The AI Paradox: What AI Reveals About Us w/ Ty Givens | S2E10

Full Episode Transcript

This is the full recorded conversation from The Human Protocol Podcast, published by The Human Protocol Group. © 2026 The Human Protocol Group. Please link to and attribute the original episode when referencing this conversation.

THE HUMAN PROTOCOL (00:01.39)
Hello everyone, welcome to the Human Protocol, the podcast about staying human while building the future. I am your host, Mykel Salomon. Today's episode is a hard truth for leaders because the biggest lie in the AI era is this. The AI is the solution for everything. It is not. AI is really a mirror. It reflects your system, your ownership, your decision making.

your culture, if your workflows are messy, guess what? AI will make them louder. If your escalation path aren't clear, AI will make them visible. If your team is burned out, AI won't fix that. It will expose it. So my guest today is Ty Givens. And we're going to talk a little bit about what AI adoption reveals on CX and support leadership and how to use that revelation.

to rebuild the function the right way. Ty, my friend, welcome to the Human Protocol.

Ty Givens (01:06.338)
thank you so much for having me, Michael.

THE HUMAN PROTOCOL (01:09.108)
Awesome. It's a pleasure. looking forward to having this conversation today and to kick things off time. I want to ask if you can give us a little bit of introduction of who Ty is for us to know.

Ty Givens (01:22.142)
Yeah, sure. So, personally, I'm the oldest granddaughter, the oldest daughter, so I have, I have, I think I have whatever those stigmas are that are attached to that are probably mostly true. I have worked in support.

in general, mean, since I was 16, but in support operations specifically since I was 19 years old. And I started out working in larger companies, Fortune 500s, probably 100s. I haven't checked the list recently, but this was, you know, like, I say all the depots, Home Depot, Office Depot. I worked at Intuit. I worked for Herbalife, I for Molina Healthcare. One of the things I learned about myself is I like startups a lot. So I used

THE HUMAN PROTOCOL (01:59.31)
You

Ty Givens (02:13.448)
head support for a shoe dazzle. That was a lot of fun, a lot of challenges. Also learned through that process that I'm more of a builder.

After joining Thrive Market in 20, I think it was 14 or 15, I worked so hard building that team and scaling that team. And I realized I never want to work that hard again for anyone other than myself. And I also wanted to give to

other leaders what I wish I had when I was doing that, which is the opportunity to lean on someone who has the expertise to help me build it out. So I don't feel like I have to do it all. And that's how CX Collective was born. So we had a name change, used to be called Workforce Pro, but 10 years in this space now, and plus the 15 of me, you know, doing the work directly as an employee. Now I do it for multiple companies in different capacity, and I love it.

THE HUMAN PROTOCOL (03:15.054)
That's the most important thing, right? That we do something that we love. Because then, even when it's complicated and maybe difficult, it comes easier than just being pushed to do something that you don't like to do, right? Yeah. Absolutely. Thank you for going over your experience. And let's get down to business, shall we? Right? So everything today is around AI.

Ty Givens (03:27.486)
Yes. Yes.

Ty Givens (03:36.7)
Okay, that's

THE HUMAN PROTOCOL (03:44.428)
Right? One way or the other. And we are in CX and support. We know the pressure that organizations are putting in our team to run operations in a more efficient way. Right? A lot of support teams and overall CX organizations are being asked to do more with less. Right? And I think that's the new constant as we move into this new feature. Right? So we're going to do more with less.

while customers still expect instant answers from us in a way. So Ty, tell us what you have seen out there from your point of view. When you deploy AI into support, what is the first thing that breaks?

Ty Givens (04:34.59)
Um, well, when we deploy it, not much breaks. But what I see, what I see that starts, you know, the whole confusion is really the lack of understanding. And this is not a critique. It's just the fact, the lack of understanding that AI is not going to fix your broken processes. It's

THE HUMAN PROTOCOL (04:40.696)
Ha ha ha ha ha ha ha

Ty Givens (05:01.658)
It's kind of like, used to, my career started in workforce management.

And I used to work with systems like IEX Totalview. I'm really dating myself. Okay, this was 200 years ago. Blue pumpkin and you know, EW film like an aspect of our product. so with products like that, the sell is always once you have the system, everything is easy. And I think that we're kind of being sold that same thing with AI. Like once you have it, everything is easy. But the reality

of any system or tool or product that you deploy, you have to take the time to step back to think about how do I need it to work for me? What am I hoping to get out of it?

Like how should it function? What are the limitations? You have to think about all those things first. And so what I find most when we go in to start helping companies leverage some of the automation tools that are available on the market today is that they think that the system is going to do all of that thinking for them. And it is, yeah.

THE HUMAN PROTOCOL (06:07.778)
Yeah, like magic, like magic.

Ty Givens (06:10.843)
Yeah, like, I'm just, we're going to launch AI agents. And it's like, right. You still have to tell them what to do. You have to check their work. There's a whole lot that goes into it. And I think that, you know, in this case, there's not like one company trying to push AI specifically. I think it's just all the stuff that people are reading and seeing and hearing about. they think this is easy. I should be able to do this. And it's like, it's not that easy. Now it will save you.

time in the long run, but you've got to take a minute to actually build it out to make sure that it is going to work in the way that you think it's going to work. And the key word here is thinking. You got to take time to do that.

THE HUMAN PROTOCOL (06:57.1)
Yes. And there is no way of sugarcoating that, We need to go through it. If not, it will be exposed.

Ty Givens (07:05.662)
Yes, yes. And it's funny, we have a client and for their sales and marketing side, because they're a SaaS platform. So within the platform, there's one version of their AI bot. And that is obviously based on functionality in the system and all of the...

the articles that are there. And then they wanted to do one for the sales and marketing side. So the one that brings the customers in. And so in the conversation where we're going, okay, you know, decisions that we have to make, do you want us to rewrite your articles? Do you want to rewrite your articles? Like, how do you want to proceed with this? And in doing that, they clearly use ChatGPT to write out requirements that they sent to us.

I could clearly tell that they didn't know what they were putting in and what came out of it was actually not very useful and I thought they're trying and I understand that and I get it but they still haven't learned that this is not a product that's going to think for you. Like you have to be really intentional about what you ask and something that I learned from some

someone else is sometimes you can ask AI how to write the prompt to get you what you need. And I was thinking that maybe that would have been helpful, but I think it is very telling to see that the outcome of what they asked for wasn't even useful. So when people are thinking that AI is going to just...

change everything for them it's like no it's gonna it's gonna expose what you know and it's gonna expose what you don't so yeah yeah yeah i mean

THE HUMAN PROTOCOL (09:01.964)
and what you did and what you didn't do.

Ty Givens (09:07.56)
There have been times like, you know, in my past where we might be working with a client and it's time to hand things off. So we're looking to interview leaders and, you know, in the interview process, I'll be listening to, know, what the answers are that they're giving me. And a lot of times it would be full of jargon and you know, things that are Google, Googleable, I'll call it, you know, things you can look up and write your answer. And so they say all the right things and all the buzzwords, but I see through that because I'm, I'm not a tell me what you would do. I'm

tell me about a time when you did, right? And when you do that, when you ask those questions, you don't get the same answers because a lot of times they're talking theoretically. And I feel like with the deployment of AI, a lot of people are operating theoretically because they're thinking that AI is going to do for them what they don't know how to do.

THE HUMAN PROTOCOL (09:39.992)
Yeah.

THE HUMAN PROTOCOL (09:59.662)
Figure it out, figure it out for them what they want.

Ty Givens (10:03.612)
Yeah, yeah, and I don't know if I could figure out a product that could tell people what they want. I would be very rich. So I don't know how to do that. I don't know a product that I shall. shall. Yeah. Yeah. So it doesn't work that way.

THE HUMAN PROTOCOL (10:10.603)
Yeah, of course.

Tell me when you find it, please share something with me at least, okay? Yeah, it's funny type, because the more we have these conversations, right, you on the field, me with other dealers and you with other dealers as well, the more that you see this formula repeating and repeating and you stop and think, and this is a small medium and large businesses, they all going through the same without putting in the work.

to try to implement the best and deploy the best AI product ever. And that means you and your people, you need to know what you want from it. And the why, the what, and the how is still on us.

Ty Givens (10:50.556)
Yes. Yes.

Ty Givens (10:56.306)
Yes, yes, yes. And I know that we can guide them to a conversation if the question is, what can we leverage AI for? We can help with that, right? The way that I describe it is think about, you know, call it 10 years ago, 15 years ago when outsourcing was like a really big thing.

Oh, well, maybe 20 years ago. Oh my God. I've been doing this a long time. I had to look at the calendar. I'm like, that's something in 2005. That's 21 years. Anywho, which is amazing because I'm 21. So how did we get here? But anywho, right? But back then, you know, when it came to outsourcing, you know, it was a

THE HUMAN PROTOCOL (11:36.382)
I know, know, tell me about it.

Ty Givens (11:49.36)
It was a big deal to get your logic-based support questions or needs handled offshore where it was less expensive. So you're building logic trees and if this then that. Or you have maybe a subscription-based company and you want to ensure that someone can follow the script without breaking based on emotion. A lot of different reasons to do it.

THE HUMAN PROTOCOL (11:53.966)
Mm-hmm.

Ty Givens (12:18.192)
with AI, we're kind of in the same spot. So think of AI as, in some ways, your outsource agent, right? Where you still have to help it understand what it is that you're trying to accomplish. And you have to take the time to decide what a success look like. And so those are the pauses that we all have to take to sit back and go, hmm, like.

I have to actually sit down and figure out what I want it to do because the C level, you know what they're getting and what they're hearing most of the time is going to be whatever is circulating on LinkedIn or wherever they get their information around how AI is revolutionary and saving companies millions of dollars. And then they're going to come to you and go, I want you to save us millions of dollars too. And you're like, okay, now if you're anything like I was when I was running teams, trying to

figure out how you're going to leverage AI is just one more thing you have to do. And you're already trying to be in the past, present, and future, and you're trying to work across people, process, and technology every day. So when they throw in one more thing, it's like, when am I supposed to figure that out? I'm just trying to figure out why we missed our forecast yesterday and got 40 % more calls and I've got 10 % of my staff out. You're tactical.

Right? So the idea of like, what could I use it for? That's something that people can help you with, but how you're going to use it is something that you have to think through. So I would say aligning yourself with an agency, a firm, a partner, a friend, whoever that can help you kind of process how you would want it to work is really key and critical to get the return. Yeah. Yeah.

THE HUMAN PROTOCOL (14:02.424)
map it out, visualize it, right? Yeah.

I agree. I agree. It's definitely what's going to define how the implementation will look like and what type of results you're going to get out of it. And thank you for pointing that out because I think we that are in support and CX, that we are running organizations and teams, sometimes we don't stop to think everything that we are already doing. And then we are getting

Ty Givens (14:18.897)
to present.

THE HUMAN PROTOCOL (14:36.322)
this added on top of everything else. And it's not only been added, it's like, we need this for yesterday. Because that's how organizations and leaders are really running this. OK, I want this, but it's not for six months from now. I want this ASAP. So you need to then scramble and going around and see, how and start putting the effort and how you can accomplish and get that done.

Ty Givens (14:38.771)
Yes.

Ty Givens (14:44.23)
Yes.

THE HUMAN PROTOCOL (15:03.338)
on top of everything that you're trying to accomplish for your team, for your customers, for your tools and everything in between.

Ty Givens (15:10.578)
Yeah, there's, I feel like the role of like leading and running a support or service team, it has changed so much like over the course of my career. And today it has a lot to do with knowing and understanding technology, being able to leverage it. And

THE HUMAN PROTOCOL (15:31.95)
Mm-hmm. Yep.

Ty Givens (15:35.866)
almost somehow keeping yourself abreast and aware of all of the industry changes. It's almost like people expect that if you are in support that you know what the latest and greatest is. And when you work inside a company and they have their tech stack, you know that tech stack. The only time that you're looking into another product or tool is if what you already have doesn't meet your needs. And then at that point you may step out and try to learn a little bit more about what's out there. But for the most

THE HUMAN PROTOCOL (16:00.611)
Yeah.

THE HUMAN PROTOCOL (16:05.326)
Mm.

Ty Givens (16:05.76)
part, we're just trying to keep our oxygen mask on so we can hand out the extras to everybody else. There's not a lot of time to think about like, is this good oxygen that I'm breathing in or am I going to pass out soon? We don't have time to think about that. so when we, and that's just a reality. Like one of the reasons why I wanted to go out on my own is I wanted to know what was there. Like I feel like it's my job now to know what is happening. And every day I'm still surprised. Doesn't that matter how much

THE HUMAN PROTOCOL (16:17.4)
Yeah.

THE HUMAN PROTOCOL (16:29.614)
Mm-hmm.

Ty Givens (16:35.68)
time I spend trying to understand new products, there's always something that I didn't see, didn't overlook, didn't know about. Had a conversation once a month, they're doing that now, And I imagine, darn, if I was still doing the day to day, I wouldn't even know that much.

THE HUMAN PROTOCOL (16:51.054)
Yeah, it's and we're seeing this more and more. It's it's becoming an imperative that leaders become technical. And I think the only way we will be able to continue to succeed is that if we really step it up and we become more technical leaders moving forward, it's not going to be just management people anymore. If you want to be leaders.

Ty Givens (17:02.642)
Mm-hmm.

THE HUMAN PROTOCOL (17:17.386)
I see this change happening very aggressive today, which middle managers being caught will at AI implementations and big companies like Amazon and others are cutting a lot of middle managers because they are consolidating functions and AI is replacing a lot of that. So for leaders, it's you being technical as well as being a good people's manager and having those skills is going to be

really needed moving forward.

Ty Givens (17:50.694)
I've seen, you know, old leaders that I worked under because I, you know, I started in support back when, you know, the leaders that I had, they were 10, 15, 20 years vets. Like this is the job that they were doing this with their career. So I learned support and support operations in a very different way than a lot of people today. I didn't go to a startup and my boss was the CEO and it's my first time running support under a CEO with no guidance. I came into startups with a lot of guidance and understanding of like how

support is supposed to function and operate at scale.

THE HUMAN PROTOCOL (18:23.694)
Right.

Ty Givens (18:27.386)
But, you know, I'll use Thrive Market as an example. When I went to Thrive Market, the job I was at previously was Seas Candies, which is very popular here on the West Coast. They ship to the East Coast, but, you know, they're very, there's no additives or preservatives in their chocolates. So there's a certain time of year where they're not gonna send anything, but it's excellent quality. Company's been around probably a hundred years at this point. However, you know, Seas Candies had, it's a huge company, Berkshire Hathaway Company.

and it has specific roles who are responsible for different parts of the function. Well, when I go to Thrive Market, I'm like, kind of dropped into, yeah, we've got Zendesk, and yeah, nobody here knows how Zendesk works. Our CTO chose it. It's not doing what we need it to do. Can you fix it? Me? Right? Like this is now your job.

THE HUMAN PROTOCOL (19:20.972)
I love this, you fix it?

Ty Givens (19:24.912)
I'm like, fix it. You know, I'm still trying to figure out why I can't measure like our call volume for the previous day because this is literally 2015, 14. So at this time Zendesk was reporting the last 24 hours, not the previous day because we weren't on the right product level to get the previous day. I had not been exposed to nickel, ling, and dimey at that level.

Well, mean, so but this is not my job. So guess what? I had to figure it out. And now we do dozens of deployments for Zendesk. But I do that from a place of understanding. Right. And I've seen like my old bosses take jobs as frontline reps.

Because the role of leading and running support is no longer just about being good with the people and handling escalations and being good with customers. It's now about programming systems and technology. you know, yeah, it's not the same job. And so it's really hard. And some of the things that I've been thinking about doing actually is putting together almost like another playbook, because I do playbooks a lot.

THE HUMAN PROTOCOL (20:21.486)
understanding how that works.

Ty Givens (20:34.408)
but a playbook that helps leaders transition into like, what does it mean to be the modern, the leader of 2026 and beyond the 2030 leader? What does that look like? Because it's different than the 2020 leader and 2010 or 2000s.

THE HUMAN PROTOCOL (20:41.134)
The leader of the future, yeah.

THE HUMAN PROTOCOL (20:52.962)
Yeah, Ty, and I'm seeing it more and more, and it's accelerating. We need to become more technical. There is not going to be any other way. If you want to succeed in your job as a leader of a support organization, for example, you need to understand your people, your teams, as much as your tech stack. Because put it this way, if we continue to deploy AI and AI continue to take over

Ty Givens (21:02.406)
have to.

Ty Givens (21:14.046)
Yeah.

THE HUMAN PROTOCOL (21:20.654)
50, 60%, 70 % in some cases, all of your operations, right? And then you're keeping humans really at the edges, keeping human just making decision. The majority of your operation, it will be AI driven and automations. How are you gonna continue to improve that if you don't understand how you get from zero to 80?

Ty Givens (21:23.933)
Yeah.

THE HUMAN PROTOCOL (21:43.5)
You know what I mean? This is why we need leaders to keep an eye on it. Listen, AI will continue to absorb a lot more of your operations that it is today. And for you to get to the next level, you will need to understand how that happens. If not, you won't be able to improve that.

Ty Givens (21:43.74)
Yes. Yes.

Ty Givens (21:52.734)
Yeah.

Ty Givens (22:01.786)
It's true. know, when I was like younger in my career, I used to get criticized for being too tactical.

and not being strategic enough as I've matured and you know got many more years under my belt. What I realized is part of the reason I wasn't strategic is because I didn't have a vision of what success looked like. So I can't get to strategically get to an end result that I can't see or fathom. So if anyone's ever said you're not strategic enough, nobody feel bad about that. If you've never actually seen the result, what are you strategizing to get to?

THE HUMAN PROTOCOL (22:26.094)
Hmm.

Ty Givens (22:41.568)
So let's start there. But the second part of that is the fact that I was tactical and that like the way that my brain works. So I have to make sure I shout out my man who hired me in the workforce management. His name is Tom. Hi, Tom. He's doing his thing right now. I'll make sure he sees this. Yeah, Thomas. He had, you know, we're talking years of the years ago, but he had this big giant spreadsheet.

THE HUMAN PROTOCOL (23:00.536)
Hey, Tom, please do.

Ty Givens (23:11.39)
where he taught us how to just record volume and A-H-T and all that good stuff, right? And I actually spent time learning the formulas that he plugged into the spreadsheets and taught myself Excel that way. Now, he wasn't one of the people who said I wasn't strategic. He never critiqued me like that. He was actually very supportive, but...

My ability to say, I'm curious about how this works. I don't understand this process, but I want to learn this process. What's something that would where, you know, I could be in a space of like touting theory. Even today, I could just speak theoretically, but I'm more practical and hands on because I'm more about let me open the hood and see how this thing works and understand why it works the way that it works.

That has actually served me well because when I go to a company that drops me in, was like, Zendesk is your problem to figure out. Guess what I had to do? I had to open the lid. And I made a lot of mistakes. I messed up. I wasn't saying that it's not perfect. It was a lot of trial and error. But at the end of the day, it was worth it for me to be that tactical and to not be so theoretic. And so I'm

THE HUMAN PROTOCOL (24:14.296)
Yeah.

THE HUMAN PROTOCOL (24:27.188)
Absolutely.

Ty Givens (24:31.45)
encouraging people to get curious about how things work. know, ask questions about how they work and why they work that way. Research it. Be more resourceful because I feel like being resourceful is something that's missing in today's world. Like we're so used to getting information handed to us so quickly that nobody wants to think through, research, understand, dig in.

THE HUMAN PROTOCOL (24:40.717)
Yeah.

Ty Givens (24:54.654)
Take a minute and instead of just asking your colleague, why does it work that way, you go figure out why it works that way. That's going to serve you better. And then as you move into, especially if you like startups, nobody in the startup is going to give you a blueprint. No one's going to tell you this is how we want it to go. You'll get the most instruction you'll get from the C level saying, want to really differentiate with the customer experience.

THE HUMAN PROTOCOL (25:16.312)
Fix it. Get me this, and you figure it out.

Ty Givens (25:23.206)
Yeah, yeah, yeah, we want to stand out, what does that mean, right?

THE HUMAN PROTOCOL (25:27.534)
Ty, I can't agree with you more on that because the more I see how things are evolving, it's going to be a lot more about execution. For you to be executing projects and implementation at a level that is needed, you need to understand how things work. You need to put in the work. It's not going to be,

Ty Givens (25:41.147)
Mm-hmm.

Ty Givens (25:54.75)
I'm

THE HUMAN PROTOCOL (25:57.26)
Delegate this and you do it for some cases. Yes, but if you really want to understand how your tool works how the process work and You probably gonna have to put it in in the work and and your brain when you go through that exercise You and I we come from from from similar backgrounds where where we we put in the work and we used to like doing things and figure it out things

Not everybody's wired like that, but it is needed. And it is an exercise that wires your brain on everything that you do to have that mindset to say, OK, there is a problem. Let me open this up. Let me start looking where the problem may be. OK? And once I get there, you're going to have a full understanding of how and why the problem happened. But that will only be possible if you allow yourself

to go through it with a mindset that says, you can do it, you can learn it, and this is needed for you to learn. And after that, once you know your process, the technology, or the tool, that's how you can start making and defining strategies to improve it. But going through it, no one can do it. You need to do it.

Ty Givens (27:11.742)
100%.

You got it.

Yeah, you gotta go through it. mean, sometimes people wanna go around things, but going through it is the best way to figure things out, even though it's tough. Every month, I do a free workshop for AI. And it's not advanced. I'm not saying I do anything like crazy, because I do it every month. It's like 30 minutes. But the idea is to just help people start thinking about how they're gonna approach it. And so I think it's the last Thursday.

every month. Like feel free to check it out because my goal is to like to give to leaders what I didn't have which is just like I needed I wanted to create a space where it was okay to say I don't know what to do and nobody feels bad about that you know it's okay if you don't know how to roll out AI you've never rolled out AI before you know that's fine.

THE HUMAN PROTOCOL (28:03.416)
Yeah. Yeah.

THE HUMAN PROTOCOL (28:14.294)
is that first step as we have been discussing today to allow yourself to say, I don't know it, but I want to try. But also this brings me to another kind of topic, similar but not similar. It's everybody today, especially if you're on LinkedIn, everybody sounds like an expert on everything. You know, it's open LinkedIn and everybody has a solution for everything. It's just like magic.

Ty Givens (28:23.208)
curious.

Ty Givens (28:36.254)
They sure do! Yeah!

THE HUMAN PROTOCOL (28:44.042)
And this conversation or what you're saying is so valid tied because what you're doing is allowing people to think it is okay not to know everything and we can figure things out. Not everybody is an expert and that is totally fine. Right? It is totally fine but you have to get started, learn it and it should feel safe saying, don't know, should feel okay.

Ty Givens (28:52.477)
Mm-hmm.

Ty Givens (29:07.646)
I actually trust the person that says that I don't know way more than I trust the person who says they know it all. Right? Because if I don't know and I'm willing to figure it out, that's a lot more valuable to me than the person who says, oh, I got it. Yeah. Yeah. I don't know the answer to that, but I can find someone who does or I'm going to look into that or I've never had that question posed to me before. And like, especially being in consulting, because sometimes people think you're supposed to know everything and it's like,

THE HUMAN PROTOCOL (29:17.688)
Yeah, that's very human. It's a very human answer there.

THE HUMAN PROTOCOL (29:35.63)
I'm sure people think when they call you, you know everything too.

Ty Givens (29:42.618)
Right! My favorite thing is when we're scoping a project and they're like, so how long will this take? Like, I haven't seen how you're set up like this.

THE HUMAN PROTOCOL (29:52.664)
don't know the requirement set up and they already went, okay, tell me please how long it's gonna take.

Ty Givens (29:56.668)
Tell me how long this will take. I'm not sure. Which is funny because I had some work done here on my house and

the my question was, well, how much longer is this going to be? And the guy said, well, I'm not quite sure because and I laughed and I was like, I actually totally get it. Now that you've explained that, I get why you would say that because you don't know what you don't know yet. And so, like I said, I trusted that answer more than it's going to take me three days for something you haven't even investigated yet. You know, be be very aware of the expert who knows it all or the person who's really good at just presenting a

THE HUMAN PROTOCOL (30:30.606)
Yes.

Ty Givens (30:33.056)
very polished appearance. Ask questions about the how, the what, the why, the when, the where, like get asked questions because what you'll find once you lift all of those layers is that they may not actually be as polished, but they might have read a lot of things and they know a lot of theory or they're leaning on Chad GPT to write their posts and Chad GPT makes them sound good, but ask them questions. And so that's like one of my

THE HUMAN PROTOCOL (30:58.072)
Yeah.

Ty Givens (31:02.976)
favorite things like I'll put posts out and let me let me be clear. I use Chudgy BT. I use it to help me sound better because I'm not a marketer. And so thank you.

THE HUMAN PROTOCOL (31:10.764)
Yeah, it's the best thought partner ever. It is the best when you use it as your assistant. Absolutely. Yeah.

Ty Givens (31:16.956)
That's it. Yeah, I'm like, this is what I'm thinking. But like help me, you know, sound a little bit more polished because I'm a straight shooter. Like I'm going to say something that's not going to be as polished, but it's real. So I'll use it for that. Now, my favorite is when I get a message from someone who asks me questions about a post because I feel like that's an opportunity for me to show them. I know what I'm talking about. I'm not making this up. Yeah. Yeah.

THE HUMAN PROTOCOL (31:25.005)
Mm-hmm.

THE HUMAN PROTOCOL (31:42.574)
Yeah, the substance, the substance comes there.

Ty Givens (31:45.99)
Yeah, ask me questions, you know, or I'll say, but I know you've helped me sometimes out put something and then the link won't work. You're like, hey, the link, thank you for that. Like I know that I, I mess up.

only on the days that end with Y. Every other day I'm perfect, right? So I'm going to make mistakes, you know? But I love the opportunity to like really express why I have a viewpoint that I have or to even tell a story around a time that I like how my viewpoint came to be because I probably went through something that shaped it.

THE HUMAN PROTOCOL (32:18.848)
I love that. It's totally needed. And you only can really find if there is substance when you start asking questions, right? When you try to go beyond what you're seeing and what people are telling you and dive in a little bit deeper. And that's when the true colors, you know, comes. And then you'll figure it out if that's if there's really substance there or not. But totally agree with you.

Ty Givens (32:33.854)
this.

Ty Givens (32:41.171)
Mm-hmm.

THE HUMAN PROTOCOL (32:47.168)
Ask questions, try to understand the what, the how, and the why, always. Don't take things just from what you're seeing or where people are attending you, because at the end of the day, it is our responsibility to make sure that we pick the best tool, the best vendor, or the best system that we want to implement. So totally need it.

Ty Givens (32:55.731)
Yeah.

Ty Givens (33:02.846)
Yeah. Yeah. Yeah.

THE HUMAN PROTOCOL (33:07.502)
Ty, when we first chat and say, OK, what are we going to talk about? And we were talking a few things. You say, we should talk about how AI is exposing our organizations, operations, leaders, you name it, right? What you know and what you don't know. And part of what we talk is really if AI breaks your support team, it's because it probably was already broken.

Ty Givens (33:12.862)
Yeah.

Ty Givens (33:28.894)
Mm-hmm.

Ty Givens (33:36.647)
Yeah.

THE HUMAN PROTOCOL (33:37.656)
People are going to be blaming everything on AI, the good and the bad. But then this conversation is just too, OK, what about if it's not AI? And really what AI is doing is exposing what's not working well already on your organization and your team. So curious to hear from you on this angle. How do you see AI?

exposing really the things that are not working well and what leaders need to think before they implement AI to make sure that we don't expose what's not working.

Ty Givens (34:20.53)
Yeah, okay, here's a very simple example. Planet of ours, we have...

been with them for many years now, but we've worked on their help center. know, once rolling out the AI bot became a priority, the idea was to rewrite the articles in a way that answered, that was set up for AI to answer, or to use for answers, I should say. We did that. Anytime we do any rewrites, we always send them to the client. The client has to review, they have to approve. And then once we get that approval, then that's what we'll load it into, whatever help desk they're using.

THE HUMAN PROTOCOL (34:44.182)
Mm-hmm.

Ty Givens (34:57.992)
Well, we had an instance where a customer was engaged with the bot and the customer asked for a product that didn't exist. And the bot gave the customer an answer based on this product that didn't exist, but leveraged an article where in that article there was language that supported what the bot said.

Right? So we're like, when we share, we're like, well, use this article to answer the question. But it says right here that you have this thing. And the client was like, well, we don't have that. But remember, anything that gets published, we send to the client for review. So it was a good example of if you don't take the time to review the information that's coming out, you could end up providing incorrect information to your customer.

THE HUMAN PROTOCOL (35:55.096)
Yeah, to the customer.

Ty Givens (35:57.51)
Yeah, because it is, it will, can hallucinate, will.

try to figure out answers. I look at it as the most eager employee who's always gonna be like, yes, okay, I got that for you. And it's like, that's not what I asked you for. And there are times when I'll be using it to like, I mentioned I'm not a marketer and I need help with marketing a lot. So I'm like, this doesn't make sense. And then it'll go, yeah, you're right. Right, because I know what I want at the end of the day. So the idea of exposing,

your mishaps is highly, highly likely. And this goes for AI agents. It goes for co-pilots. If your processes aren't organized, it's a really good test of how well you've documented different information, how well you provided instructions to your team, because this tool is only going to do

what is in the documentation for the most part because it will make mistakes but for most part it's going to be looking for that answer and it's going to give it back to you and if it gives you something back that's not right and you need to check the source of your information and if you check that information and you're like well I wrote it this way but I intended for it to be this way understand that your agents are filling the same thing.

THE HUMAN PROTOCOL (37:18.808)
Well, yeah.

Ty Givens (37:21.854)
You know, and so you have to kind of be rigid. Like back in my two days, I wrote this performance improvement plan that was so airtight that HR was like, this person, there's no way that this person is going to survive this. I wasn't trying to not have a person survive. What I was trying to do was get the exact results that I needed. And I think you better write a really tight performance improvement plan for AI.

THE HUMAN PROTOCOL (37:45.902)
Yeah, I agree. Documentation is key and things that we all need to be careful about because once you put a shot button in front of the customer or an AI solution in front of the customer, yes, mistakes will happen, but you don't want to be repetitive in a way that mistakes are being scaled, scaled, scaled because what's going to happen, you're going to break trust with that customer if you really don't pay attention to it.

Ty Givens (38:16.028)
so true.

THE HUMAN PROTOCOL (38:16.056)
So this is not a thing that it will take one day to implement. You really need to put in the time to make sure your documentation processes, how you want things, escalation path, And where the issues are going to be routed if the AI bot or your first level of AI doesn't solve it. And think it in a way that time that this is not a one-time implementation and you are done.

Ty Givens (38:22.365)
Yep.

Ty Givens (38:44.924)
No, it's not.

THE HUMAN PROTOCOL (38:45.214)
It's not. This is what the rest of your life will continue to be because you will need to continue to train and and curious to hear from you on that part of. Are people really thinking what comes next after you implement this?

Ty Givens (38:49.167)
you

It's true.

Ty Givens (39:01.606)
No. No.

No, because it is similar. Like I said, it's kind of like when you get the new tool that's supposed to predict your workload or whatever the case may be, you think I'm going to plug this in one time, I'm going to run it and I'm going to go on about my business. And what you have to do is once you get the output, you need to still be reviewing it. So on a weekly basis for any of our clients that are active, we literally go over what people are searching for, which tickets or sorry, which articles create

created tickets, which what did they search for and was that accurate information? It's it's a and then we're constantly rewriting the FAQs. We're rewriting it to make more sense because maybe we did that wrong or maybe we said this wrong. So it is it's literally a constant thing. Now it's a heavy lift up front and then there's maintenance. And so you need to set some sort of cadence to touch it to make sure that it's doing what you expect it to do the way that you expect it

THE HUMAN PROTOCOL (39:49.71)
Bye.

Ty Givens (40:04.446)
to do it and if it's not then that's your time to course correct. But yeah there's like I said I would be really rich if I could figure out how to like set it and forget it and that yeah I would be good but I don't know how to do that.

THE HUMAN PROTOCOL (40:18.541)
The ownership still ours. The same ownership that is to implement it, the same ownership that you will need to continue to own it after you implement it. And it is hard work, because processes changes, new system, new processes, you need to continue to keep updating your knowledge base for this to be as successful as you want it to be. And as you say,

Ty Givens (40:21.618)
Yeah, it is.

Ty Givens (40:30.782)
Absolutely.

Ty Givens (40:41.363)
Yep.

Ty Givens (40:45.212)
Yeah, 100%.

THE HUMAN PROTOCOL (40:48.504)
keep an eye on rewriting FAQs and checking on processes. It's a lot of work. It is a lot of work. You need to have your team structure to absorb that after you go live with the solution, because someone will need to do it.

Ty Givens (40:53.95)
Mm-hmm.

Ty Givens (41:05.554)
Yep, it becomes another job. Like sometimes we'll go in. You know, like I said, I come from a big company. when my days at Intuit, we had a quality function. We had a training function. had tech. At that time, we were using a Bias switch. So we had dedicated telecom engineers that worked with us. We had an admin for our EWF. So you have all these different roles.

now those roles are not as prevalent as the on-premise servers are not a thing anymore. Everything is in the cloud. And now you don't have to know how to program anything to launch it. You can just kind of roll it out. So it's just a different world. But when we get into companies, some of the companies that we know and love, really great brands. And we get invited in to work. And I'm like, they don't have any of this stuff set up. So you're not alone if you don't. They don't know. And what ends up happening is

is that they think that the support manager, the customer service manager is supposed to handle all of those roles.

THE HUMAN PROTOCOL (42:14.275)
Yeah.

Ty Givens (42:14.778)
And so a lot of what we do is go, well, like, let's break down what the support leaders role actually is and give them success criteria because we want them to feel good. Especially if you're experiencing a lot of turnover. We have a client who I think they were going into their fourth support leader. And I was like, you guys got to like slow down and build a foundation because a lot of people can come in and run support, but not everybody can build the foundation. That's the key and critical piece. So let's build.

foundation so that somebody can come in and run it successfully.

But in doing that, at the end of it, what we do is we look at like, what are their norms going to be? Do they want a quality program? Some clients want to have a person on site or, you know, in their office that does it, that's fine. Do they want to roll with AI? Whatever those things are, we then have to build a role around that, whether it's a coordinator role that is going to look across multiple facets, because maybe AI is doing a lot of the heavy lifting and they're just doing the spot checking.

THE HUMAN PROTOCOL (43:14.829)
Right.

Ty Givens (43:19.68)
Or they might want a dedicated resource that we have to train up on best practices in these different areas But the reality of it and what I'm trying to get to is that it's not a one-person job There's not one person who can do all of those things that I quality Checking a and I'm going old days on you not today's AI world But quality checking a five-minute phone call can take 30 minutes to an hour because you're gonna have to listen

THE HUMAN PROTOCOL (43:33.249)
Mm-hmm.

THE HUMAN PROTOCOL (43:47.838)
If you want to it right. Yes, absolutely.

Ty Givens (43:50.354)
Yeah, yeah. And that doesn't even include the prep time for the meeting when you have to talk to them, the actual coaching conversation, the following up that happens after that. All of this is happening for one five minute call.

Right? and those are the parts, that's the invisible labor that the rest of the company doesn't see or that even the C level or depending on who you report to, maybe your VP doesn't understand what goes into it because we should be quality checking more.

Sir, ma'am, it literally takes me, you know, just to get through a five-minute call, it takes me 30 minutes. Then I have to meet with them. Let's add on another, you know, 45 minutes. So one five-minute call is 75 minutes of work. Well, can you shorten it? Do you want a good experience? know, which one? So there's a lot that goes into it. My workforce management background gives me the ability to quantify that time to help people justify why we need to hire for roles or go for AI.

THE HUMAN PROTOCOL (44:51.062)
It's different skills, right? And it's hard sometimes to find someone that can handle everything. But as you were saying, these will require different skills, depending on what you want to achieve. And for that, you need to start thinking, once you go live with it, what are the skills that you're going to need to maintain that operation? Because if you're really investing all this time upfront to deploy and implement AI, you're really...

One, make sure that you maintain it the best you can so you can continue to improve it afterwards. It's also a changing mindset on how our organizations, especially in support and CX organizations, are going to change because roles may be changing as well as we are just. From all the efficiencies that AI should be providing, roles should probably be evolving.

So it's going to be interesting to see where this goes and how things evolve time. And one thing for sure, with all this AI conversation, really what I'm seeing lately is that the question is changing from who answered to who is responsible. And that's where my mind, at least my mind is lately, is like,

Ty Givens (45:51.742)
Yeah.

Ty Givens (46:10.138)
yeah.

THE HUMAN PROTOCOL (46:13.344)
We know it's AI, it's embedding into our operations already, chat bots, different other ways to handle level one and sometimes even level two issues. But who is responsible is the thing that we all need to be thinking ahead and start thinking when something happens, who is accountable for that outcome. And to really build an AI

Ty Givens (46:38.258)
That's it.

THE HUMAN PROTOCOL (46:43.48)
first organization, whatever, thinking about the outcome is really how you're gonna then define the rest of the AI solution in your operations.

Ty Givens (46:56.766)
So true. So true.

THE HUMAN PROTOCOL (46:59.458)
Yeah, so outcome first, then you define the rest, right? But when a bot gives the wrong answer, who's responsible? AI or the...

Ty Givens (47:11.23)
That's a good question.

You know what? It actually, I'm gonna follow Mullen's word and say it's gonna come down to the person who programmed it. Because it's going to do what you ask it to do. so there are times when it hallucinates, there are exceptions to every rule, but I would say look at the source of the documentation that you provided and look at the instructions that you provided. And if you didn't say don't hallucinate, stick to the article.

THE HUMAN PROTOCOL (47:22.199)
So.

Ty Givens (47:45.012)
And you should have.

THE HUMAN PROTOCOL (47:48.192)
If you don't want to get blamed, put in the work, right, Ty? It's still on us. It's really us. It's our responsibility. We own the outcome. So put in the work. Yeah, I agree. Let me see what I got next here for you, Ty. Curious to pick up your brain on this.

Ty Givens (47:50.674)
That's right.

Ty Givens (48:00.104)
We are.

Ty Givens (48:09.585)
Okay.

THE HUMAN PROTOCOL (48:16.366)
Do you have anything from what you have seen and what you know so far on your experience with AI and support overall? Do you have anything that you are reluctant to see AI doing it in support, for example? So if there is something that you think AI should not do, what would that be?

Ty Givens (48:38.372)
I would say.

complex issues that live in mostly a gray area, meaning that it doesn't fit the typical if this, then that type of criteria. And or it requires a certain level of empathy, right? Because when all is said and done,

that.

If you want to piss someone off, send them through an endless loop of the same thing without showing any true understanding for the experience that they're having. I think that if we are looking at, you know, it could be high stakes, let's say it's holiday season, if someone orders a custom piece of jewelry for, you know, their college student who came home and that jewelry is not going to make it on time.

then I would say, know, when you're developing your criteria of what the bot handles, custom, misdeliveries, you know, like just totally bypass it, get it to a person. There are certain situations that require a level of humanity that, you know, it is just necessary.

THE HUMAN PROTOCOL (50:03.532)
Mm-hmm.

Ty Givens (50:08.858)
And so I think in those cases, and if you want to show the customer that you genuinely care, because yeah, you can have a bot kind of like do some of that and talk to the person, but there's going to come a breakdown where it's like, no, you're not understanding what I'm telling you. My child came home from whatever universe, we'll go with Tuskegee, my niece goes to Tuskegee. So, you know, my child comes home from Tuskegee, I bought this piece of jewelry for her, and it is not right, or it's not going to be here, and she's going to be going

before it gets back. Don't send that through an endless loop. That requires a certain level of pause. Let me sit down with you. Let me understand the issues. Let me see what we can do. It's a little bit different than here's 20 % off your next order because there probably won't be a next order.

THE HUMAN PROTOCOL (50:50.936)
Hmm.

THE HUMAN PROTOCOL (50:55.458)
And we all have been through those endless loop calling 800 numbers. And sometimes when you know, and we all hate that it's a, let me transfer here. let me transfer there. That's why when people say, I don't want to put AI to talk to a customer. And, but if AI is going to give you a faster response and take care of the customer need faster than what the ticket may take you and a human to jump on it. Why not? Right.

Ty Givens (51:02.173)
Yeah.

Ty Givens (51:24.274)
Yeah. Yeah.

THE HUMAN PROTOCOL (51:25.44)
If it's going to alleviate the endless loop because your contact center is maxed out, why not?

Ty Givens (51:32.872)
Yeah. And also give the agent an opportunity to think critically and to like solve something and to really feel good about what they're doing instead of just always doing the same thing. It's like, that's kind of boring. It's like, I want to actually feel good about solving a problem that was complex and that makes people feel valuable and useful. And so let them do that. Yeah. Yeah. True.

THE HUMAN PROTOCOL (51:40.785)
Absolutely!

THE HUMAN PROTOCOL (51:54.242)
what we should be doing as humans. No one wants to keep repeating the same over and over again. I even look back when I got started as well, me fixing all these issues, every single ticket the same as the previous one. And when I look at that tie, I feel a little bit sad why we were doing that. You know, it's just repeating and where was the pride and the passion of you doing something valuable?

Ty Givens (52:15.004)
Yeah, yeah.

Ty Givens (52:23.421)
Yeah.

THE HUMAN PROTOCOL (52:23.834)
And so I see all these really, all this revolution with AI, I see a lot of benefit myself. yeah. Any metrics that gets worse after AI is being rolled out that you have seen?

Ty Givens (52:31.752)
Definitely.

Ty Givens (52:40.988)
Yeah, sometimes CSAT originally doesn't do so well. And I think it's just because the, depending on what happens, depending on how you're rolling it out, if you're not testing things properly, if the wrong answer comes out, because let's have a conversation real quick about CSAT. CSAT sometimes goes down.

when the customer doesn't like the answer provided, even if the answer provided is in line with your policy. So when you're dealing with

THE HUMAN PROTOCOL (53:15.512)
But you're talking about the customer knowing that it was AI or a shot bot that was providing the... it doesn't matter, right?

Ty Givens (53:24.19)
No, no, I'm customer wants to return an item on day 45 and you have a 30 day return policy. And they go through the bot and they get stalled and the bot says, sorry, you can't do that. It says it nicely. Did I help you today? No, you did not, right? I'm pissed off at you. You didn't give me what I want.

THE HUMAN PROTOCOL (53:33.408)
Okay, yeah.

Ty Givens (53:46.462)
And so when I say that it can go down, it has a lot to do with really the true understanding that sometimes customers are not satisfied, not because they didn't get kindness or respect or they weren't treated well. They didn't like the outcome of their request. And so you can see that a lot with CSAT, I mean, excuse me, when you deploy AI that your CSAT goes down

THE HUMAN PROTOCOL (54:06.946)
Yeah. but yeah.

Ty Givens (54:16.416)
because it's just the policy, right? Does that mean that you change your policy? Not necessarily. Policies are there for a reason, but you just operate with a different understanding. And I think that that's a learning moment that you can use for your agents who will feel a certain way if a customer says, I was dissatisfied with my conversation with Ty as an example. But when we look into it, you find that Ty only was upholding the company policy. Ty's tone was right.

THE HUMAN PROTOCOL (54:24.174)
Mm-hmm.

Ty Givens (54:46.336)
delivery was right, but she said no to something that she's supposed to say no to. So that's why the customer is not satisfied. And I think that if we can teach our frontline how AI works.

and how to really align themselves with it too. Let that be a coachable and teachable moment on like, hey, you can help us create accurate outcomes. So the only things that you're getting are the things that require a lot of thinking, it will help. In that instance, they may also feel less tied to the fact that they got a negative CSAT rating because they upheld policy. Because sometimes we can't take ourselves out of situation. It happened to me, it's horrible.

THE HUMAN PROTOCOL (55:27.522)
Yeah.

Ty Givens (55:30.656)
I don't want to overlook that. I don't want that negative CSAT. But if we could apply it to an agent and they can see it from a leadership perspective, they might be like, I kind of understand that. I kind of get that. So I think, you know, there's, what's his name? Why am I drawing him? Like I can picture his face. He's my buddy. He's at, he's up front. why am I drawing? Anyway.

THE HUMAN PROTOCOL (55:55.48)
Kenji? Yes.

Ty Givens (55:56.712)
Kinji, yes, Kinji. He has his own company and they have CSET for AI. So this is something that he's tackling. there are solutions out there and ways out there to get around it. But I think that's the first thing that we'll see. And it tells you a lot about the policy.

THE HUMAN PROTOCOL (56:17.72)
Yeah, absolutely. It's going to be interesting to see how all this evolves and what are we getting the different flavors because different industries, customers, they have different needs, So it will be interesting sometime in the future to look at reports and see how things are behaving. For example, does customer care or not?

about receiving a human versus an AI support. What do you think?

Ty Givens (56:51.324)
I don't think so. think that customers care that they get correct information in a timely manner. Don't make me wait two days to tell me something that a bot could have told me instantly, because that's going to piss me off. But if the answer is no and I get the answer upfront, just like, I might still have to talk to an agent. But if they just reiterate, that's fine. But don't make me wait two days to hear something that is very clearly a yes or no. Help centers are lovely because you can do your keyword searches and find the information.

it so much easier because you can ask exactly what you want and get your answer right away.

THE HUMAN PROTOCOL (57:26.132)
I agree. think customers don't care. If you implement it right and if your chatbot or your AI support is keeping the same level of quality that your customers expect, I think they wouldn't care about it.

Ty Givens (57:29.075)
Yeah.

Ty Givens (57:39.016)
Mm-hmm.

Ty Givens (57:44.638)
And transparent. That's the other thing I think people miss is if you're, I mean, you have to say like, this is, you know, I'm an AI assistant or whatever. But if you're transparent about it they get a correct answer, they're fine with it. I think it's when you make it very difficult for them to talk to someone and you try to pretend that they're talking to Molly and that's not Molly and you know it's not Molly. So, yeah.

THE HUMAN PROTOCOL (58:03.768)
Yeah, yeah. Don't ever do that. Make sure that the system is designed in a way that if the customer needs an agent, you send it to the agent, right? And that's the way it should be. Awesome, Tye. I have one last question for you, okay? So a piece of advice you would give yourself, but that yourself

Ty Givens (58:16.616)
Yep.

Ty Givens (58:24.647)
Okay.

THE HUMAN PROTOCOL (58:32.544)
reflected today is a younger version of Ty. So what would you tell that girl seeing her today, especially facing this new feature with AI?

Ty Givens (58:45.886)
I'm so glad that you're curious because even though the curiosity today, the feedback that you're getting is that you're not strategic enough for you're not a visionary. But that curiosity is going to pay off for you tenfold. So don't change that about yourself.

Keep going, keep pulling back the layers, keep questioning things because that is what's gonna make you very successful in years to come. Yeah.

THE HUMAN PROTOCOL (59:25.452)
Love it, love it, love it. I think the future will be very good with those that allow themselves to be curious and to try new things and to adopt and adjust. Beautiful. So Ty, thank you for telling us the truth about what AI exposes. It's really been a pleasure to having you with us today. Thank you, my friend.

Ty Givens (59:52.05)
Thank you, it's so good to be here.

THE HUMAN PROTOCOL (59:54.35)
Thank you. And if this episode challenged you in a way, send it to a leader that is deploying AI to support right now. Follow the human protocol because the code may change, but our humanity will continue to be the constant. Thank you all and see you next week.