Ep #15 - Scaling Ethics: Architecture, Accountability, and the Future of Leadership w/ Ryan Roberts

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 PODCAST (00:01.812)
Ethics can't be an afterthought anymore. In a world run by algorithms, integrity has to scale as fast as the code does. We have built systems that can make decisions at the speed of light, but how we build leaders who can think at the speed of consequences. Welcome to The Human Protocol, the podcast about staying human while building the future with AI. I am your host, Mykel Salomon.

Technology is scaling faster than our governance, faster than our laws, and in some cases, faster than our ethics. From AI power recruiting to autonomous decision systems, we have reached a point where the biggest risk isn't what the machines do. It is what we forget to design for. Because ethic isn't a policy. It is an architecture.

Today, I am joined by Ryan Gus Roberts, who helps founders and executives build companies that survive scale. A former US Air Force fighter pilot and senior strategy officer, Ryan, thank you for your service. Ryan went on to lead teams developing advanced weapons risk modeling and simulations before transitioning to entrepreneurship. He founded and later sold

a defense contracting firm, while running it remotely from a cell bot in the Caribbean. An experience that taught him that freedom and scale both depend on a strong architecture. Today, as Director of Operation Averical Motion, Ryan applies systems engineering and behavioral science to leadership training and organizational growth. He is the creator of the Architecture Maturity Index.

A framework that helps leaders uncover the hidden dynamics that determine whether their organization will scale or stall. His current work, the Vertical Architecture Intensive, builds on that foundation using the same methodology to identify the critical vulnerabilities that quietly kill growth and make fixing them clear, measurable and executable. Ryan.

THE HUMAN PROTOCOL PODCAST (02:28.14)
Welcome to the Human Protocol!

Ryan Roberts (02:30.096)
Yeah, thanks Michael. Glad to be here.

THE HUMAN PROTOCOL PODCAST (02:33.226)
Awesome. We're very happy to have you today. And thanks for taking the time to have this conversation with us. And to kick things off, I usually like to get these conversations started on a more personal side of things. one, which is one question that is when we remove this bio, right, that I just went over.

What LinkedIn knows about you, what Shaji PT knows about Ryan, who is Ryan at his core?

Ryan Roberts (03:05.541)
Wow, what a great question. And unrehearsed, good for you. You know what? I tell you what, and I think people who know me well, certainly those close to me would tell you, maybe one word is curious. I've learned for a long time to remain curious. So I think that's an adjective that I...

would apply to myself in a lot of means and applications. Certainly it started really young as a curious person with an engineering type of mind. I learned really early how important it was to be more than surface level curious, know, how to contingency plan, be curious at a deeper level of what are second and third order impacts of a decisions. So yeah.

THE HUMAN PROTOCOL PODCAST (03:53.038)
Mm-hmm.

Ryan Roberts (04:02.382)
I would say I'm curious and that's potentially why I think my life story is a bit more non-linear than others. Certainly my career and the places and the things I've done is because I remained curious. So I appreciate that question though.

THE HUMAN PROTOCOL PODCAST (04:11.694)
Mm-hmm.

THE HUMAN PROTOCOL PODCAST (04:18.482)
And no, and I appreciate you being so personal about the answer, which it makes all sense, right? That curiosity took you to places in life, right? And different directions that without you being curious, probably you were not going there. So I think that makes perfect sense. And especially for the conversation that...

we are about to have today, which is based on scaling ethics, Architecture, accountability, and the future of leadership. So that makes total sense. And thank you, Ryan, for being so personal about it. So we know we have an ethical dilemma, right? Especially in times of AI that is not discussed a lot.

So I'm curious just to set the foundation of what this conversation will be about. What does it mean for you when people or when we are gonna be talking today about ethical architecture? What does it mean to you?

Ryan Roberts (05:32.39)
Yeah, right. Well, first of all, I think, as we all know, when this word comes up, ethics, that it's not morality, right? Morality is a very personal and internal thing. And then ethics has to be, you know, it's the group think, it's how the organization or larger or as a population.

THE HUMAN PROTOCOL PODCAST (05:40.034)
Mm-hmm.

Right.

Ryan Roberts (05:56.824)
or as a world, how do we, those are the levels of ethics and how those frame decisions based on group think. And so when we think about ethics and architecture and why it's important, it's because once we step out of morality into ethics, we are now talking about a system, right? We're talking about people.

We're talking about, and it doesn't matter if it's just a small group, it doesn't matter. That's the ethics of the group. If it's an organization, if it's a group of organizations, if it's a country, that now makes it a system. you know, to me, it's incredibly clear, right, that physics, if physics governs the natural world and we have sciences like biology that govern life,

then we know what governs systems, and that's architecture. And business is a system. And so when we think about how organizations are going to implement considerations for AI and where that comes across various domains, it's architecture, which means now you have to have a plan. You don't react, right? Architecture and the systems and the frameworks that govern that are what govern

THE HUMAN PROTOCOL PODCAST (07:13.632)
Hmm.

Ryan Roberts (07:21.122)
systems including the ethics of AI.

THE HUMAN PROTOCOL PODCAST (07:27.138)
Right, and I think that's perfectly tied to your engineering background, right? Because I was thinking about how your engineering background really influenced how you think and how you approach leadership. But I think you just answer right there. It's all about the architecture that you put behind it. And based on that, that's where you make decisions. So I think they're perfectly tied between them.

Ryan Roberts (07:53.764)
Yeah, I'm sure that's true. I've always had an engineering type of mind as far as being curious and making sure we understand systems enough to ask the right questions and consider what questions we're not asking. My role, of course, I am not a licensed professional engineer. My background, flying jets just happened to lead me to leading teams of

THE HUMAN PROTOCOL PODCAST (08:22.808)
Right.

Ryan Roberts (08:23.292)
really smart, really great engineers, which led me into a part of my career where I dove deeply into systems engineering, which is what this is. Systems are systems. It doesn't matter if it's a guided weapon and you're trying to consider all the multitude, hundreds of thousands of components that could fail in the system that lead to what we would call a failure or in business of vulnerability.

THE HUMAN PROTOCOL PODCAST (08:38.093)
Yeah.

Ryan Roberts (08:53.678)
It's all the same. so, yeah, I think it's just my mind that was always curious. And then when it led to risk assessment, which is what I did for years, it all kind of was a capstone of all of the years that led up to it, including, you know, trailing over into not only leading those teams, but also business. And I realized how it's all the same.

THE HUMAN PROTOCOL PODCAST (09:15.779)
Right.

Ryan Roberts (09:18.898)
It's all the same system. It's all the same type of frameworks, the same criticality analysis you have to do. with regards to today's conversation, AI is certainly a mechanism that could lead to a vulnerability or a failure in business. And even amongst, I would say more than that, businesses, economies.

It's an incredibly important consideration that we don't just wait around to react. Which I think everybody might feel or many people might feel we're already doing that. But that's okay. That's okay. We have to continually to push with an engineering mindset of what are the potential failures and more incredibly more important is what are we not considering.

THE HUMAN PROTOCOL PODCAST (09:51.927)
Right.

THE HUMAN PROTOCOL PODCAST (09:58.787)
Mm-hmm.

THE HUMAN PROTOCOL PODCAST (10:13.084)
boy. It's exactly that's the point of what keeps me awake at night, right? What is it that we're not considering? Because we're always hoping and sometimes, you know, we build systems hoping people would do the right thing. But I don't think we're building systems enough to make it hard for people to do the wrong thing. And I think in all these AI conversation, I think that's the part that I'm missing on it. It's the

how we keep ethics in what we're building as far as making it harder for really people doing the wrong thing. And that's kind of what keeps me awake at night. I don't know if people are having this mindset of really building these systems in a way that it's keeping ethic as part of the conversation.

Ryan Roberts (11:13.65)
I think, I wouldn't say that it keeps me up at night, but to me, that entire sentiment is well founded. Look, if businesses exist, presumably to make a profit and to have and keep customers and to generate top line revenue, then why do they fail, right? Well, you know.

THE HUMAN PROTOCOL PODCAST (11:24.558)
Hmm.

THE HUMAN PROTOCOL PODCAST (11:30.68)
Bye.

Ryan Roberts (11:40.272)
We know that the overwhelming number one reason is they run out of money. Right? And so if that's the case, then clearly had they known some period before, they would have made a different decision, which means it was a blind spot. Whatever led to them running out was something they either weren't aware of at all, which happens all the time, or something they clearly didn't think was as critical.

THE HUMAN PROTOCOL PODCAST (11:45.198)
Mm.

THE HUMAN PROTOCOL PODCAST (12:01.923)
Right.

Ryan Roberts (12:06.156)
And so I think it's the same thing here. If running out of money is the number one vulnerability that kills a business, failing to architect for AI could be the next. And so you have to make sure that we are looking at AI, think, to some people feels unknown. And that should be the first signal of risk. That if you feel like you don't know,

then in business if you can't audit it, then you've got something critical on your hands.

THE HUMAN PROTOCOL PODCAST (12:39.31)
then there's a problem there, right? And I was reading the other day, I think it was a Deloitte Ethics and Trust report saying that while 67 % of technology executives say AI ethics is a top priority, only 24 % have a way to measure it or are even measuring. So how can that be a concern?

Ryan Roberts (12:42.194)
That's right.

THE HUMAN PROTOCOL PODCAST (13:08.716)
without you being able to measure and see how things are working. when I read things like that, it's in part of back to this conversation, if the accountability and the frameworks are not, or the systems are not there to measure accountability on this system, what are we building?

Ryan Roberts (13:28.818)
That's incredible. What was the number? 24 %? 24 % said they.

THE HUMAN PROTOCOL PODCAST (13:31.87)
Only 24 % have framework or accountability in place to measure.

Ryan Roberts (13:39.258)
Wow, well that should scare 76 % of business owners. I mean, really scare, right? This goes back to what we were just saying about when you do risk assessments and figure out what's critical in an organization, right? When we measure criticality in a system, we are just multiplying the likelihood of the occurrence, right? The rate at which that thing could occur. We have plenty of data, we know why businesses fail.

THE HUMAN PROTOCOL PODCAST (13:45.495)
Yeah.

THE HUMAN PROTOCOL PODCAST (14:02.882)
Mm-hmm.

Ryan Roberts (14:06.342)
then the multiplier is the severity at which that can impact the business, right? But we have to divide by the inverse of the detectability, right? Systems fail to the level at which the failure or the vulnerability is even detectable. That means inherently in their system, we've got 76 % of people, business owners and businesses, at least that many, that

THE HUMAN PROTOCOL PODCAST (14:30.766)
Mmm.

Ryan Roberts (14:35.184)
have now identified something that's not just potentially severe, it's critical. It becomes critical because you have admitted, I don't know anything about it, I don't know how to audit it, I'm not even sure what I should be looking at. Detectability falls incredibly low, and it seems like instead it's just replaced with this blind trust. And I can tell you that the 80 % of businesses that fail because of cashflow issues would wish they had some more insight.

THE HUMAN PROTOCOL PODCAST (14:40.3)
Right.

Ryan Roberts (15:04.388)
And I think that just tells us alone exactly why this potentially is not being analyzed across the domains that it should be. It's not just risk and compliance either. We'll talk about that, I'm sure, because we're seeing some compliance and some walls floated. But it's also the systems and the operations. And it's interesting what you just said about how business owners are intending to potentially use it.

Because now you're in the strategy domain of business. So it's not just the operational piece. Is your business strategy relying on AI? So you've got several domains, it seems to me, that potentially have some critical failures on their way if you don't figure this out.

THE HUMAN PROTOCOL PODCAST (15:53.128)
And it's how can you build accountability loops in your organizations and system if you don't have a way to measure them, right? If you don't know what you don't know. And this is part of what, when I read things like that or reports like this, it's okay, I think there is trouble because if people cannot measure, then how do you know how it's performing or how are you gonna improve it or how do you know that this is behaving as you expect it to behave?

So it's troublesome in a way, right?

Ryan Roberts (16:28.514)
Yeah, I mean, there's no question. I mean, that is exactly what we're talking about. If you can't audit it, then you can't possibly have the visibility, which means you've got a detectability issue. You've got a critical issue on your hands. And I'm glad that there's discussions like this and what you're doing to kind of start exposing this, right? It's like anything that's new in business. It's like any paradigm shift.

THE HUMAN PROTOCOL PODCAST (16:34.478)
Mm-hmm.

Ryan Roberts (16:54.17)
And what can kill businesses that don't adapt is just this burying their head in the sand of the, I'm just going to trust that it's going to work itself out. And we've seen it over the last few decades, right, as we've entered more and more of the digital age with paradigm shifts. And if you don't adapt and understand, then businesses fail because they don't. And so I think many can agree that AI may be even a more

critical aspect because of all the pieces of the business that could affect. I mean, there's a lot, right? We start thinking about all the ways that AI and not considering it in a business where it touches. you know, it's the ethical understanding of where and why. But I guess it's first of all, what do mean by AI, right? I feel like a lot of business owners

potentially don't know enough to know how much they may have exposure to various AI systems, right? Like some people might think, know, well, we don't, you know, we don't have an exposure. I don't think that is potentially true almost at all anymore in business. And that's potentially one of the poor assumptions.

THE HUMAN PROTOCOL PODCAST (17:58.446)
to it. Yeah. Right.

THE HUMAN PROTOCOL PODCAST (18:08.962)
Yeah.

THE HUMAN PROTOCOL PODCAST (18:13.912)
Right. And for even a lot of people, they may have been using AI, they don't even know that they're using business owners as well. But it keeps, you know, how do you see ethics in business today, especially in AI driven businesses? Is that still an afterthought? I'm not hearing a lot of conversation from business owners talking about how important is the ethical dilemmas and things to keep in

in this relationship, or in these systems that we're building. How we forgotten about that? To talk about how we keep this system as ethical as they should be, or it's just about making money that we are just focusing and worry today.

Ryan Roberts (19:07.148)
The good news is I think you can do both. In fact, think you have to. Because the same audit that's going to be required to have an understanding where it exists in your business, what's coming up in the future, right? That's part of the strategy that leadership should be thinking of in one of the domains. Then again, it goes back to you have to have enough of an understanding to at least know

THE HUMAN PROTOCOL PODCAST (19:10.572)
We should. I think we should. Right.

THE HUMAN PROTOCOL PODCAST (19:24.396)
Mm-hmm.

Ryan Roberts (19:33.434)
what your blind spots must be and then at least where you sit currently. Because you can't ask ethically without understanding what exactly you should be studying. Right, I mean, we're talking about, let's just talk about the way AI might be used in a business for, let's say, automations, right? Well, if you don't have accountability inside,

And like you said, if we're not auditing it, we don't know the accountability of what the automations then not only do there. Well, there's lots of reasons why we might have critical vulnerabilities and failures and. Imagine biased or. Flawed output, right? So you got that. And so what's the ethical issue with bad data that you may be putting out to clients or misunderstanding yourself?

THE HUMAN PROTOCOL PODCAST (20:00.469)
Mm-hmm.

Ryan Roberts (20:27.41)
Imagine getting that data and that leadership executed confidently, right, on the wrong data. We talk about this a lot in systems. Are we even aiming at the right target? And how do you know if you're not auditing the systems that are helping you make decisions? You mentioned feedback loops. Do we have systems in place to make sure that whatever we think we understand and whoever is responsible in the organization has feedback so that every time we can audit those?

THE HUMAN PROTOCOL PODCAST (20:33.602)
Hmm. Yeah.

Ryan Roberts (20:54.404)
And then I think what happens with new technologies like this is the drift just starts becoming invisible. And that I think is the same like a lot of technologies that have burned businesses in the past, where you allow that drift, you know, not only become invisible, but you're almost bury your head in the sand. And now you're not keeping up. And there's definitely not only going to be ethical implications.

but your business is going to, know, I mean, losing trust. Look, we're all still humans, right? But humans, AI, at least so far, is not cinching on its own. And so it's still a tool. And at the end of the day, humans and behavioral psychology is what drives human decision making. And what we at least think is best for our lives. And that's what drives business.

THE HUMAN PROTOCOL PODCAST (21:24.622)
Absolutely.

Ryan Roberts (21:52.1)
Certainly it's what drives capital markets. And AI now becomes this tool. And if you aren't understanding how that might affect other people and the trust in the, even if it's something as simple as you are allowing systems to drive decisions and those decisions are wrong, then imagine that the trust you lose as an organization, not the least of which is all of the ethical exposure like we're seeing in Europe finally bringing up about.

collecting data and personal data. There's just a lot to consider for businesses.

THE HUMAN PROTOCOL PODCAST (22:27.0)
Yeah, absolutely. the over relying on it, it can cause that when a system fail, that's because a human being stopped asking the hard questions, right? Or it was not kept in the middle of this whole interaction or relationship. And that's part of what I also think about is the over relying on businesses and people implementing these solutions.

not really thinking it through, right, and not keeping the safeguards in place that will guarantee that we architect things for success and not for failure. And that's still up to us human beings to be part of this conversation, right? It's...

It's interesting to see where we're heading as how much are we going to continue to have leverage on building these systems, architecting and designing it, but also making sure that there are check marks that human beings are still part of the conversation.

Ryan Roberts (23:37.298)
Yeah, and I think, you know, unfortunately, the data is clear that most businesses don't spend enough time architecting. As you've already mentioned that they chase the dollar, right? How many businesses just chase top line revenue, top line revenue, and that's what they chase. And unfortunately, that is just an ignorant bias of what actually has businesses succeed. Right. In fact, we know that one of the contributing factors

that is one of the mechanisms that leads to business failure is trying to scale too early, right? And so we could say the same thing here, which is that if you're not architecting for this potential vulnerability, just like you would for anything else, then you're exposing yourself. But like we've already said, we already know that most businesses fail. In fact, in the online space, it's about triple the rate of failures.

THE HUMAN PROTOCOL PODCAST (24:28.814)
Yeah.

Ryan Roberts (24:35.536)
So there's plenty, you're right, right? It's scary numbers how quickly online business is there. Yeah, yeah, absolutely. And so if that's the case, then we already know that inherent to a lot of businesses is not the architecture that should happen. Instead, they chase top line revenue as the,

THE HUMAN PROTOCOL PODCAST (24:35.878)
wow.

THE HUMAN PROTOCOL PODCAST (24:39.763)
It is. I did not know that.

THE HUMAN PROTOCOL PODCAST (24:54.766)
Mm-hmm.

Ryan Roberts (25:02.006)
as the only benchmark that means something in all of the underlying systems, especially when businesses are trying to scale, all, you know, don't support. margins fall too small. You got all kinds of other issues that leads to. And if that's the case, we have also now just identified that most companies are certainly not architecting AI and the risk and vulnerabilities of how it might show up in their business.

And for those that, you know, kind of get past that viability stage and they're trying to expand systems and they should be looking at ways to ethically implement AI and be part of a bigger conversation so that they continue to build trust. Because at the end of the day, again, this is a, have customers and how do you influence them by getting them to trust you and the product or service that you have. If you're not architecting this as part of that.

THE HUMAN PROTOCOL PODCAST (25:44.462)
right?

Ryan Roberts (25:58.726)
then not only you're likely to fail, but you're just going to be part of the statistic, which I think I'm just trying to come back around to what you said, which is that is probably what keeps some people awake at night, is knowing that the data tells us that most people, most businesses are not going to give this the consideration they should.

THE HUMAN PROTOCOL PODCAST (26:21.39)
Totally. think we've spent decades, Ryan, getting good at designing, making sure that engineering efficiency was there, architecture, that we should not just drop everything just because of AI, because that's still the backbone of what's going to continue to build solid systems moving forward. And I think...

And this is why this conversation is also so important. It's back to the accountability conversation, right? We spend decades on engineering efficiency. We have gotten so much better at everything that we do there, better at coding. This is part of why AI exists today. That is, I don't think it's just for us just to say, let's leave this for AI to do everything, because that's when I think we're going to be in trouble.

We need more accountability back to this equation, right? That's what I think it's missing a little bit right now.

Ryan Roberts (27:26.116)
Yeah, I think it's more than a little bit potentially. I think it's a giant blind spot. Look, if we've known for decades the type of things that businesses should be looking at inside the various domains, right, across strategy, across systems and operations, across risk and compliance and data, go to Market Engine, which again, I could go back to what I said about top line revenue, that businesses loved and love, love, love to

THE HUMAN PROTOCOL PODCAST (27:33.699)
Yeah.

THE HUMAN PROTOCOL PODCAST (27:49.25)
Yes, absolutely.

Ryan Roberts (27:54.172)
to spend their time as if business is the go-to-market engine. And that's just one domain. if we already know that historically that business leaders aren't particularly great at making sure they're doing vulnerability assessments on a regular basis, and we know that. Why? Because of the failures. Then I think it's at least a reasonable implication that something new

like this isn't being audited. And like you said, if the idea that you're just going to trust something without understanding it means that you're not measuring it, which means it's now not detectable. The failures you could have, the bias in the information. And again, all of these things can eventually, you know, even the smallest business could be absolutely.

crushed under the weight of mistrust that will come later because you decided to outsource your accountability to a machine. that's the whole reason for feedback loops. And the way we build those systems is so that the accountability remains in the loop, and so does the humanity in this case.

THE HUMAN PROTOCOL PODCAST (29:01.518)
Yeah, completely. Yes.

THE HUMAN PROTOCOL PODCAST (29:10.561)
and

Ryan Roberts (29:20.782)
And it's just really simple, right? A system you can't audit is just simply a system you can't trust. And this system and this particular tool is no different than any other tool we use in business.

THE HUMAN PROTOCOL PODCAST (29:28.014)
Yeah.

THE HUMAN PROTOCOL PODCAST (29:35.054)
And that's where you as a leader, it's not only that you need to understand the consequences of what you're building. You need to care about those consequences first. And that's where ethics comes into play because it's not only building something and make money out of it. What are the consequences of what you're building? And not only thinking that this will always work, well, sometimes it will fail.

What are those consequences as well? And do you really care about it to a point that you do something and you build better systems, right? And that's where I think leadership of these next three to five years will really need to pay more attention on if you are building a product, what is the impact of what you're building on us human beings? And what are the safeguards, right, that you...

you will be putting in place to make sure that you're keeping those systems accountable and you continue to have people making sure that the system is designed for what you envision and what your company's vision and mission is, right? But especially for those things that is not supposed to be, which is those type of conversations that I wish every business leader think about.

What is it that I'm building? What is it for? But also what is not for and what I'm doing to make sure that people don't get there in a way.

Ryan Roberts (31:14.608)
Yeah, think that's, I think you're onto something. I think it needs to be, you know, whatever we can do to raise the awareness, right? Because you can't, you can't automate ethics. I mean, it's, it's, it's, you can't. And if you, can only architect for it. Right? Right, right. You can't automate that. It's literally, you can't have feedback loops to know the decisions that are being made in the dark. So.

THE HUMAN PROTOCOL PODCAST (31:27.692)
You can't or morality, right?

Ryan Roberts (31:42.706)
I think that's why this needs to be a greater understanding. And I think you're aware that not enough business owners and leaders even understand the breadth and the depth of artificial intelligence because it's not just large language models. It's not just single-framed automations. There's a whole lot of it, including data collection.

THE HUMAN PROTOCOL PODCAST (31:57.294)
Yep.

THE HUMAN PROTOCOL PODCAST (32:07.534)
you

Ryan Roberts (32:12.37)
Now, like I said, we talked about what's going on in Europe and some considerations. There's just a lot to consider. And I think the good news is I'm seeing more and more leaders across a lot of industries that are now pursuing more education in AI. I see it all the time. I think that's great because those are going to be the leaders, I think, in the next few years. As, yeah.

THE HUMAN PROTOCOL PODCAST (32:35.202)
I think so.

Yeah, I think so. And especially in the next five to 10 years, think leadership will be redefined, not by what leaders are building, but what they are preventing, which is, for me, is the most important part of this. It's not about we are excited about what's being built today. It's just amazing. You know, we are getting a lot of efficiency. We are getting a lot of good things, but

there is not a lot of conversation about when things go wrong, what will happen to us and what are we doing there. And that's also, think, it needs to be part of this conversation. And you can only prevent things going south or going wrong if you first care about it in a way that you are architecting and designing your system to work the way they're supposed to.

think it's important, not only what we built, but what we prevent from happening.

Ryan Roberts (33:42.812)
That's right. That's right. Absolutely. It's about the detectability, right? High severity times high occurrence times low detectability is critical, critical in the system. So I think that's exactly what we need to do is the more visibility, the more light we can shine on it, the more education around it as opposed to, I think it's an overwhelming thing for a lot of leaders that they just, feels like something new and scary.

THE HUMAN PROTOCOL PODCAST (34:06.369)
it.

Ryan Roberts (34:09.518)
And the good news is there's a lot of bright people out there that can explain it. mean, I would die. And now I'm just thinking about that. My guess is we're going to see more AI consultants in the future to understand, you know, the blind spots.

THE HUMAN PROTOCOL PODCAST (34:20.812)
Mm-hmm.

THE HUMAN PROTOCOL PODCAST (34:24.79)
Yeah, and to potentially explain and talk to regular people how they can leverage those technologies in a way that people that are not in technology or tech savvy can really understand how to get the benefits out of it. I think I agree with that as well myself. So Ryan, I'm just curious and I'm wondering if you can...

Tell me a little bit about this framework, the Architecture Maturity Index. What is that about and how that was built? If you can tell us a little bit about around that story, that would be great, because I'm interested and I know people will be as well.

Ryan Roberts (35:12.71)
Yeah, I suppose it was built over decades, right, of a career of seeing risk in systems before I understood it in business. And again, it goes back to my comment I made earlier about businesses that tried to scale too soon. It's a common vulnerability. And so because of that, there needed to be a way to measure like

THE HUMAN PROTOCOL PODCAST (35:35.16)
Yeah.

Ryan Roberts (35:41.904)
what are the, we know the failures, right? We can look at historical data and we know that we can pull out the laundry list of all the ways and the rate at which those type of mechanisms show up in failure in business. And if you look across and what we do inside that in the AMI is look across eight domains, right? So we're looking across various.

sectors of the business, as I mentioned before, it's leadership's capacity, also on top of strategy and systems that go to market engine. There's eight domains and across those domains, we wanna go and find the vulnerabilities by measuring the criticality. But we start with an index that's measured of how mature the business is right now in each one of those domains.

And it becomes a, to some degree, a systems engineering equation of how ready they are across those domains. But the criticality gets layered on based on, like I said, an additional multiplier, which is what's observable. So you can imagine that, like, if we onboard a client and I go through a questionnaire before I get the data, then I already can see the gap.

potentially between what they perceive and the reality. Then on top of that, the data can be analyzed to see where they have additional vulnerabilities and things that may be unobservable. And there was really two considerations in this whole framework that leads to this type of system engineering to make sure that you take something that might be really complex.

and make it right so that a leader can look at it and go, ah, those are my three most critical things I can do this quarter to make sure. And, and then I would say that, that again, I started young with the need to have contingency plans and to understand risk. And that goes back to me specifically, um, on my journey to, um, to become a fighter pilot in the air force. I had my own unique.

THE HUMAN PROTOCOL PODCAST (37:37.506)
that I need to work on.

Ryan Roberts (38:00.722)
issues that should have precluded me from being able to get there. I was told that it was impossible for me. Well, for me, that now became, unfortunately, I had a mentor who kind of opened my eye to understanding risk and considerations for contingency planning in such a way that it rewired my brain to essentially go through life always prepared to ask the second and third order questions that might lead to it.

And then I spent years running modeling and simulation for guided weapons. And in that type of engineering analysis, you do something that's called FAMICA, failure modes, effects, and criticality analysis. And it's the same thing we're talking about across a business analysis with slightly different mechanisms. But it's the same, that you are finding the mechanisms that fail.

the rate at which they fail, how bad they can fail, and how observable it is. And then on top of that, having been exposed across several business best practices, we all hear of Six Sigma, right? And certainly larger organizations might employ people that understand these principles. And the problem with that, from a statistical understanding, is that Six Sigma only measures variability.

THE HUMAN PROTOCOL PODCAST (38:57.496)
Right.

Ryan Roberts (39:23.466)
That's what you're talking about when we talk about standard deviations. And that's why originally that was designed by Motorola, I think, in the 80s. It was to solve a problem in variance so that they could get it down to component failures very, very small. Well, that's great with a large sample size and manufacturing. AI and these other issues highlight that this isn't...

THE HUMAN PROTOCOL PODCAST (39:47.437)
Right.

Ryan Roberts (39:53.094)
You largely don't have large enough data sets. Humans are involved, so it's not a normal distribution. And so for all those reasons, that's a failure if you think you're using that to analyze your business, except for only a few places where the statistical analysis would line up to something useful. And so altogether, that's where these ideas all just compounded into what is a better system, which is

THE HUMAN PROTOCOL PODCAST (39:55.736)
Mm-hmm.

THE HUMAN PROTOCOL PODCAST (40:07.224)
You're right.

THE HUMAN PROTOCOL PODCAST (40:19.832)
Mmm.

Ryan Roberts (40:20.07)
this architecture maturity index that first measures that, but also a vertical architecture program where we stay involved in measuring the criticality and then measure it over time, presuming that leadership is going to take some action on those things and then measure it quarterly so that you're headed towards what the ultimate goal is, is survivability, which at the end of the day, when we talk about business success, it's about how you can survive.

THE HUMAN PROTOCOL PODCAST (40:44.63)
Right. Yeah. Yeah. Awesome. This is very interesting. And I think this will be very beneficial and helpful to early stage founders and startups, right, that may not have the guidance or may not fully understand their business. And having this may help them to better understand what they need to do to survive.

Ryan Roberts (41:10.316)
Well, yeah, and that's the whole thing that sometimes and it's understandable, right? Human nature, you've shown that you've got a product or a service that is viable. And now all you can think is let me just sell more, sell more, sell more. And that kills a lot of businesses because you don't put the systems in place and you don't do the vulnerability assessment, including the domains that AI would potentially pose risks.

THE HUMAN PROTOCOL PODCAST (41:23.682)
Yeah.

Ryan Roberts (41:37.296)
And so you don't do the analysis so that you are on a firm foundation to now build the scaling machine that you think you want to. And so it's incredibly important for founders and something past the startup phase to understand how you can apply these principles. And still, even though it is highly complex, but you want it delivered in a way that is just simple and actionable. And so that leadership can go.

Okay, that makes sense.

THE HUMAN PROTOCOL PODCAST (42:07.854)
don't understand what is it that I need to do and where do I need to start and the reason why, right? I think which is important for these early stage founders when they're looking at everybody wants to scale at the end of the day, but is the how you do it that is gonna make it or break it. And we see it all the time, right, Ryan?

Ryan Roberts (42:29.522)
It's not only that we see it all the time, it's just this misconception. I was having a discussion in the last couple of weeks with somebody and it made me think of, the Princess Bride and the line where he says, I don't think that word means what you think it means. And that's how scaling is. like, if you think scaling your business is making more money,

THE HUMAN PROTOCOL PODCAST (42:50.198)
Mm-hmm.

Ryan Roberts (42:58.182)
you're missing the rest of the equation, right? The scaling means that I can bring in top line revenue without a linear correlation to the effort. And so to do that, you have to have this type of assessment done so that you can put systems in place so now that you have a gap that can grow of I can repeat these processes instead of just the I do more, make more, or even worse.

THE HUMAN PROTOCOL PODCAST (43:01.23)
Hmm.

THE HUMAN PROTOCOL PODCAST (43:18.839)
Right.

Ryan Roberts (43:26.222)
And generally the case, is completely blind to all the things that actually need to happen inside your business before it's too late.

THE HUMAN PROTOCOL PODCAST (43:36.59)
It's about people and processes and how to build efficient processes in the middle, right? And that's the magic about building successful businesses. And it's like you say, it's critical for you to understand very early, what is it that needs to happen and how do I need to structure myself and even architecture behind the scenes to sustain the growth.

of what may be heading your way before it's too late.

Ryan Roberts (44:11.058)
That's right. And that is usually the issue with founders and early stage leaders is obviously if they've shown that what they offer is viable, right? It has value. Then often they're highly tied to how do we make that more valuable or how do we find more people? And it misses most of the other components that go into making sure. Like, do you have the financial controls even in place?

THE HUMAN PROTOCOL PODCAST (44:29.411)
Right.

Ryan Roberts (44:40.516)
Are you crystal clear about the tripwires that you've set up that you're going to make decisions agnostic to what's going on over here, right? Like, just want to, and that often is not the case in businesses. And obviously the things we're talking about with risk and operations using AI, are you even aware of all of these factors that will pile up on you silently?

THE HUMAN PROTOCOL PODCAST (45:04.536)
Right.

Ryan Roberts (45:07.09)
until it causes a conflict that such where your margins aren't nearly what they need to be to continue to grow. And as we all know, that is key to success and longevity in business is growth. Stagnation kills a company too. So you have to make sure you have systems that are ready to continue growth. And it's just overlooked because rightly so the founder has an idea, has a service, it's great and they're passionate about it.

THE HUMAN PROTOCOL PODCAST (45:35.822)
Absolutely. Yeah. And we, we, we often see it. People are very passionate, right? And that's beautiful to see, but there is a lot more to it for you to be successful.

Ryan Roberts (45:36.754)
What? Yes!

Ryan Roberts (45:44.27)
There's a lot more to it. Yeah. Let the engineer go do that piece, right? You know, I spent a lot of years flying jets. And I can tell you between engineers and operators, there was always this little bit of a, of a riff of, you know, where, where a pilot be like, why in the world this engineered? Yeah. Yeah. Like, why would the engineer do something like that? And the engineers like pilots don't understand, you know,

THE HUMAN PROTOCOL PODCAST (46:01.056)
of fighting each other.

Ryan Roberts (46:08.754)
But it's true even with operators of business and those that understand what it's going to take to avoid the vulnerabilities so that you have longevity. It's two different skill sets. And founders would be very wise to make sure they find people and systems to put in place so that they can continue to be passionate. But it's often not a sprint, it's a marathon. And so let's make sure we put the systems in place when we're ready.

THE HUMAN PROTOCOL PODCAST (46:14.798)
Mm-hmm.

THE HUMAN PROTOCOL PODCAST (46:22.008)
to be.

THE HUMAN PROTOCOL PODCAST (46:37.942)
I love that. And you can use that phrase for everything. It's really often, most of time, it's really not a sprint. It's really a marathon, and you need to prepare for that. And that's where designing systems for failure and back to the conversation is really what you can prepare yourself, kind of to anticipate the worst, and that way you're gonna be prepared for the worst in a way.

And when you design things and when you think about it, there's so much beauty and logic on thinking what may go wrong instead of designing of what I envision for a process or a system to work for. It's such a beautiful thing to think about that, because you're going to be definitely designing more resilient systems and processes in a way.

Ryan Roberts (47:31.76)
And this is what I would call the foot stomp. Foot stomp, this is an incredibly important point, which is that you brought up thinking about the contingencies or the things that could go wrong. Look, there's been studies done, and founders often are very passionate. They've got something that works, and they have this vision. But they don't spend enough time thinking about all the things that could go wrong. And studies have been done to show

Like we take two groups of people, right? We take one group that has a clear vision. They set the vision, you know, and we across groups, you could do things like vision boards, all the things that you might think of about being very clear. That is an incredibly critical piece in neuroscience of making sure that now you can picture what it looks like so that your brain has less friction and it understands how it, you know,

THE HUMAN PROTOCOL PODCAST (48:28.14)
Hmm.

Ryan Roberts (48:28.646)
can see it, and then if it's really bold vision, that's healthy too because a piece of your brain called the reticular activation system now gets involved where now it's constantly aware. All important. But the data shows in these studies that they can take a second group and do the same thing, but do one additional thing, and that is to track and write down and correlate every possible thing that could go wrong. And then consider that.

THE HUMAN PROTOCOL PODCAST (48:53.975)
Uh-huh.

Ryan Roberts (48:56.024)
And the increase in the achievement of the goals went up 60 % by the second group. And this is why this is important. This is why systems engineering in a business works. Six Sigma did its thing to try to understand some engineering and some statistical analysis around variance. But systems engineering and the best out there will understand that you have to plan for

THE HUMAN PROTOCOL PODCAST (49:03.086)
Wow.

Ryan Roberts (49:24.496)
what could go wrong and most often it becomes critical if you don't know. And that's why, that's right. And it is so overlooked. Look, in military strategy, it's a step we have to be very clear about contingency planning and it's best practice across all kinds of systems, not just in your strategy, but all the way through all your risk evaluation across all the domains in your business.

THE HUMAN PROTOCOL PODCAST (49:28.632)
Yeah. Contingency plan and build your contingency plan around that.

THE HUMAN PROTOCOL PODCAST (49:43.086)
Mm-hmm.

Ryan Roberts (49:52.978)
of what could go wrong and then ask it again and go a few layers deep, What if, you know, now what and then what and then what so that you are trying to uncover all of the things that are invisible that make it critical.

THE HUMAN PROTOCOL PODCAST (49:57.526)
Yeah. And then you keep pushing and keep trying more. Right.

THE HUMAN PROTOCOL PODCAST (50:09.452)
And at the end of the day is preparing you, us to survive, right? Because you're only going to be able to survive as much as you are prepared for the worst. And this is what we are talking here today. And especially in this conversation for business leaders, if you really want to survive, you have to think about and prepare for the worst because that's what's going to get that your business survives really hard time and challenging times ahead.

And this is all part of these exercises can prepare you for tough times ahead.

Ryan Roberts (50:48.518)
Yeah, and we touched on this earlier, but that's exactly right, is you have to prepare ahead of time, right? And it goes back to a principle in systems engineering that your failure falls to the level of which things are observable. again, whether it's AI in this case and how it affects businesses and why ethical decisions

will actually affect, literally cause risk to rise and becomes potentially critical in a business because you can't just ignore it. can't, not only can you not just ignore it, you can't also have a very thorough plan. You can't just surface level say, this is, I put this person in charge and they're gonna go do this thing. That's not enough. You have to at least run an analysis of

where things could be critical that you're unaware of because that is what is going to kill your business. And it doesn't matter which mechanism it is, including AI.

THE HUMAN PROTOCOL PODCAST (51:57.144)
Beautifully said, Ryan, beautifully said. So thinking on this discussion today about ethics and architecture, right, and how they are kind of correlated and one with the other, how do you see the blueprint for future leadership? And what do you want to see leaders and business leaders do more in the future?

Ryan Roberts (52:22.812)
Well, I would say whether it's this particular mechanism, is AI, that causes some consideration for risk across an organization and potentially, again, has ethical implications for communities and beyond organizations, that business leaders need to consider the whole business as a system. And that's how they should be analyzing their business.

not just about what feels good and setting big goals that are based strictly on revenue, but how about setting some goals about how clear you are about the systems that run your business? Look, the best companies out there that are wildly successful have incredibly robust systems in place to make sure that the systems that

that it can support the massive amount of revenue that sometimes they can generate. And sometimes, in some cases, especially with rising technologies like AI, are rapidly growing. But you have to reinvest those funds into systems that can detect where the risk might be coming from. And it's critical, especially with when we mention, again, like young companies, startups, founder-led companies, because

The criticality raises too by actually one more factor actually that is a multiplier and that's what we call the exposure factor. How exposed are you to any particular failure, right? What stage is the business and what's the current velocity and what systems are there for resilience? And so you might imagine, you might have a hiccup with something like making a poor decision about AI or any other mechanism if you're a mature company and have

lots of cash reserves. If you're a young company, you can't do that, right? And so I just think that like what I would what I think business owners need to do and leaders and founders is to make sure that they surround themselves and and take advantage of systems like we offer, which is to just highlight what you likely aren't looking at so that you set yourself out up for

THE HUMAN PROTOCOL PODCAST (54:13.518)
Mm-hmm.

Ryan Roberts (54:37.776)
Longevity in business, which again is going to lead to the success. So overall, that's what I would say. I would say that passion is great about your service or your product, but it will not carry you. And what will carry you is relying on a system that just continually shows you over and over again what the next vulnerability is.

THE HUMAN PROTOCOL PODCAST (54:49.208)
Mm-hmm.

Right.

THE HUMAN PROTOCOL PODCAST (55:03.488)
Awesome. What would you tell government officials, what would you like the government to focus on in all this conversation with AI? What are the areas that you would like to see more action from policymakers, for example?

Ryan Roberts (55:22.51)
this one's tough, right? Because, you know, government serves a purpose, right? And it should be for the people so that we can enjoy more liberties. And this is where AI gets really sticky because, right, and again, I'll circle back to what we know is going on in Europe right now and some discussions about

facial recognition technology and collecting that data. Well, you would call that information asymmetry in business, right? If somebody else or another company or even inside your organization has more information, that's leverage. And so if you have companies that are allowed to collect data without consent, that's information asymmetry.

THE HUMAN PROTOCOL PODCAST (55:58.882)
Mm-hmm.

Ryan Roberts (56:19.396)
which in another word, we would call control. So I think that it is a consideration. Look, I'm not politically affiliated with any particular parties. I do feel strongly that government was created for our purpose and for the people and not the other way around. And I think that liberties should be protected at all costs. But that gets very sticky when we consider

that there is a responsibility that the government should make sure to enforce when it comes to information asymmetry that organizations are.

THE HUMAN PROTOCOL PODCAST (57:00.406)
and how to level the playing field in a way, which we know that in this whole new future, those that are using AI will have, you know, leg ahead.

Ryan Roberts (57:10.225)
Right.

Right, and so if some organizations can afford to collect more data because of the AI, then what's the answer? Like how do you protect everybody's information or make it available to everybody? Like you tell me. You tell me where the answer is. Because if one company is allowed to collect it, then I guess we should say everybody should be allowed to have the data. So what, are we going to open source all of the data that's collected with

THE HUMAN PROTOCOL PODCAST (57:20.802)
Yeah. Right.

THE HUMAN PROTOCOL PODCAST (57:41.134)
there.

Ryan Roberts (57:41.139)
biometric data that's now being collected on a daily basis in parts of the world.

THE HUMAN PROTOCOL PODCAST (57:43.872)
boy, that's a conversation by its own in a way. That, my goodness. We don't really have the answers right now. I think we're going to have to see where this goes and how this whole thing evolves and see what are the global standards that, you know, because

Europe is doing something, the US is doing something else, Asia is doing something else. I'm curious to see where all this is to get to. Hopefully in a place that we have some safeguards and global standards around data privacy and things of that nature. We shall see, right? We shall see. Ryan, one last question, okay, before the end of our conversation.

which is what would be your piece of advice to a young Ryan, you looking at him today and in this new future with AI? What would you tell a young Ryan today?

Ryan Roberts (58:49.928)
wow, no question that I would tell young Ryan. If you're saying given the current environment, if I was a young Ryan in today's environment, I would tell any I would tell my young self as early as I could tell my young self. You know, potentially at least down into my teen years to start becoming incredibly educated on AI, right? Because we know that.

THE HUMAN PROTOCOL PODCAST (59:01.112)
Right.

Ryan Roberts (59:19.666)
The rise of the term AI seems to me to be linear in the last five years or so with the rise of large language models like a chat GPT. Because it seems like that's what people think you're talking about. When really there's so many more domains of OAI, right? So, and usually the folks who are in tech or certainly developers, they understand.

THE HUMAN PROTOCOL PODCAST (59:36.93)
layers.

Ryan Roberts (59:45.508)
some, you know, the difference between that or what it might take to build an agent or, you know, various pieces that we might consider how AI is being put in, you know, into work. And so I think a young, I would tell my young self to get very smart on this. And it's even different, right, than some of the technology we've seen over the last few decades, right? The internet changed things, you know, dramatically.

the way we then later on shopped online. I there's just been a few different shifts and then mobile devices, which by the way is about to go through another transformation. yeah, yeah, yeah. Apple seems to be the only one hanging on to the mobile device while other companies are investing into wearables and other virtual environments. But I just think a young person should get

THE HUMAN PROTOCOL PODCAST (01:00:22.762)
So we're seeing wearable devices now as the future.

Ryan Roberts (01:00:41.062)
should spend a lot of time in this space and be very, very, very smart about it because that is going to be the future, at least foreseeable. And I think everybody should at least agree on that. And that if I would tell young Ryan that, then I would tell any person that, that spend some time and start understanding.

THE HUMAN PROTOCOL PODCAST (01:01:00.878)
Yeah, absolutely. That's the very sound advice and that's for really everybody, not only a young Ryan, but a young Michael and everyone that is really listening to this. Ryan, thank you very much for joining us today. Really, really appreciate you being on the show with us. Okay. Thank you. Appreciate it.

Ryan Roberts (01:01:18.736)
Yeah, Michael, was great. Really fantastic. I appreciate it.

THE HUMAN PROTOCOL PODCAST (01:01:22.284)
Yep. And thank you for reminding us that ethics isn't really a department, it is design. And you have shown us throughout this conversation today that accountability is really an act of architecture. One decision, one process, one human at a time. And for you guys watching us, today's conversation showing you how you think about leadership, share this episode with a friend, a colleague, a founder, or even a policymaker who's shaping our digital future.

Follow the human protocol and join us every week as we explore what does it mean to stay human while building the future with AI. Thank you all and see you next time. Take care.