Most CEOs have already concluded that AI belongs somewhere in their business. Few have translated that conclusion into a decision. The gap between agreement and action is where the real risk sits — not in the technology, but in the absence of a policy and a plan for getting a team to actually use it.
Bryant Oberg, founder of Human Co-Pilot, spends his time inside that gap. He advises CEOs and teams on AI adoption and governance, and the pattern repeats across engagements often enough that he's named it: leadership agrees AI matters, then stalls on the same question. "The problem that everyone's coming to is well, how?"
For a solopreneur, "how" is a personal decision. The founder is the only user and the only one who bears the consequences of a bad prompt or a leaked file. Once a company adds team layers, that decision stops being personal and becomes organizational — and most CEOs haven't noticed the shift.
In the absence of a stated policy, employees don't wait for one. They adopt AI on their own terms, using personal tools to handle company work because no one has told them otherwise. Bryant calls this "shadow AI" — employees moving proprietary and client information into personal chatbots or agents, without the company's knowledge, simply to keep up. It isn't rogue behavior. It's the predictable result of a governance vacuum. "Not having a policy is not adequate anymore," Bryant says, "because people are using it regardless."
That reframes the CEO's actual decision. It isn't whether to allow AI use — that's already happening. It's whether to bring it into the open, under terms the company sets, or let it continue unmanaged.
Bryant's operating principle for advising CEOs through this is direct: "Letting the human run the AI, not letting the AI run the human." The distinction matters because AI is non-deterministic — it doesn't return a fixed answer the way a spreadsheet formula does. It generates a response shaped by the input it's given, so its usefulness depends entirely on the judgment of the person directing it. Teams that don't understand this either disengage entirely or hand the tool more authority than it has earned.
To explain why teams freeze rather than engage, Bryant returns to a historical analogy that anchors his advisory approach. When automatic elevators replaced manual ones in the 1920s and 1930s, passengers were reluctant to step into a machine with no visible operator. Elevator companies didn't solve this with a technical fix. They stationed an attendant in the elevator — not because he was needed to press the button, but because "you needed someone to give you the comfort in order to get in the elevator."
Bryant applies the same logic to AI adoption at the team level: "CEOs, to an extent, have to use that paradigm with their employees. They have to get people comfortable to get in the elevator." Capability was never the barrier. Trust was.
The framework points to two concrete moves, and Bryant is explicit that neither is a heavy lift.
The first is a written policy — what tools are approved, what data can and can't be shared, whether settings like model-training toggles are on or off. He frames it as a starting structure, not a finished system: "These are not super complex things that take a ton of time to roll out for companies... just a three-page policy." The goal isn't to resolve every edge case. It's to give a team enough of a landscape to operate inside, rather than leaving each employee to draw their own boundaries.
The second is deliberate reassurance — the elevator-man role itself. Bryant treats this as something a CEO needs to actively perform, or sponsor someone else to perform, before expecting adoption. Skipping it doesn't speed things up; it produces a team that's either avoiding the tool or using it unsupervised.
The sequencing is the tradeoff most CEOs miss. It's tempting to treat policy and comfort-building as optional preludes to the "real" work of implementation. Bryant's position is that they're the prerequisite for it: "You have to take people and educate them first... then you can really worry about building the house."
The most common objection Bryant hears from CEOs isn't skepticism about AI — it's time. "I'm too busy" is the recurring line, and he treats it as a diagnostic rather than an excuse: "If you're too busy, that means you need someone else to help you. You're not going to get to go do it yourself."
That reframing carries a real cost calculation. Every week without a policy is a week of shadow AI accumulating in the background — proprietary information moving through personal tools the company can't see. Every week without deliberate trust-building is a week a capable team stays hesitant, amazed only in theory by what the tool could do for them. Neither cost shows up on a dashboard. Both compound.
Bryant's framework doesn't hand CEOs a finished system — it hands them a sequence: name the policy, build the comfort, then let capability follow. The decision that separates companies isn't whether to use AI. It's whether the CEO treats governance and trust-building as the first move or as something to get to later, once the tools are already in place. By the time the gap becomes visible — in a data exposure or a stalled team — the sequencing decision has usually already been made by default.
Bryant Oberg is the founder of Human Co-Pilot, where he advises CEOs and teams on AI adoption and governance. Before pivoting fully into AI following ChatGPT's release, he spent 15 years in finance, including as a partner at a hedge fund, before building and losing his own AI-powered compliance business within months. Connect with Bryant on LinkedIn or at human-co-pilot.com.
__________
Jeff Holman is a CEO advisor, legal strategist, and founder of Intellectual Strategies. With years of experience guiding leaders through complex business and legal challenges, Jeff equips CEOs to scale with confidence by blending legal expertise with strategic foresight. Connect with him on LinkedIn.
Intellectual Strategies provides innovative legal solutions for CEOs and founders through its fractional legal team model. By offering proactive, integrated legal support at predictable costs, the firm helps leaders protect their businesses, manage risk, and focus on growth with confidence.
__________
The Breakout CEO podcast brings you inside the pivotal moments of scaling leaders. Each week, host Jeff Holman spotlights breakout stories of scaling CEOs—showing how resilience, insight, and strategy create pivotal inflection points and lasting growth.
Listen and subscribe on your favorite podcast platform:
__________
Want to be a guest—or know a scaling CEO with a breakout story to share? Apply directly at go.intellectualstrategies.com.
TRANSCRIPT SUMMARY:
00:00 Humans Must Lead AI
01:02 From Finance to AI Entrepreneurship
02:46 Escaping AI Productivity Pressure
07:14 A Failed Business Sparks Reinvention
12:22 Understanding Nondeterministic AI Responses
14:27 Allow AI to Admit Uncertainty
17:18 Keeping Humans in Control
20:53 Shadow AI Creates Business Risk
23:20 The Elevator Operator Analogy
24:45 Start With Policy and Education
28:01 The Fractional Chief AI Officer
31:36 Too Busy to Implement AI
FULL TRANSCRIPT:
Bryant Oberg (00:00)
you needed someone to give you the comfort in order to get in the elevator. They put the elevator man to stand there. They have to get people comfortable to get in the elevator. If you're too busy, that means you need someone else to help you.
Jeff Holman (00:12)
Welcome back to the Breakout CEO podcast. I'm your host, Jeff Holman, with Intellectual Strategies. I'm here today with Bryant Oberg. Bryant, great to have you on the show.
Bryant Oberg (00:21)
Jeff, thanks for having me. It's great to be
Jeff Holman (00:22)
Yeah.
I'm I'm glad to have you. You you started and you run Human Copilot. So I'm really curious to to talk about that today and and what you do, you know, who work with, what you build, and how you help scaling CEOs in their businesses. but but when we're talking before the show, just for momentarily, you mentioned something that I think is is worthwhile for people in our audience to know, and that is that you're not necessarily coming at this, you know, the the building the AI tools and such.
from a purely technical standpoint, right? You're you're coming at this from more of a business, maybe even a I'm gonna guess a human a human standpoint, hints the name of the business.
Bryant Oberg (01:02)
Yeah, totally. so I spent the first fifteen years of my career in finance. I started in banking. From there I moved to the buy side. I was worked at a hedge fund, then became a partner at a hedge fund. And I was on that trajectory to kind of stay in it. I was then I moved on and did some consulting. And boom, November twenty twenty four, Chat GPT hit me in the face and I haven't really looked back. I've looked back a few times to to see the past. But
ever since then I've I I saw already then and even though it wasn't so long ago in AI terms, it was about ten years ago. Yeah. And I saw then the trajectory of where AI can take businesses. In fact, it helped me build a business in six weeks. And so I've started to lend those services of what I've been able to build for myself to others, and that's really where human copilot originated from.
Jeff Holman (01:54)
I love it. I love it. Well, the the whole concept of of kind of coming at things from a different angle is is super intriguing to me. I feel like that's one of the maybe the superpowers that I have when I come into things. Just you know, I I draw on an engineering or a legal or a business background d and and kind of transl I I think of myself as a translator a lot of times between different different people. I I'm excited to hear what you've done to build your business and and also
how because a year and a half, that's I guess maybe close coming up on two years, November 2024 was not that long ago in terms of time, but with the acceleration of what people can do, what people have been doing through AI, that is almost an eternity relative to, you know, 10 years ago and what people are doing. So I I'm excited to hear about that.
Bryant Oberg (02:46)
Yeah, I think it's actually it's it's a very interesting thing for for people like us and and entrepreneurs and business owners where we feel like now, because the whole world's our oyster that we should be able to do so much, we have to set one of the things I talk to my clients about is setting realistic goals. Right. You don't have to take over the whole world tomorrow. Is, you know, what are we trying to what pain point are we trying to solve for? And let's just solve that pain point. Let's not solve every single problem right now.
and that goes into ancient construction and I can talk the whole AI craze, a as we dive into it. But I just think it's an interesting thing. I mean, I know that personally I'm not satisfied over the last year and a half because well, if I did ten ten things, I think I should have done a hundred things. And if I did a hundred things, you know, AI could have maybe do a thousand things. Right.
Jeff Holman (03:33)
that's a that's a fair that's a fair statement. And I I'll bet a lot of I bet a lot of our listeners are are saying themselves, you know what, there is an AI overwhelm aspect to this, right? I don't know, I don't know if there's a a a cultural term for it yet, but you you feel like you should be doing so much more, right?
Bryant Oberg (03:48)
So there is, people don't really like the term. It's actually called AI psychosis.
Jeff Holman (03:53)
okay. I haven't heard that before.
Bryant Oberg (03:54)
Sounds a little too intense. I I don't think that's where I mean there there's a medical term for it, but you know, where people actually go to the hospital, it's very sad. I shouldn't I shouldn't even smile at it. But it it it but I'm talking more in the fact of people kind of being stuck in their minds of I should be performing 10X where I'm at. And it's yeah you have to bridge that gap. And that's really where the human co-pilot comes into into the aspect of, you know.
Letting the human run the AI, not letting the AI run the human.
Jeff Holman (04:22)
Mm, yeah. And I think I think on that note of the call it AI psychosis or whatever whatever term it is, that's not it it's a it's a different version of what's been out there before, right? Like people how many how many years have we dealt with people saying, I really should be raising money, I need V C capital, I need it I need P E money and and in in reality, how what percentage of the businesses really have that? A a a small, small percentage. They're the they're the the the sexy ones, the ones that you see, but
But so many other businesses are are being built. And I think that the the I don't know, the the culture is coming around to the this whole bootstrap mentality. Like there's huge benefits to bootstrapping, and especially in d if you can do it with AI. But but VC is not the only way. And and yet five years ago, especially before 2021, I think everyone would have said, You gotta have VC. You just like you're you're not gonna hit a home. Nothing's a home run without without VC. So and and I think we're seeing that turn. So I think we'll s
Yeah, we'll see where the balance ends up with with AI itself, but but certainly for those of us that are in control of the AI in our lives, or not overwhelmed or developing psychosis by it, it's a huge benefit. I mean, I I just got off a call with a with a client today and we were talking about, you know, they're a smaller client, but the the stuff we can do to help them with their contracts that are
very large contracts for the size and stage of company they are and the type types of companies that they are contracting with, the things we can do with their contracts is just so much more I'm just gonna say thorough, maybe sophisticated, but at a minimum so thorough that we and we can do it so quickly that we couldn't have done fifteen years ago when when I had clients who had, you know, hundreds of contracts and they're just like, you know what, Jeff, we just don't have time to we don't have time to manage it all.
We ha every contract has deadlines and renewals and and we we try, but we do what we can, but we can't do it all. Now it's kinda well, probably can do it all. even, you know, even even on a on a somewhat scaled basis, you you probably can do most of it, if not all of it, because AI really is at the point where it's it's giving it's it's enabling things, information. I'm gonna call it
information management that you just didn't have before.
Bryant Oberg (06:52)
Yeah. That's a great way to put it. That's a great the information management is a great way to put it.
Jeff Holman (06:57)
Well so tell us about human copilot and and where that started w what kind of drove you to to make that switch and what what happened in what happened in October of twenty twenty four that that made you say, November, hey, it's time.
Bryant Oberg (07:14)
Look, for all the breakout CEOs out there, yeah, most of you probably have a time where you failed. So I'll that that's where I start actually, which is it it was in October of twenty four, I was building a business for something called BOI compliance. BOI compliance is people who run small businesses might have heard of it, they're account about it, but you had twenty five million small businesses in America that had to file this paperwork and you're gonna be charged a five hundred dollar fine per day.
Starting january first, twenty twenty five.
Jeff Holman (07:46)
And for those people who don't remember, that was beneficial ownership information. You had to say who who was actually the owner of the company, right?
Bryant Oberg (07:53)
Thanks, Jeff. Yeah, exactly. so people were up in arms about it. It wasn't fair, this and that, but it was coming, it was coming to you and you had to do it. You know, so I did a little research and I thought I was doing consulting at the time and business consulting and restructuring, finance consulting. Thought, you know, most people don't know about this. If I offer it as a side service, I could really, you know, help people and probably also be a great lead generator. So I built a program and or I started to build a program, a way to do it. Never built the software before, never knew anything about.
computers really. I could, turn on turn it on and go to Gmail, but right, like I wasn't I wasn't the computer guy. Yeah. And I within six weeks I built a basically a whole company that that helps you file your B BOI compliance. Yeah. And, you know, it went from a lead generator to we it's twenty five million people, but only about two point five million people have filed. There's a huge on tab market here where people when it comes to compliance, people don't just
They they want to hand it to someone else and say, you do it. And so I said, look, I I'm gonna start building this into like a full fledged company. And it was because of AI and Chat GPT that I was able to do that. And so that's where I started. I say I failed, and for all those, you know, the CEOs out there, it was because on March 2nd of that year, Donald Trump, President Trump, decided to basically change the way the law was read, but in essence killed the BOI compliance.
And so overnight the business went away. and it was going through this and that. But it was it was an interesting, it was a fun little four months where I thought, it was so exciting, everything was coming, and then boom, crashed, right? But it was one of the best things that happened to me because it led me into AI and then I made a full pivot to be solely AI. And that was how I got started. And it was because during that time I was talking to family and friends, and I said, you gotta work with Chat GBT, you gotta work with AI.
They said, no, it hallucinates. It does things I don't want it to do. It's not so smart. Said, look, you're just not really using it right. Be patient with it. This thing is this thing is growing or is getting smarter, smarter exponentially. Just wait six months. And so I was trying to get people on the train early. And no one would listen to me. And I was like that weird cuckoo AI guy now. And I was like, wait, I'm a finance guy. What you mean I'm the cuckoo AI guy? So I said, okay, how can I get people to sit down with me?
And I found a price point that was enough where my family and friends would sit down with me and pay, but they didn't feel like I was forcing them to be my client. And it was for waiting
Jeff Holman (10:34)
Were you they were hiring you to just introduce into AI?
Bryant Oberg (10:38)
Yeah, teach them how to use Chat GPT. Okay. I said I said, let me sit down with you for an hour. I will show you I will show you how I use AI and I'm gonna teach you how to use it for your life. Not
Jeff Holman (10:48)
Different
definition of friends and family funding, there isn't it? Exactly. Like they are revenue. I I
Bryant Oberg (10:54)
They are revenue, exactly.
Jeff Holman (10:56)
I want to touch on this just for a second because I I think a lot of people are still in this mode. You know, this somebody I talked to this morning was talking about, yeah, I I really need to get into, you know, I need to get into that more. We need to need to dive in. I'm not doing enough of it. and I feel like, you know, people say it hallucinates, which it does, you know, very, very off the wall stuff sometimes.
But but I feel like the the real qualifier here isn't how bad how bad AI is, right? Because it's so powerful and it does so many amazing things. Yes, it does things that are n unexpected or even wrong. But I feel like the I feel like the qualifier should be way more about am I expert enough to use this properly versus why is this so bad when I just blindly rely on it?
Right. If I'm blindly relying on AI, and we see people do that, we do projects or build companies or whatever, and you're just blindly you you don't have the expertise to say, wait a second, that's that's past the guardrail that that I know is there. you're building something different. So so I I'm not sure that it's that the emphasis in the conversation. I I think we say it's an AI problem when maybe it's more of a human expertise problem and we're just trying to build with something that's
more powerful and a little more chaotic than we know how to control.
Bryant Oberg (12:22)
So the thing I kept hearing with AI when I was down with people is I just wanted to do what I wanted to do. Yeah. Which is a funny thing, right? Because and this is something I don't know if a lot of people actually understand about AI, that it's something that's called non-deterministic. It's what it gives its gr its greatest greatest power, but also what creates the most frustration for people. Non-deterministic means that the response it's going to give you is not determined until it
runs through its entire universe of of knowledge and gives you an answer from your prompt. As opposed to computer program where it's just when you put Excel two plus two, you'll always get four. Yeah. Right. And that gives its ability to really, you know, let's say think. Let's let's use that word think. And that ability to think is where it, you know, some of its superpower really lies. But people get very frustrated because they don't understand in order for it to think well about what you want it to
answer you about, you have to give it all the information. You need to be able you you need to know how to actually interact. You need to ha put it in the right data points, garbage in, garbage out, right?
Jeff Holman (13:33)
Well and and with and with I like the Excel example because the the stark contrast is it is it is deterministic in the fact that when you put in a correct formula it it will do what it's what what it's supposed to do and only that. and when you put in an incorrect formula it will give you it like the the default is it it gives you an error code. You you you can't do it because it's an error. So everything that isn't predetermined is just classified as an error. Whereas with AI
I'm not sure anything's classified as an error per se. It it it just says, well, this this is the context you gave me. And if I look at the world of the LLM that I'm dealing with or whatever, right? And I and I probabilistically put together a bunch of words that I think would be a good response to this, something's gonna come back and it and we don't know that it's an error. We don't exactly we just we don't know because there's not an error code default.
Bryant Oberg (14:27)
So one of the best things you can do, and there was a paper written last summer, August twenty five on this about how to stop hallucinations. For those who know hallucinations are when AI makes up an answer because they don't it's not able to to draw on the correct answer, is to give the AI permission to say I don't know.
And all you have to do is put a rule, make a rule, set the memory, go into your settings and find the place to put it and say, if you don't know the answer and and give it a percentage, right? Say if you don't, if you're not 90% confident of your answer, say I don't know. If you're not 50%, if you if you don't think you know the answer, you can tell me I don't know. Right. Because its default is to is to not is to always try to give you an answer. So it's it's a very it's a small little tip and trick that you can use, but it's effective.
Jeff Holman (15:17)
Yeah, and I don't think I've I don't think I've heard somebody say that before. I have heard I have heard people tell you that it's ineffective to tell it what not to do. Because a negative instruction doesn't really get processed the same way, right? It's like, well, how does it know when it's not doing it, right? But if you tell it what to do, then it has a path to go forward. And so that what you're saying makes perfect sense, although I I'm not sure that I've I've thought of it that way.
Bryant Oberg (15:43)
Yeah, yeah. It it's it's really interesting that you know, it it just doesn't have permission to say I don't know because who would talk to a chat bot or any AI instrument that would constantly say I don't know?
Jeff Holman (15:57)
Right. Right. Right. Yeah. Psychologically that's not that's not very appealing, is it?
Bryant Oberg (16:03)
No, it's not.
Jeff Holman (16:04)
And and and frankly it's I think there's some humans that way too that they're not they're not programmed to say I don't know and they will they will just say stuff.
Bryant Oberg (16:12)
So let's go, let's jump to trick number two that I'll share. If you ever get confused about how to interact with AI, just think about how you interact with people. Right? Because that's what it's trained off of. That's, if the the the thing about AI is it's so big we can't even comprehend the amount of parameters or data points it's using. All of those data points are really coming from human-entered information. So if you're confused about how to interact with AI, well, think about how would you interact if it were a person? And
making that switch in your head should help overcome a few hurdles that you come to with AI.
Jeff Holman (16:48)
I I think and that brings us to the kind of the the focus of of your company, Human Copilot, right? Like this is w what's the human focus here? go into that a little bit for me. What what makes what makes your approach different? Is it is is is it what you just explained? You know, how do how do we interact with AI like we would a human, or is there more to it?
Bryant Oberg (17:08)
It's there's more to it. So that that I think is a it's an interesting I call it a trick. It's a it's not a trick. It's a a modality,
Jeff Holman (17:17)
Yeah really.
Bryant Oberg (17:18)
right? Yeah. Guardrail of what you do. You know, I have different depending on who I talk to, I say, you know, do you want to treat it like an employee you trust or employee an employee you don't trust? Right. That's et cetera. But the human aspect to me is really actually the user i and who's the co and making sure you know who the co pilot is.
Right? Is the human the co-pilot or is the AI the co pilot? And the big thing I work with my users on and my my clients is that it's the you have it you have the human and then you have the co pilot. and the co pilot's the AI. And it's how we started the conversation, but it's making sure that the AI is not guiding you, but you're guiding the AI. And this I see all the time because again, what does AI do? It's looking for
answers to fulfill whichever direction you've pointed it in. Right. But what happens is at the end of its answer, it's always kind of guiding you somewhere because to keep the conversation going. And a lot of times people get sucked into it and they forget what their goal is in the first place. And they're no longer able to what I say guide the AI and the AI guides them.
Right. And we can talk about this on a chat bot or we can talk about it in much larger systems where you're dealing with a multi agent system that, is dealing that's handling your sales or your or your inventory or anything like that. But I mean, it it's it sounds strange, but we have to keep people thinking and not just totally giving over their minds to AI, but yet learn how to use AI to its most it its utmost capabilities.
Jeff Holman (18:59)
I I will say that that from a a human perspective, there there are times when I feel like AI is trying to do to me what some social conversations do. I'm I'm a somewhat of an introvert. you know, I'm I'm I'm kind of programmed to listen. That's my engineer coming out of me, or maybe not coming out of me. I'm not sure which way it is. And and so I I'll I'll find sometimes that conversations go on and on and on and
And the other the other side of the conversation, they just they just keep moving the conversation to wherever they want to go. And I I keep listening and I'm like, okay, yeah. And I engage and they're like, you're such a good listener, and I'm like, Yeah, I am, and you are a great talker. So so you know, I feel like that that's what AI would do to me if I if I didn't ever step in though. Because so many times AI is like, do you want me to build this now? And I'm like, No. Like like we're not even close to that. You know? Yeah.
How do how do you when you work with somebody who's in because you've got the people who are building AI, you know, you're sat down with your friends and family and you're like, Hey, let me teach you how to how to use AI in your personal life. But then you've got kind of this a lot of times I think, and it depends on the business, right? In in small businesses, the founder and the CEO are the one and the same, and they're the person building the technology and they're the releasing it. And so maybe they're the direct user.
of AI. But when you start to get into a a company that grows a little bit bigger and you have a CEO leading a team and and maybe some layers in there even, how should CEOs in that type of position think about interacting with their team who are the direct users of the AI? Are there complications where that maybe level of detachment from the actual use of the AI
brings different dynamics into play on the team.
Bryant Oberg (20:53)
Yeah, mean people are very scared of it 'cause they think it's gonna take their job. Right. And they think the CEO CEO is gonna take their job, so they're very hesitant to use it. Or they're very frustrated 'cause the CEO hasn't implemented anything and so they're doing what's called shadow AI and they're using AI on their own personal systems and taking all the company information and proprietary stuff over to a personal AI chatbot or agent and and using that without the company knowing.
to help facilitate their work. Right. And so I think it's the CEO's job to the the first thing is really, and and this is something I'm talking a lot with my colleagues about in this space is the governance of AI. Right. What is your policy for it? And not having a policy is not adequate anymore. Right. Yeah. It because people are using it regardless. And depending on what you do, do you want your team using it in this shadow capacity?
Or do you want them using it in the open? And if they are doing it in the open, how do you get it where you feel like all your team members are a part of this? and there are some things, and I you know, the thing I keep going, I always go back to kind of how I started the whole business, which was these one-on-one sessions, or now I do group trainings, right? And it's about getting people to adopt or really adapt to the AI that's that's being that's supposed to be used because the buy-in is the first thing. It was like we were saying before.
The the models are really good, right? The the AI is there to really benefit you. But if you don't know how to use it or you're not using it, it's not gonna benefit you. So you have to get people to over that hump. And there's a there's there's a great analogy I like to use, which is the elevat the analogy of when elevators came into existence. Okay. So in the thirties, early the the twenties and thirties, when the elevator was invented, right, you
Before what did you used to have? I'm sorry, when you had the actual the the automatic elevator where you could press a button, right? Before what did you used to have? You used to have it it was manual, okay? So you had the elevator man and you had the and then when it became automatic where you press a button and you went up and down, which was like magic, right? Kind of like AI today, people were terrified. They wouldn't go into the elevator. So what did the elevator companies do? They put the elevator man to stand there.
In the elevator
Jeff Holman (23:19)
Back in the elevator.
Bryant Oberg (23:20)
back in the elevator, right? To press the button. Right? He was the elevator operator, right? Because you didn't need him to press the button, but you needed someone to give you the comfort in order to get in the elevator. And I feel like CEOs to an extent have to work have to use that paradigm with their employees. They have to get people comfortable to get in the elevator. And once they get in the elevator,
Then the then most people are like, wow, I didn't know it could do this, or wow, it it did my report in fifteen minutes. And most people are amazed. But if you have a bunch of people who are not willing to get into the elevator, it makes your job really difficult as a CEO of, you know, anybody who's y not operating as so as a solopreneur.
Jeff Holman (24:09)
Yeah, I know I I can I can imagine that. I even on my team we have six attorneys and and a a few staff members. So we're a small law firm what we do, but I I I know for a fact, at least I I think I know, that the AI usage among our team members varies significantly, right? And and I think that's pretty typical in the legal field where you have some people who are they want to use it extensively, some people who don't wanna ever touch it. And so that's really interesting. Is the is the AI policy the elevator man or is it is who
Who's the who or what is the elevator operator man, i in these scenarios?
Bryant Oberg (24:45)
Yeah. it's gonna vary be it's gonna be e each team's gonna be different, but it starts with the conversation, the policy, and really education. Right. Is it you know, some people are very risk averse, so it's teaching them with a policy what's okay, what's not, right? you know, there's a big difference between having a free chat GBT, a personal chat GBT, and a chat GBT business, right? Yeah. And
It's there's a big difference between having the toggle on where your data can be shared to train models and not. Right. And just, you know, these are not, you know, super complex things that take a ton of time to to roll out for companies. But most people just haven't either had the time or the or the wherewithal to put it put in a just a three-page policy, right? We use this type of AI. It needs to be done this way, right? your toggle always has to be off. You're not allowed to use these.
Data reports in your personal AI. And then once you create that, let's call it that landscape, you can really worry about building the house, which is going to be your house in in this is going to be, you know, the house of AI, which is going to, you know, make your company fly. But you have to you have to take people and educate them first.
Jeff Holman (26:05)
Yeah, it makes this is super timely. I you you you wouldn't have known this, except I'm gonna tell you that my marketing team this week came and said, Hey Jeff, we really need to turn the web search on for our for our team account so that we can use it for some of the marketing things we're doing, you know, with the podcast and such. And I'm like, Yeah, that's fine. We've had it turned off for a reason because, you know, w we can't do we can't put client information and law firm confidential information into the into chat and you know, have it go s be sent out to a third party search.
company and w they have their own policies and it like like we as a law firm we just can't do that. And so I actually had to update our policy yesterday because I said we we need to we need to turn on web search, but we need to have some guidelines in place for how our legal team marketing it's not that big of a deal. You know, they're they're not they're not using client information. Our our attorneys and staff are definitely using client information and we need to have policies in place so that if web search is turned on, we are still
protecting the the information i the right way i in any chats that might involve confidential or client information. So just just yesterday and today I've I've been implementing that and, you know, turned on web search with some very strict guidelines about when and where that can be used and how everywhere else it can't be used. So
Bryant Oberg (27:28)
That's great. You're a step ahead. You're really a step ahead.
Jeff Holman (27:31)
okay, okay. Awesome. I'm I'm ahead of crowd. well I was thinking too, and when I asked that question, who's the elevator man? I I you know, I I guess one of the obvious answers is Bryant. You you're one of those you're one of those guys, right? So I I imagine there are people who bring you in to say, Hey, listen, our team's not totally comfortable with with the this new fangled elevator, the automatic elevator yet. Like, you know, could you they need some practice, but they need a little little guidance and comfort along the way. is that something you do?
Bryant Oberg (28:01)
A hundred percent. Yeah. I come in and the the role is sometime called the C AIO now, chief a fractional chief AI officer. Yeah. CTO, whatever you want, a chief technology officer, but it doesn't have to be, given a title. It's a consultant that comes in and tries to, you know, make the e the playing field even, right? Why is a consultant great to bring in? Because they're a third party that doesn't have a stake. If everyone gets mad at them, they leave at the end of the day. If everyone's really happy at them, they leave at the end of the day, you know.
And it allows everybody to bring both their, their baggage but also their strengths to the table in whatever I mean, this is just general consulting, but when we're talking AI specifically, and figure out exactly what is the impediment for the team to take the next step to really it everyone agrees AI is a difference a difference maker. The problem that everyone's coming to is, well, how? Right. And so
I actually do both those things, right? Is I ch I I do the education and I get people over the hump to say, We're doing this AI thing and then I actually get us get teams and companies to the how does it make the difference.
Jeff Holman (29:12)
Yeah. Well I I love that you're mentioning that because I think there are a lot of I get a lot of pitches for the podcast, frankly, from people who are they they offer AI
Bryant Oberg (29:20)
Do
you need any I any I consult?
Jeff Holman (29:22)
That's right. Any I d but but so there are a lot of people out there doing that, but and and and that may be part of your business model also. but I love that we're talking on here about not just everybody should be building AI, but there's this there's this psychological aspect to getting your team on board before it it's giving them the permission or the skills or the comfort level before they before we even dive in and say, You gotta be building AI for content generation and distribution and
C CRM management and like yeah, that all that's important and probably a lot of companies should be doing that. but this is this is the conversation that might be the precursor to a lot of them pulling the trigger.
Bryant Oberg (30:08)
I I I have seen that and I suspect so for many, many companies.
Jeff Holman (30:13)
I love it. Well well I appreciate you coming on the show today, Bryant. if people want to get a hold of you, they wanna talk to you, they wanna they wanna hear about your elevator man services or or other things that you're doing, what what's a good way to connect?
Bryant Oberg (30:27)
Yeah, sure. So Human Copilot is or LinkedIn are my two two main modalities that connect. human co-pil is human dash co dash pilot dot com. and you can find me there. You can connect to me on WhatsApp through the through the website or fill out one of the forms. You can take an AI quiz or even do a fill out a little survey to see where AI could help your business. And and LinkedIn is also a great way to reach out. I respond to
hold messages or just posts or anything. So those those are the two best ways to reach me. I'll drop my email is also Bryant at humancopilot.com. That's human-co-dash pilot.com. So just my first name at human copilot. And it's it's been great be the show, Jeff. It was a it was fun. And I hope I hope some people learned a little bit about AI or at least became a little more human with their AI use.
Jeff Holman (31:22)
I love it. And I'm gonna ask one more question. I just because I think it's appropriate here. What's the what's the question or the thought that somebody would be asking themselves that that they they should flag them to say, you know what, I really I probably should be calling Brian?
Bryant Oberg (31:36)
Too busy.
That that's the one that's what I hear so much is I'm too busy. If you're too busy, that means you need someone else to help you. Okay. Because you're not going to get to go do it yourself, right? Whether it's you need the AI policy, you need the training, or you need some AI tool implemented. Almost every CEO across the board agrees that AI needs to be somewhere in their business. Most CEOs are too busy to even know where to begin. And so they have to call up somebody like me to really help them figure out.
where th where they need to start, how they can implement it and what the change is gonna be.
Jeff Holman (32:13)
Okay. Well that's a that's probably a question that a lot of people in our audience are is a statement they're saying to themselves as as we speak. They're like, I'm not sure I have time to listen to this podcast, let alone go build everything else I need to build. So good. Well, Brian, it's been a pleasure. Thanks for coming on the show.
Bryant Oberg (32:30)
Thanks very much for having me.
Jeff Holman (32:31)
And thanks to our audience for joining us on another episode of the Breakout CEO. Be sure to follow or subscribe on your favorite podcast platform. And if you enjoy the show, a rating or a review goes a long way. Our mission is to promote the stories of breakout CEOs in scaling SaaS, e-commerce, and tech companies to equip peer CEOs with valuable perspectives and confidence.
Thanks again for joining us on this episode of the Breakout CEO. I'm Jeff Holman, and I'll see you next time.
