AI Announcer (00:06.924)
This is Prompt This, the podcast for business leaders who've had it with AI hype and want the real deal. Clinton Greg, cut through the noise to bring you analysis and playbooks on using AI to launch new ideas, scale your business, and stop getting left behind. Now, here are your hosts.
Welcome to the podcast. I'm Greg. My cohost, Clint and I started this podcast to explore how business leaders are using AI in the real world. We found that our friends and colleagues aren't just talking about AI, they're actually building with it. So we decided to bring them in one by one to share their stories. And that's prompt this.
And I'm Clint. So I founded a successful customer relationship management software company years ago where we delivered AI powered CRM tools. Recently, I wanted to get deeper into today's AI, the most amazing new tech out there. And so I started asking my friends what they were doing with it. And now I'm a podcaster interviewing smart business leaders to unpack how they have started using AI in their companies.
And again, I'm Greg, after years of running high performing sales teams that led with AI, I too am now a podcast co-host. I focus on helping sales and marketing leaders accelerate the use of today's AI to deliver record breaking results.
Greg, what did you think about that recorded intro of ours just a moment ago?
Greg (01:37.324)
Well, when we first did it, I was really excited about it. I thought how cool was the process and, and, it sounded neat when we layered it over some, music. But then the more I listened to it, you can really tell it's an AI voice.
Yeah, it's a little bit too fake. I think we should try something different in the future. But hey, we were playing around with AI and voice technology and we learned a lot through the process. And I still think there's some really cool ways to apply AI into voice. Speaking of voice and AI, we've got a great guest today who's going to dig into that topic even further.
Yeah, looking back on today's episode, I mean, we're really going to learn some things about taking voices of customer conversations and what AI can do with it. And it's fascinating.
So be sure to stick around with us until the end of the show where we have our AI challenge where you learn more about how to use AI in your business. And with that, let's jump into the podcast.
Clint (02:40.248)
Hey everyone, welcome to the podcast. I'm Clint. We have Charles Hicks, founder and CTO of Reliable CX, a strategy consulting firm helping businesses bring AI into their customer interactions. Welcome Charles.
Greg.
Charles Hicks (02:53.87)
You guys have great voices, by the way. I'm feeling soothed just listening to y'all talk.
I'm excited because it's great when we have a high caliber guest here. And I got to ask, what made you say, I got to build a company that combines customer experience and AI.
I've really been a student of customer experience over the last 15 years, think, realizing that I could make my living kind of...
critiquing companies, but also helping them and kind of digging into what are those hard problems and why do they exist? There's so much interesting wisdom that you find out about organizations and people and systems. But I was also fascinated by the idea that AI kind of allows us to automate the mundane and humanize the exceptional.
And that's more of an idea than a promise, I think. But...
Clint (03:58.67)
hold on, hold on. Automate the mundane and humanize the exceptional. I like that. I just gotta roll that through my mind there for a moment.
customer experience is something we all get to partake in in some way, shape or form. I'm sure we've all had good and bad experiences and I've always been fascinated by what were the underpinnings that made those happen. so...
Customer experience and AI, I think, represent a really unique opportunity to take the boring pieces out of interacting with a company or make them more interesting in some way. We started reliable with the idea that we could actually help companies reimagine some of their operating models around these new technologies in a way that dramatically changed the quality of life for their employees and their customers.
You got some practical examples in there you can walk us through, like what are some common solutions?
we record and transcribe.
Charles Hicks (04:59.564)
almost all of our internal and external calls, our internal meetings, our external meetings, we pull next steps and follow ups and tasks out of those using the tools that we have in our business. And then if we need to do things like generate customer quotes or account summaries or strategy documents about an opportunity, we just take those transcripts, we put them into a Google Doc, we pump them through ChatGPT and we ask it whatever questions we want.
to get answered and on the balance it's better at analyzing those and answering those questions than a human is. And so that was kind of surprising the first time you do it you're like, this isn't going to work. And then you get the results and you're like, actually this is better than what I would have done. That's kind of…
faster.
and heck of a lot faster and heck of a lot cheaper. There's definitely areas in your business where you can get better results for about 10 % of the cost. If you can run it through these models and these models use it correctly and they understand the context and all that stuff. And the accuracy is above human level. And it's kind of finding those use cases that make sense for your business that I think will see that
foothold around AI. yeah, like starting around communications, just doing smart stuff with transcripts. You know, I'd say just start there. There's a lot of brainstorming, there's a lot of exploration that you can do. You don't need to buy a lot of software. Just play with some transcripts, some chat GPT, and kind of see what you can get.
Greg (06:36.61)
I can imagine, you know, whenever I've run a customer group or anything like that, what I always realized was just there's customer data everywhere and it's never in one place, no matter how hard you try, but there's like 10 times the data everywhere else. And I'd imagine AI could finally, you know, help you do things with the data that's everywhere.
Yeah, and I think this is where it gets really interesting because the rate of technological progress around AI tools, we can break it down very quickly, but when you start to break it down, you realize the arc of where things are going and I think where we're going to be in the next three to five years. just to back up, most of these AI techniques were developed in the 50s. They've been around for a long time. We just haven't had the confidence
computational power or the training sets or the data sets to make the most use of them. Over the last 20 years, we've created an immense amount of data online and we've basically trained these basic machine learning models how to understand human speech and language. That was kind of the first major step. Then after that, they actually developed a deep understanding of like if I give them a paragraph of text,
it can tell me what that paragraph is really about. So that's natural language understanding. That was kind of the next step. if you think about that, we taught computers how we think and represent ideas. So that was step two. And then last year when DeepSeq came out and everybody was kind of losing their mind, the subtext of that was really that DeepSeq had figured out how to train reasoning models without spending a ton of money.
And so the LLM... Just hit that real quick. question. So the fact that if you want to do 2 plus 2 as a math problem, and you want to feed that to chat GPT, it needs to use a reasoning model to solve that.
Clint (08:33.87)
It's a resting model.
Charles Hicks (08:47.298)
problem. So you can kind of divide the LLM up into kind of general understanding of human language and thought and how those ideas get kind of managed and grouped and then how it can actually do tasks like solve a math problem, find the dinner recipe, etc. Those are kind of the two major halves of it right now. so reasoning becomes interesting because reasoning allows you to then break bigger problems down into smaller problems and then give more of those problems
over to the AI to solve them. And so when we look at fitness and commercialization of reasoning, we look at the types of tasks and the duration of tasks that those models can support. the punchline is, two years ago, Chatchit P.T. could barely do math. Now it has a PhD in chemistry, science, law, medicine.
and it can...
understand that literature, that research, those diagnoses, and give you some feedback or input or even some basic decision making capabilities around those questions. Not to say that it should be trusted in all cases, but the point is that from a capability perspective, the AI models are accelerating quickly in their understanding of deep kind of human sciences and
How you find this? know maybe maybe in Clint's companies, everything works perfectly. But in some of the companies I've worked in, some of you talk about it's good at problem solving. Some of the biggest problems I've ever. Faced in internal company problems, issues, attitudes, things like that. How do you find it in the business world? mean, you know, there's a whole bunch of problems that that.
Greg (10:44.864)
I wish it was around when I was solving.
Yeah, it's a great question. just even back up and share some nuggets of how I use it for solving business problems real quick. So I'll use ChatGPT as a strategist. I go chat with it. I feed it context about the companies that I'm interested in, their financials, their...
their initiatives, the things that I know about them. And I ask it to answer questions and help me think through different approaches. I think on the balance, it's like having access to an expert in that area just at your fingertips.
Research Assistant, sounds like what you're saying? Hey, shifting gears real quick, so where does this take us with jobs,
Very much so.
Charles Hicks (11:35.926)
I think a lot of people are going to lose their jobs to AI, I think that the nature of work is shifting.
First off, don't think anybody's safe. I don't think it's low level people versus high level people. I can see AI replacing executives, mid-level managers, low level employees, all across the board. Everybody's under threat. I think if...
If your job has traditionally been clerical in nature or where you're just moving information between groups, mean a lot of mid-level management is just moving information around between groups, I think that function of those jobs is going to decrease over time. So I think the nature of the work will shift. I think it'll shift more into management roles where people are managing these AI systems. They're supervising them, they're training them, they're tuning them, and there'll still be a lot of people required to do that.
As we've done contact center projects where we've brought in
virtual agents for handling level one issues. Ultimately that leads to a reduction in that contact center staff. We've had pretty good success helping companies re-skill those employees into knowledge management roles because the knowledge that you feed these systems becomes really important. So you need to curate that knowledge and feed it a good diet of information so you need people to manage that. And then the training and the coaching, like we used to
Charles Hicks (13:11.632)
schedule one-on-ones with our bots weekly, just like you would an employee. And the one-on-ones, the bot a lot of times will bring up issues it's struggling to respond to or things like that. So you end up kind of coaching and managing these bots to a certain degree as well. I don't think the jobs are just disappearing. I think they're shifting. There's definitely some that are going away, like I think probably 20%.
of like tech jobs or whatever I think will probably go away. just to give you some early indications, there's been studies done on efficiency gains for using like Claude or Cursor for software development. And when I first started playing with it, I was like, this is going to slash development costs by 90%. But the real numbers are closer to like 25%.
These tools, while they're great at making big changes to your code base, they can get lost and struggle with a lot of little things. And so you still have to kind of fine tune the details and really kind of manage these things in a different way. So it's going to be an impact. don't think it'll be as big and as scary as everybody thinks. And I think the nature of work will shift and it'll be more interesting to a lot of people.
Yeah. You know, one thing Clint and I talk about a lot, I was just curious what you're seeing since you're talking to so many leaders out there. Like what are the biggest traps you see them falling into with, with AI?
There's a lot of people that think that, so it's fun to talk about. So there's a lot of talk and not a lot of action. And I think, know, I've noticed this pattern where sometimes people, their identity gets wrapped up in an issue.
Charles Hicks (15:00.478)
And so I think one of the traps executives are falling into is, my gosh, this is so difficult. We're never gonna figure it out. Let's just keep talking about how difficult it is, right? And so, you know, I'll just point to the fact that there's real data out there about how companies are adopting and finding success with AI. It's not this open question. And if you're just not looking for it, then you're
you're kind of practicing willfully, a willful ignorance is, is maybe a nice way of saying it, right? It's like, it's like the, I love working in tech. It's like pull your head out of the sand and just look around. Right. But, but, but like, the, the rate at which these conversations move is accelerating and people kind of have a
way to say it.
Charles Hicks (15:57.44)
a time block in their head for how long they're gonna dwell on something. so the rate of innovation means that the conversations are evolving quicker, the learning is happening faster. The way I think about it is like, we just released a new video game to the world and everybody's trying to figure out how to play it.
Right, and there's some kids that are, you know, they're up 23 hours a day playing this game so they can write the strategy guide and then there's other ones that, you know, that are just playing it here and there. And, you know, so they can talk about it. And so I think, I think you've got to kind of get serious about it because it is shifting operating models.
So in there, I'm hearing, don't get stuck in an analysis paralysis, just do it. Okay, got that. Now, let's say I'm getting started. I'm a CEO of a, I don't know, $50 million manufacturing firm and I'm getting started with AI and I'm recording customer calls and that sort of thing. What's another trap that comes up after that?
I think the next trap ends up being like you're going to find information in those calls that doesn't jive with your understanding of your business.
you have to kind of get to the bottom of it. Is it your understanding that's wrong or is it the interpretation of the calls that's wrong? In my experience, 90 % of the time, it's that person's understanding of the business. so being open to just being open to learning that you're gonna learn something you don't know and probably makes you uncomfortable, I think.
Clint (17:30.478)
We're back to that ego topic. Yeah.
You find that leaders want to get into AI and kind of want to bring this in the business or do they feel like they have to? And in the engagements, you're kind of getting that feeling where they're like, all right, we're at this point, we got to do this. Like I've seen two kind of engagements when we bring people in, when I've been in there. And I'm just curious, I'd imagine people would be excited about this, but what are you actually seeing?
I see that most of the executives that I've been working with, they're fairly scared that they're going to make a big mistake that's going to have cascading penalties like a regulatory thing or something, some bigger time bomb that they're not aware of.
I hear people throughout, like, security and hallucinations. The people that aren't, that are trying to find a reason to not do it seem to come back to those two reasons the most frequently.
Yeah, and I think there's an even more fundamental threat to businesses, which is as these reasoning models get better, right, if your business is just a variation of a public data set that ChatGPT or Claude can train on, that's a threat, right? So I think a lot of companies are also trying to figure out where's the safe space for doing AI.
Charles Hicks (19:10.88)
in their business that's not gonna erode their advantage or give away too much of their secret sauce to these other bigger vendors that are doing scary stuff with data that people aren't even aware of, right? So I think that's kind of the other area of uncertainty that's holding people back is they don't know if they're like mortgaging their future with these technologies. And I think that's just where you just have to look at
where the data sets are that are unique to your business that can't be easily consumed. And I think that's also why communications is a nice starting place for a lot of.
Hey, I want to shift gears here for a second and go down a different path. So, Charles, you've talked a few times about data in here, and you've talked about that conversational data that you talked putting it into the chat GPT to find out those threads. Tell me more about how companies should be thinking about data today.
data.
Charles Hicks (20:13.87)
This is another area I think people completely are kind of missing the mark. First of all, your CRM data is probably complete garbage. Or it's out of date, right?
We spend a majority of our time polishing it, so you won't say that. That's very offensive already. This is very offensive. Greg.
I have a question for you. What's the relationship between salespeople's performance and their ability to keep the CRM current?
Greg (20:52.034)
Well, we used to say things like selling is the commission side. CRM polishing is where you get your base pay.
Greg (21:05.934)
That's how, that's, that's how we operate. that, the worst your CRM data is in a roll up to the executive team, the more noise they bring down and get in the way of us selling. let's do this. Move forward and then go do our, go to what we need to do. And that's what it sounds like from a sales floor. So I love the way you asked me, let's ask Clint.
Well, he thinks of CRM data. I think he has a little expertise in that area.
always found that the sales reps that keep a tidy room are the best salespeople in the end. The people that keep themselves organized and on top of things and use CRM as a way to keep that organization in place. And yeah, the side benefit is to educate upper management about what's going on in their little piece of the business.
Alright.
Clint (22:02.094)
But that's just kind of my take, but I, you know, I've got a whole career building serum tools. I find I kind of, fall back on that as a nice, like, Hey, if you use my software, that just shows you're really, really organized and you're going to be better at your, at your day job. And that just kind of sounds like kind of a, bit of a rationalization. And this, uh, this has been a great conversation. I'm enjoying every piece of this year. Last topic. What are your favorite AI tools that you think people should be paying attention to you right now?
Yeah.
Greg (22:32.141)
yeah, yeah, this guy, I think Charles knows.
Yeah, so, okay, I got a whole bunch.
Cool. All right. Let's hear this toolbox.
I used so I use chat GPT for basically creative thinking business strategy work Kind of branding ideas and stuff like that I I find it's very helpful and kind of allowing me to organize my thoughts into something that I can it helps me break down my big ideas into smaller ones and then I'll use lovable
which is, it's like, if you haven't played with lovable, you should just go to lovable and check it out. It's like a chat GPT but for websites. So you can just type in a website that you wanna build and it'll build it for you. So for, yeah, so for iterating on concepts,
Charles Hicks (23:30.158)
You know, instead of having to do it in my head or draw it out in Figma now, I just go build it in Lovable. And then I show it to people because that just cuts down on the feedback loop. So, you know, what used to take weeks to kind of communicate an idea, I can do it in minutes now. So think about that. I haven't found a good way to build presentations yet. I know there's some good tools out there, but that seems to be an area.
For me, the areas that I haven't found good tools is presentations and spreadsheets. I also like perplexity, Claude, you know, any of the tools with deep research are pretty nice. Especially if you have open questions, like, like if I'm doing market research, most of the deep research tools. there one that you use that,
Anybody can just sit down and start working with and playing with tomorrow?
I tell you what, I've really enjoyed like Sora and the whole image generation and video generation piece. Combining that with Lovable, I've been building kind of video game concepts with my own art and all kinds. The range of expression of what you can put together, like bringing your ideas to life with these tools, I think is really the coolest.
Greg and I have been playing around with song generation sites like, know, and and then playing around with AI voice. Voice interaction like, yeah, the Levin labs.
Greg (25:05.88)
Yeah, the voice is fun. Yeah, we had a good time with that one.
I've done the AI headshots. That was actually a really good experience. So I use Sector. Sector.ai. I think it's like 60 bucks and they give you like, I don't know, thousand headshots or something like that. But as somebody who hates taking pictures and like going out and dealing with a photographer and getting dressed and all that crap, really cool.
Now what tools are?
Charles Hicks (25:35.136)
I would encourage everybody to do it. It's just of fun activity to go through. And you can kind of see what that user experience is like. It's kind of a different, it's a little bit of a different user experience from just a traditional product. So yeah, I would say Sektas, cool.
Most of my stuff has been kind of those tools. We're also a Zoom partner and Zoom has an AI tool as well that's built around their platform and I found that one to be quite useful, at least in terms of working with conversations on their platform and stuff like that.
As we get ready to wrap up now we have a segment that we will be doing in each of our podcasts called the AI challenge.
The AI challenge is a takeaway assignment for our audience to begin to get comfortable playing around with different AI tools. So Clint, how does the AI challenge work?
Well, keeping in the theme of today's discussion topics, we actually have two AI challenges for our audience. We have blog articles on the promptthis.ai website that describe the step-by-step instructions to learn how to use these AI tools we're talking about. See the show notes below for the links to the challenge instructions. Greg, what's the first AI challenge?
Greg (26:49.038)
Well, today we heard all about using Chat GPT to analyze recorded customer conversations to discover insights like what your customers pain points are, which competitors get mentioned the most, you know, things like that. So what you're going to need to do is have access to a few Zoom recordings to get started here. Upload the recordings, prompt Chat GPT to analyze it and see what comes out.
Nice, that'll be a good one. The other AI challenge is to refresh your LinkedIn profile image with a headshot that's newer than that picture from 2015 that you still have up there. Well, what about you, Greg? When's your LinkedIn profile headshot from?
Hmm, probably about that 10 year range.
Yeah, time for a new one. Time for a new one. So you two, you've got to take away Simon yourself. All the link in the show notes to step through using sector to create an AI enhanced headshot.
Right.
Greg (27:46.584)
Well, it has been fascinating and thank you for being here.
appreciate your time today, Charles. What's the best way for our listeners to track you down if they want to follow up with you and learn more about what your company's doing? Sure.
So I'm on LinkedIn as Charles Hicks. We also have website, a reliable.cx. And you can see what we're all about there. Thank you guys. Yeah, cheers.
Thank you, Charles. You have a great day.
Thanks for joining Clint and Greg today. You can find all of the Prompt This episodes and more in-depth articles at www.promptthis.ai and be sure to click the follow button below. We look forward to having you back.
