna be able to do it folder by folder so that you are effectively managing the knowledge bank that you’re providing it.
Kristi:
Yeah. I could
01:22:00-
Say just this folder, but as soon as you give access, it has access. So
Terry:
Yep.
Kristi:
Really mean? And those are the things you gotta think through. And it’s the same thing with the, it’s more complicated with the agents because it just gets a a lot more complicated.
Kristi:
And because of the intricacies of everything you’re asking it to do is typically more and more integrations and more systems.
Terry:
And I think that there’s also the nuance of people not necessarily recognizing the difference between an automation and a true agent. You know, simple, not, well, I won’t say just simple, but steps in a process that it follows the same each time. Do this action, do that action. I mean, you can do that to the, today there’s tools if this, then that, Zapier,you know,they may, there’s all sorts of different things.
Terry:
In my mind, a true agent is one that needs to use a knowledge bank to make a decision in terms of what should its next step be. It’s not a natural flow through where it actually has to make,
Terry:
A decision and it’s potentially interacting with other agents to get information or to
01:23:00-
Provide information that in my mind, is a true, you know, agentic environment.
Kristi:
And in many cases you have a master that’s controlling the other agents. That the master agent is the one that’s making the decisions of telling who to go where, what to do. And that’s what I meant by when I said, I’m not messed with, ’cause the master agent is making all the decisions.
Terry:
Yep.
Kristi:
So when you talk about human in the loop, where does human in the loop, and this is a question I’m throwing out, where does human in the loop belong with an agent? The master agent’s making the decisions make sure the other agents know what they’re supposed to be doing. The master agent is the one that I would ever want to deal with.
Kristi:
I don’t wanna deal with the other agents, I just wanna deal with the master agent. So what does that make the workflow really look like? So there’s
Terry:
I think personally, and to answer how I’m, where my head’s at the moment in terms of how you manage that is that, there’s trigger points and there’s checkpoints
Kristi:
Yes,
Terry:
For that master agent and even the master agent itself. There should be two in my
01:24:00-
Mind, there should be two, o overlaying agents of governance.
Terry:
There should be one, a compliance, which makes sure that agent is following the instructions that it’s been given originally. And that’s the only overriding. And then the other one is the, behavioral or ethical, component to make sure that you’ve got one, that it’s doing what it should be, and two, that it’s doing it in the way that it should be doing it.
Terry:
And I think they’re two very separate. And your feedback is from the master agent and the two supervisory agents, so to speak. They put the guardrails around this thing to make sure that it works, and people don’t even think about that other aspect of it.
Kristi:
Yeah. And see, and wonderful that I’m talking to someone that’s really thought through some of this, because those are the things, when I think, when I say things like think thinking, ai, you’re way more down that path of thinking in AI than most executives are, or most, IT people or, and I love my IT guys, but most of the people are saying, Hey, I
01:25:00-
Want agents, or let’s, or even implementation people who are jumping the gun and saying, I wanna build you an ag agentic, blah, blah, blah, just because they’re excited about N eight eight.
Kristi:
Eight NA, whatever, n eight n
Terry:
And I didn’t yet.
Kristi:
They wanna build the agent because it’s cool and they just wanna hand it to you because don’t want, they don’t have it, they don’t own it. They’re gonna code it for you and you own it. And you have to worry about the processes after the fact. And they may not know the questions to ask you to prompt responses because they’re a coder and they’ve never had, they too don’t know exactly the right things to ask.
Kristi:
So it just depends on, what may not deliver, I think those may not deliver. Because similar to Gen ai, a whole bunch of people jumped in at first and didn’t know everything they needed to know. I think my reasoning for saying AI agents, ’cause a whole bunch of people are gonna jump
01:26:00-
in and code and ask for and do way before they’re ready to answer the right questions or ask the right questions to the ecosystems what they really should be.
Terry:
I, think it’s one of those things that we all want, but I think as we’re all learning, we don’t know. I think some people know. But there’s gonna be a lot of people diving in that, that aren’t gonna be prepared to do the, ask the right questions or do the right things. And it’s not their fault.
Kristi:
It’s that we’re all learning.
Terry:
You know, we sit in this live chat space, particularly around lead generation through websites. So we sit in between marketing and sales. And the way that we run this business is that it is 100% reliant on delivering value If we’re not delivering value, then, and if we’re not delivering value, then that usually means we’re destroying value or it’s an opportunity cost for the organization.
Terry:
And there has to be a lot of governance around what we do, and we’ve got all of that built around our human beings. We do it with people. It’s not done with ai.
01:27:00-
We do have some AI that we are using now to help with the governance side of things in terms of supervision, in terms of suggested training paths for particular individuals based on the analysis of their own, behaviors.
Terry:
The really interesting thing for me has been that we’ve been, it would be. dollars in the bank if we just switched everything off and went to AI and we could do everything with ai, we’ve got a decade of experience. We’ve got millions and millions of data points to train the AI on. All anonymously, of course.
Terry:
But in terms of what actually works in getting these conversations right, in terms of understanding the knowledge about our client’s products in terms of understanding what’s important to those website visitors and the questions that they’re asking. We don’t have AI anywhere interacting with any website visitor, and I can’t see that we will in this foreseeable future, not for our existing clients.
Terry:
We might put a different offer together that has, if you want ai, you can have it, but our clients are of the opinion that, their deal sizes are such that they don’t want the
01:28:00-
Risk of a chatbot or AI interacting with their customers at the front end. And I think that, it’s the reason we’ve got a business.
Terry:
That’s what gives me some faith moving forward. And I do think that there’ll be this time of, you know, AI fatigue is going to roll over us all and we’re just gonna actually wanna talk to people again. It’ll be a bit like, COVID and actually finally getting out to do some face-to-face stuff again.
Terry:
But, look, that doesn’t mean that I don’t think there’s not a place for ai. There is most definitely, but I have not yet. With all of the information, with all the data that we’ve got, with all the smarts we’ve got in our business, I have not been able to build, an agent. an agent to interact with website visitors, with the objective of giving them an excellent service experience, and yet for our clients, an excellent sales person to sit in there and convert that on the way through.
Terry:
Now, it’s, I’ve seen, there’s plenty of products out there at the moment. I’ve watched and I’ve tested probably getting close to a hundred now, I think. And I’ve seen some more recent ones with some pretty well known people
01:29:00-
On LinkedIn, posting about it. And it’s okay if it’s only a, you know, a couple of hundred bucks a month, offer, it doesn’t matter if you get a few wrong, but when you’re talking six and seven figure deals
Kristi:
These individuals are interested in when they’re coming to the website, that’s a different story.
Terry:
You don’t wanna be mucking around with something that, uh, you know, it gets eight out of, eight out of a hundred wrong. if it gets,1%, 1% wrong and that’s means, a one point something million dollar impact to the pipeline that year. Well, that’s not the sort of thing that you wanna be, running the risk on.
Terry:
So it’s been interesting. look, I’ve loved our conversation today. Thank you so much. and certainly for being the, the flagship, episode of AI Walk the fine line. I, is there anything, Kristy, that I should have asked you? that I haven’t yet?
Kristi:
Yeah, I don’t, I mean, I could sit here and talk all day about this stuff. I don’t think so. you know, and I’d be happy to do another episode once you get further down if you wanna do something else Again, I, can’t wait to see what you do and keep seeing them.
01:30:00- Terry:
Thank you. Thank you. Look, I think one of the areas that I would like to dig in a bit further with you moving forward maybe it would be good to have a conversation about building proper frameworks around AI for the governance aspect. I’d like to have a conversation about that later on because I think that’s somewhere that,if there’s one area, or one question That leaders should be asking before, they integrate with another AI tool is, you know, what’s the government governance aspects here?
Terry:
What’s the security, what’s the safety aspects and what are the guardrails and what are the trigger points? And, how do we know that it’s gonna do exactly what it promises to do? Because I’m yet to see, I’m yet to see any AI deliver exactly what it’s been promised to do.
Terry:
And I think in the go-to market world, some of the worst things I’ve seen, there have been these AI SDRs, doing outbound cold emails at bulk, that work for a little while. it’s interesting there, there’s some that are quite good to start with. But they soon peter out, it’s like they don’t have the ability to either modify on the run or they get into
01:31:00-
Some sort of looper.
Terry:
I’m not sure what it is yet. I haven’t explored it deeply enough and I’m not really interested in exploring it deeply from, that perspective. I let others do it and they can stumble and fall four.
Kristi:
Yeah. Anytime that I’ve talked to my implementers, I haven’t one that was implemented, but anytime that we’ve talked about, when somebody’s wanted a chat bot, SDR, they’ve always said that they would have to custom develop it and they would use agents to do it,
Kristi:
and
Kristi:
They wouldn’t use one of the ones off the, that’s ready-made.
Terry:
But even custom development. Well, I mean, that’s what we did. That’s what we’ve done.
Kristi:
I like, okay.
Terry:
And so it’s like even when you do that, it’s,
Terry:
It still comes down to the LLM that you’re using. It still comes down to the LLM that you’re using. What it’s learned, I mean, there’s many tighter guardrails you can put in, but anyway, I mean, look, that’s a conversation for later, but I think that, the, my position is pretty simple. If you wanna use AI chat bots, then you know, low value, high volume, e-commerce type stuff, t-shirt shoes.
Terry:
At the other end, high technical, low
01:32:00-
volume stuff, put in a Cisco route manual, then let an engineer chat with, it using keywords all day long. but I think, you know, anywhere there’s emotion or it’s a high value, sales conversation, then at this stage, Yeah.
Kristi:
Okay.
Terry:
It wouldn’t be wanna keep it human in the loop there.
Terry:
Kristi, Thank you so much. I’ve enjoyed it immensely. I appreciate you. being on the chat today and
Kristi:
Sure.
Terry:
We’ll chat again soon.
Kristi:
Absolutely.
Terry:
Okay. Bye.
Terry:
For now.
Kristi:
Bye.
2026 © All rights reserved Chat Metrics
Made with ❤ in Melbourne, Australia
