Pondering AI

Drew Burdick designs AI systems to multiply human capacity, prioritizes great experiences, and values the serendipitous magic of human connection and collaboration.      

Kimberly and Drew discuss building with badass teams; curiosity and innovation; building momentum with AI; human relationships and rapport; proprietary knowledge and expertise; long-term thinking; AI agents as teammates; pricing in human experiences; designing for humans vs. bots; regulation and accountability; societal guardrails; the mid-market squeeze; actions companies should take now; investing in people; and keeping community front and center.

Drew Burdick is the founder of StealthX and the CLT Startup House. Drew parlays his deep background in design and solution development to help companies deliver exceptional experiences with AI.

Additional Resources:
 
A transcript of this episode is here.   

Creators and Guests

Host
Kimberly Nevala
Strategic advisor at SAS
Guest
Drew Burdick
Founder - StealthX

What is Pondering AI?

How is the use of artificial intelligence (AI) shaping our human experience?

Kimberly Nevala ponders the reality of AI with a diverse group of innovators, advocates and data scientists. Ethics and uncertainty. Automation and art. Work, politics and culture. In real life and online. Contemplate AI’s impact, for better and worse.

All presentations represent the opinions of the presenter and do not represent the position or the opinion of SAS.

KIMBERLY NEVALA: Welcome to Pondering AI. I'm your host, Kimberly Nevala. In this episode, we're pondering human-centric design with the founder and CEO of StealthX, Drew Burdick. Welcome to the show Drew.

DREW BURDICK: Yeah, Thanks for having me. Glad to be here.

KIMBERLY NEVALA: Absolutely. Now, I know that you are a designer to your core, but your early education was actually in mass media, communications and poli sci. So I'm wondering how that learning has influenced your work over the years, if at all?

DREW BURDICK: Yeah, great question. So it's funny, when I was in college, I didn't really know what I was going to do with my life. But I was very fascinated with human psychology, sociology, things like that. And I remember distinctly, I think it was my sophomore year in college, I took a ethnographic research class. And little did I know that years later, I would be using that all the time.

A lot of the work that we do is trying to understand how people interact with each other, how people communicate, how people think, and design things for them that make their lives better. And so it's kind of funny, it was a bit of a winding road. But a lot of the liberal arts stuff that I learned through college is very, very applicable now and in ways I didn't realize, for sure.

KIMBERLY NEVALA: That's awesome. Now anyone that looks at your work will see, that you are simultaneously giddy, and I mean that in all the best ways, about the affordances AI provides today. And equally, if not more so, giddy about building, and I quote, with badass people. So really delivering great experiences for humans. And I'm wondering how you situate, or strike a balance, in thinking about when and where to use AI or not to use AI?

Particularly in the face of this very pervasive, pervasive narrative of AI as an everything machine. Which when taken to its ultimate extreme, really sort of puts humans in the place of being-- situates humans as being surplus to requirements. So how do you think about This

DREW BURDICK: Yeah, so I'll start with the badass team. I'm really having a blast working with people who are also kind of builders or are curious, leading with interest, and wondering. And so a lot of us on the team, we're building every day. We're trying to solve new problems in new and interesting ways. And as you might imagine, one of the most valuable human skills is the creativity and the innovation.

There's this really great story about Bell Labs. I think it was sometime in the '60s or '70s, where this one particular team across the board they were consistently delivering the most patents out of like any company, any team. And I think Harvard, or one of the universities, who did a research study on this team, and they discovered that this team was innovating primarily because they were at lunches with these other teams. And they were just talking about what might be, what could be. And what they were doing in one part of the business, they were cross-pollinating on the other side of the business. And so I believe that humans, that's going to be the most important thing is the creativity, the collaboration. I call it the serendipitous collisions when you're working with people that AI can't really replicate. But on the other side, I think with AI, there's a lot of really cool opportunities to use it to do the things that, honestly, a lot of folks don't enjoy.

It's the triaging your email inbox and helping you organize information and collate stuff into a usable format. Those are things that we're doing all day. I use this phrase of AI is going to make everything fast and easy. And the thing that's going to stand out more than anything is differentiated brands and experiences. And I think also just the trust. Trust is something that people don't know who to trust. They don't know is this real or is this not.

And so I think it's going to be more and more important that we all invest in cultivating trust and relationships with other people. Because ultimately, at the end of the day, we're still serving people. People are the ones buying things. People are the ones that are living, right, that are actually experiencing things. So yeah, anyway, that's kind of our viewpoint.

KIMBERLY NEVALA: So in this ecosystem where you have said, and others have said, that AI can scale up, often mediocre, content and apps overnight, how do we actually, as organizations, as businesses, ensure that we're focusing on the right things and not just the same things that everyone else is doing? And in that scenario, how do trust and perception show up at all?

DREW BURDICK: It's interesting. We spend a lot of time on this thinking about what's the right problem to solve with AI. We have a lot of our clients, and folks that we work with, they're asking that question right now. The last six to 12 months have just been a pretty wild time to be alive. And folks are continually coming to us, and they're like, we know we need to do AI stuff. We don't really know where to start. We don't know what the right problems are. What should we do?

And there's some, I would say, standard things that everybody, every company, should invest in. Which is if you're not using Copilot, ChatGPT, Gemini, Claude, you should probably be using something like that and teaching folks how to use it, how to think about it. You should probably use a note taker to help with capturing notes and things like that.

But then when it comes to these custom solutions, like agents, people talk a lot about. We have this opinion that to build momentum, to get the organization to really rally around what's possible, the really big trend we're seeing is focusing on highly visible problem that's very acute and is not dependent on internal data and systems. Because a lot of the pain points that enterprises run into is they're like, well, our data is locked up in this warehouse and it's not accessible. And there's security concerns. And legal concerns.

And so we've kind of really started saying like, let's find a highly acute, highly visible problem to help build momentum around the value of AI to serving the organization. And then you said trust. You talked about that. I think right now because everything is- it's moving so fast. It's like the Wild West. People trust people. And so the folks who are out, and this is what I believe in. I want to be in it every day, learning, growing, and improving our knowledge so that we can help other folks who are scared.

They don't know what to do. And so they trust that other human beings are doing the due diligence to make sure that it's a good tool to use. So anyway, that's how we think about it. The other element is the human experience piece. There's a lot of discussion around how do we do something that's a high impact, low effort use case for AI. And that's a great way to think about it.

But there's also when it's a customer facing, or even employee facing, is it going to make their life easier? Is it going to make their experience richer? Are they going to be able to focus on high value activities? And we fundamentally believe that this should enable people, but not be one of those, oh, we don't need these people anymore. So there's also a lot of work that's being done around reskilling, upskilling people to help make sure that they can leverage the tools to deliver more value and also continue to have a job.

KIMBERLY NEVALA: And are there examples that you particularly like, whether that's in your work with your clients or that you've seen out in the wild, of folks taking this approach in not only leveraging the AI to good effect, or an AI system to good effect, but also navigating that turn too with the humans? I don't want to say in the loop but navigating to rethinking or upskilling the humans that are involved.

DREW BURDICK: So we just recently did a project with an organization. I'll just say they're in the financial services world. And one of the tasks that a lot of their investment research team does is manually go look at white papers and websites and newsletters and LinkedIn to try and find new investment opportunities through fund managers. And people were spending hours, and hours, and hours doing this stuff. And it's a fairly small team that does this.

And they came to us and were like, hey, this is a very visible thing because this is like how we get new deals into the pipeline. It's also acute because we have limited people to solve this problem. We think there's a way that AI could help us. We don't really know what to do. And we're like, that's a great. That's perfect. We can do that. So we, with a very short amount of time, built an agent, it was actually a series of agents, that were solving different pieces of this workflow with an interface that allowed them to see AI is making recommendations like, here's a potential fund manager that might be opening a new fund.

It was scoring them based on the ways that they thought about it. And also providing them with interesting data. It was enriching with data that said, here's why, basically, AI felt like it was a good fund. And they've loved it. And it's made their life tremendously easier. And it's allowed them to focus more on their clients.

They're able to spend more time and energy thinking about their clients, and their relationships with these fund managers, versus combing through LinkedIn every day or looking through some of these websites. And so that's just one example. I think there's a lot of things like that where people are spending so much energy on finding the information, analyzing the information, doing something with the information, that they're not able to do the things that are truly most important and uniquely human. Which is the going and having conversations with people and building the relationships and the rapport. So that's just one example.

KIMBERLY NEVALA: Now you can imagine in that scenario, that the information that the AI system is surfacing could look a lot like the information that a similar AI system would be surfacing to a competitor. And whether it's you're using this to help inform strategy, or to perform research, we know these systems always collapse or move towards the mean. We also know that there's a tendency to over trust or over rely, sometimes, on those.

So even in a simple situation like that where there's clear value, how do you address those types of factors to ensure that AI is an input to the decision, making input to those decisions, and not the decision maker, If at all? Maybe it is a decision maker, I don't know.

DREW BURDICK: No, no, it's a great question. So throwing back to what I mentioned a little bit ago about ethnographic research. So we spend quite a bit of time sitting with the team and helping understand how they thought about what a good fund manager looked like. What were the characteristics? What were the patterns? And then we used that to train the agents on that. At the end of the day, though, they still have years of experience and expertise that they have to apply.

But it's rather than looking at the entire ocean of potential people, they're using their proprietary knowledge, their experience and expertise that we've distilled into this agent. And it's making the nets, where they're like, all right, I only have to review these 30 people versus these 3,000 people. And so in my opinion, what a lot of companies should be investing in is there's really important knowledge that people have in their minds.

How do you distill that into something that can help be a multiplier for people, and their talents, and their capabilities? And then they can spend more energy on the stuff that really matters versus a lot of time on just trying to get to those final 30 that they're really evaluating.

KIMBERLY NEVALA: And if we think about that judgment and experience, then you could then draw the conclusion that says these AI systems favor, yes, the prepared but more the experienced. And there's a lot of discussion right now, and I want to pull back to that thread you said about upskilling and reskilling, for the sort of broken career ladder. Where, how, do you come up and get that expertise, learn that judgment?

Well, it's probably from doing some of the grunt work. Not suggesting here, by the way, that's the only way to do it. But it has to be developed at some point. And you have said in the past that, perhaps, expertise in industry, or subject matter expertise, will be the new frontier. Do you still believe that is true? And how should we be thinking about going after that?

DREW BURDICK: It's interesting. So since the last time we talked, I was on a panel with a few folks. And we had a Q&A session. And one of the guests, or the attendees, said, basically, hey, I have an 18 year old. What should I do? Where should I point them with the way the world is? And I told him. I was like, honestly, I don't know. The world is changing so dramatically.

And as a parent myself, it's a tough market right now because it is difficult. Those first couple of years of experience, you're really just immersing yourself in some industry, or skill set, and learning. Some of the things I've recently written about, and been thinking about, is this idea of apprenticeships. How to provide opportunities for folks to be in the weeds with someone.

And if you recall, hundreds of years ago, apprenticeships were actually often not paid. And they were this opportunity for you just to trade your time in order to get experience that you can then parlay into a career. And I think that we may see some more of that because a lot of organizations are having to make a hard decision on, well, can we really afford this 22 year old that's out of college, has no experience and expertise? Or can we just have AI do it for us?

And I think a lot of folks are picking AI. And I think we do need humans. I mean, the pool of talent is going to get dramatically smaller as time goes on because we won't have those three to five years of experience folks, because we're not nurturing them right now. So I think organizations that are thinking in the long term, thinking beyond just the next short-term gains of, oh, we just use AI to solve this problem.

They're thinking about, well, what happens when all of our mid-career, later career folks start to move on? we're going to have a gap here. I think they're going to probably need to invest some more intentional dollars in apprenticeship kind of programs, internship type programs. Because I do think we're going to be in a world of hurt in a few years when all of our recent college graduates just haven't been able to get the experience that they need to be able to provide value to the economy and to organizations.

KIMBERLY NEVALA: Yeah, at the end of the day, we still need people.

And I think there's this question, and you alluded to it very early, which is where does innovation come from? And yes, I think AI systems in some cases, because of the spread of the data that they're pulling, can pull threads together that you might not otherwise do. But a lot of times that innovation comes from humans addressing a constraint. There's some constraint.

Or pulling, as you said, those serendipitous ideas together from places that you may not get from the AI because it wouldn't even think to connect them. So it will be interesting to see how we both nurture and reward that type of interaction moving forward. I did notice you've been posting recently and you had made a statement. And I think the language is sometimes important. And it can help us and hurt us in different ways.

And you talk about having AI teammates, or AI agents, or AI employees. I don't remember the exact word you were, but it was on there, employee, teammate. And I think this terminology in some ways is a strange conceit. We could just call them AI programs. We could say I this is how I do my role, and I use these AI programs to help really expand and engage on that.

And the language is interesting because I think if it is applied uncritically, it risks conflating people and programs, people and AI systems. and perhaps set some inappropriate expectations, or assumptions, about both the capabilities and value brought to the table by both. So how do you think about that? And is the language just what the language is? Or why is that terminology something that you lean into?

DREW BURDICK: It's a really good point. I find that when I'm talking to folks who are fairly low maturity in AI, understanding what's possible, the word agent loses its meaning. People are like, what is it? It's just like it gets thrown around a lot and people are like--

KIMBERLY NEVALA: AI might have lost its meaning as well, to be honest.

DREW BURDICK: It's sort of like, OK, AI agent. What are we really talking about here? And so often when I'm speaking with business leaders who are trying to wrap their head around it, I'll use this. Think of it as a teammate. Let's just set the agent word to the side, and let's just talk about what does a teammate mean? And a teammate is-- if we set, obviously, human teammate.

But when we talk about AI teammate, it is a system that is able to interpret information, take action on your behalf. You can delegate responsibilities to it. It often has a job description, or in a human case, you have job description. In an AI case, now, actually, it's similar. There's a bit of a job description. And I'll just give you a specific example. So we use Claude Code quite often. That's one of our core tools that we use every day.

And I have a C-suite, I have an actual COO and then I have a COO agent. And I'm able to go to Claude Code and say, hey, COO, blah, blah, blah, blah. And I can issue some sort of need, or goal, or problem. And it deploys teams of agents that are teammates that will actually go and tackle that. And it creates a plan. I have red team, yellow team, green team, blue team. And they have different purposes. And it is pretty insane what these AI teammates are able to accomplish.

So I do think I think it's a really important point that you're making. I will say like I probably was a little careless saying the word teammate, because it does devalue, potentially, human teammates. But I think a lot of people scrolling on LinkedIn, or whatever, and they're seeing agent, agent, agent. And it's like, what even is this? And what's a skill, and a plugin, an agent, and a sub agent, and a LLM?

And it's all these terms that people are just like, I don't know what this means. And you're like, oh. Just think of it as a digital teammate who can take actions, has a job description. It has the ability to do certain things. That's what this is. And that helps people orient to what it is and also what's possible. Because a lot of people still struggle with what can it really do.

KIMBERLY NEVALA: Yeah, it is hairy. Back in the day, I was guilty of telling folks, and this was back in the early days of data governance saying-- because we got brought in a lot of times to say, we've done this once. We took a run at this wheel and we failed. And so we need you to come back and do data governance, and we need you to call it something entirely different.

And I used to say, I don't really care what you call it, as long as you're clear about the definition. And then you understand how other people are going to perceive that when you're talking about your XYZ program, just not calling it data governance. Or you're using this term that other people use in a different way. And I suppose the thing that I worry about most there is how folks, employees, and organizations view that.

And you mentioned people are sometimes afraid of it. And I think sometimes I'm worried that language, and the use of that language, sets people up into a competition with systems. And I think you and I both know pretty well when folks like yourselves use it-- I'm going to push back, because I know you are open to the conversation and to take it. Again, you're not using it uncritically. But in other cases, I think that it sometimes can actually add to fear and hesitation.

And/or it plays into this area where we get reliant on it because we expect a level of discretion and judgment that we would get from a human teammate. Again, this probably all goes back, though, to good programmatic and system design, I would imagine.

DREW BURDICK: I don't know if you saw this really interesting article that went viral about two weeks ago. It's called "Something Big is Happening." Very interesting read. Matt Shumer, I think, is who wrote it. It's come up a lot. But what's really fascinating about it is since the most recent model releases from OpenAI and Anthropic, the capability has gotten incredibly incredible. And in addition, the interesting thing now is that AI is training AI.

So I think there's going to be this exponential curve. And so truly a year or so ago, I was communicating, hey, these are the things that AI can't do. This is what humans can do. And increasingly, AI is more and more and more capable of doing things that a year ago, two years ago, were thought impossible. And so I think there is a bit of, I don't know, hard reality that a lot of us are going to have to face of, OK, AI is actually, truly in a lot of ways, operating as a human teammate.

And I think the reskill, upskill piece is like, how do we equip people so that they can elevate? Because I don't think everybody's ready. I don't think everybody's ready for how quickly this is accelerating, and how much better it's going to get in the near future. The model providers were pretty smart. They started first with making it good at coding. And now that it's great at coding, it can train itself. Which is really weird if you think about it. So anyway.

KIMBERLY NEVALA: Yeah, we'll see where that all goes for sure. And interestingly enough, I think there's probably an argument to be made that we need to be even more careful with the language when that happens. We could do that. But now if we shift our focus a little bit and look at the consumer, or the customer, experience, you have said that humans are becoming the luxury tier of experience. What did you mean by that?

DREW BURDICK: Simply put, if everything starts being done by AI, if a human does it, that's a premium service. That you're having now a person who is going to be going and working on whatever it is. So I think as everything gets faster and easier, as all this stuff kind of gets commoditized, as AI starts to consume a lot of these elements, there will be this opportunity, this interesting counterpoint, counter side, where folks are going to be like, I want more human made things.

I want to go somewhere and talk to real people. And I want to be eating food that humans have made, not robots. I want to look at art that humans have created, and not something that was generated in 60 seconds. So I think it's just the natural, in my opinion, progression of the way that we're all going. If content is easy to produce, it's going to be more valuable that someone sat down and wrote this out by hand. That's a very valuable artifact, right. That's a very valuable piece of work.

Or if someone really-- like this. We're having a conversation. This is not AI generated. This isn't produced by some kind of artificial whatever. Like we're just talking and we're just having a conversation. And I think these are the kinds of things that are going to become more and more valuable because it's I don't know what to trust, but these are real humans. And they really care about the craft. And they care about the work. And they are putting their time and energy into this thing.

And so I do think that it will become the luxury tier, as you said.

KIMBERLY NEVALA: Interesting. And how do we make sure, particularly those of us who are working in the field-- and I think we have to admit to some extent that we are in a privileged, sort of, position, and are immersed in ways that the vast majority of humans on the planet are not, many of whom don't have basic access to internet and things like that.

But there's also been this thread that I found interesting about do we need to design systems for bots? Or are we deciding are we designing those for humans? And I'm wondering how you think about this sort of design for the bots versus designing for a human experience. And there's these elements where-- I'm going to forget who this was, but a very prominent academic.

And he said, we need to start, for instance, formatting articles so that generative AI can consume them. Or actually, I think it was on your podcast, someone was talking about CAD and saying, well, now we have this amazing way to interface with the program with language. And so all of these people who like to draw and use their hands need to learn how to instruct the machine with their voice.

And my thought was, as someone who cannot draw a stick figure that credibly looks like a stick figure, and I wish I was joking about that, I was like, well, but if someone wants to draw, then let them draw. If part of their process, whether it's the tactile, they're moving it on the screen with their mouse, or they're putting a model together, because that's how they think about it, I don't want to see us start to limit how we think about, and how people can engage, with materials.

So I can only get it through a generative AI chat bot, particularly if I'm only constrained to the providers that are here today. Or everything that I want to do, I have to do with my voice, because maybe that's not the way everyone wants to do that. So how do we make sure that we're balancing those forces, if you will?

DREW BURDICK: It's a great question. I mean, to be honest, like, I think a lot of us are still wrestling with it, right?

KIMBERLY NEVALA: Yeah.

DREW BURDICK: My thought going back to the beginning of your question of this designing experiences--

KIMBERLY NEVALA: My rambling?

DREW BURDICK: No, no, no. Designing experiences for humans and designing experiences for AI. I think there's this really interesting thing that's happening right now where if you look at-- most people are familiar with the concept of search engine optimization, SEO. And over the last two years, we now have AEO, basically artificial intelligence optimization. As well as GEO, generative engine intelligence optimization.

And people are now designing and optimizing experiences for Claude, and ChatGPT, and Gemini to find content and serve it up to a customer or user. And it's working. It's interesting to watch. You can optimize things for AI because AI looks for certain types of patterns and certain types of ways. And so there's a lot of interesting conversation around this idea of agentic experience.

How do you create an experience? Well, there's two sides of it. There's the how do you create an experience for humans where it's powered by agents, and making their lives better and easier? But then also, how do you create an experience-- I don't know. It's probably not the right word. But create a tool that makes it easier for AI to access certain information and do things on behalf of the person.

I had somebody on my podcast really recently, and he's a CEO for a text messaging platform company. And we had a really good discussion around if you're familiar with RCS, it's like a rich text messaging that's recently come out. And so we were discussing that. And one of the things that I kind of posited, or put out there, was I think we're quickly approaching a world where all of us are going to have our own personal assistants that are going out and doing things on our behalf.

And it'll often be, I think, assistant to assistant. So if I'm trying to schedule time to meet with you, or someone else, my assistant's going and coordinating with yours, and coordinating, and scheduling it on the calendar. And you and I aren't necessarily having the conversation. It's doing it on our behalf. So how do you make that system easier so that it's not getting stuck on some permission issue or something like that?

And so those are some of the problems that we're actively trying to solve right now. We're building a product at the moment, and it's this idea of creating a platform that's recognition anywhere work happens. So this particular company, they're a B2B SAS company, and they do employee recognition and rewards. And one of the challenges is, you think about how employees work, they're just recognizing each other in Slack, and Teams, or in person, or in meetings where you're like, hey, go to this thing and put a hey, Kimberly, you a great job on that podcast. Awesome job.

People don't do that. It's just you're adding friction to the process. So we've been thinking a lot about how do we bring this to them? How do we bring it to humans and also proactively help identify, hey, you said something really nice about Kimberly in Slack. Do you want to nominate her for this award? Or do you want to give her a gift card or whatever?

And so it's kind of interesting, because it's we're designing for both. How does the agent interact with these other tools so that human experience is better? And then also like how do we continue to make sure that the experience for humans is as good as it possibly can be, accounting for the fact that AI is going to be doing a lot of things on their behalf? So it's interesting. We're trying to solve those things now and it's really tough.

I think everybody's kind of still in the middle of what is the right way? We don't know.

KIMBERLY NEVALA: And I recently had Theodora Lau on the pod. And she talks about AI as the OS moving forward. Or as the interface. Which I know you have also been positing and hypothesizing, even working towards, for a while. Where a lot of the work that we do, and the interactions, are inside AI environment. And in some ways, I think-- not different than I think the applications that you just talked about, where I start to think about how do we find a product? How do we find a service?

And she's in financial services. She's saying there's been some research recently. And what they found is really what is getting surfaced right now within these chat bots is maybe nine or 10 companies. And they're only the big ones, right. They're not the new innovative fintechs. They're not the ones in emerging markets. They're not any of these smaller companies.

And so where we are right now because-- I think, conceptually this makes sense. What I'm worried about is right now we've developed sort of a Google redux that has significant lock in for these companies that have been out the door first. And it's sort of one entry point to rule them all, where they have a lot of control over who gets seen when and mediating what options are shown.

How do we then think about that particular issue? Because given the current-- outside of folks having to pay to play, which I'm sure will come -- but again, still favors folks that are really, really large.

The architecture of these systems still today favors the folks who already have large web presence. And we could say, well, people can start to use bots. And you can create lots of content. And you can just flood the zone. But truthfully, there's always someone who can flood the zone, right? Bots, there's no end in sight to this. This doesn't seem like a reasonable way to approach AEO or GEO. So how do companies think about this? And is this something that we should be concerned about at all?

DREW BURDICK: I think we should. Yeah, we definitely be concerned. You all probably saw OpenAI, ChatGPT, just recently rolled out ads. And it's funny, because the Super Bowl commercial, Anthropic is kind of poking fun at it. But I think it's a sign of things to come. I think what you're going to see is the people who can afford to buy attention in these channels are going to start to own the attention.

And the only thing that I think that is really going to solve it, maybe, or address it in some part is what we had talked about when you and I did a fireside chat at UNC Charlotte a few months ago. I think the regulators are going to have to take a more intentional look at this. They're going to have to hold these big model providers accountable. Just think about Google over the last decade, or two decades, since the search engine has really kind of become now like a giant, massive 800 pounds gorilla. There's regulation now that's put in place.

And social media. You think about when Facebook first came out in 2004, there was no guardrails. And I think Zuckerberg's mantra was move fast and break things. And look where it brought us. We have a lot of issues. And so I think we're kind of in that same like '04 to '08 era of social media, and really even maybe to 2013, where it was like there really were no guardrails. Everybody was kind of doing whatever they wanted. It was a little bit kind of crazy.

And people made lots of money. And a lot of people got beat. And then now the regulators woke up and are like, oh, we got to actually regulate this stuff, right. And I'm hopeful that people will wake up and pay attention and be like, hey, we need to put some guardrails on OpenAI. We need to put some guardrails on Anthropic. We need put some guardrails on Grok and Gemini.

We have to put something around this so that it can protect people. Because I do believe that already there's lots of examples where, I think you said, you search for something, or you ask a question, and it's showing you certain things. It's not giving you a true view of here's all the possibilities. It's here's the 10 that I'm giving you. And then you're really only picking from those.

And it's not a far cry to say, as we move forward, people are going to buy those spots. I want ChatGPT to recommend my product or service versus my competitors or versus an open source solution or a better solution. So, yeah, I do think it is a very important time in our history as humanity to really think about what are the guardrails that we're putting on as a society. What are we going to accept?

And I do think, as folks like you and I that are kind in the arena, so to speak, that need to raise the flag and say, hey, this is important. We really need to figure this out because it's only going to get harder. And the big model providers are only going to do more of this because they're motivated by profit. That's all they're focusing on. And so I do think we have to figure that out.

KIMBERLY NEVALA: So all of that being said, I know you are an absolute optimist, and enthusiast, about the technology, and about our ability to, hopefully, maybe, learn. I don't know that we have a good track record here about showing that we've learned from our prior mistakes, But. Hopefully we will get there soon. But based on what you're doing both internally within StealthX and with your companies, are there a few key pieces of advice, whether it's mindset shift.

See, now I'm proving I'm not AI. I'm going to just embrace all my stuttering and stumbling around as proof I'm human. Hope it works in my favor. Let's see. That you think organizations need to attend to today?

DREW BURDICK: Yeah, I think a lot of folks aren't paying enough attention. I think they really need to actually take this seriously. I'm finding in a lot of organizations, they're like, we tried AI in 2023. And it just was hallucinating. And you're like, every now month, it is exponentially better. And in the last 12 months, it's like night and day the amount of things that are possible.

So I think it's like, number one recommendation that I could give is take it seriously. Actually pay attention. Because I think that there's a lot of organizations that are just really, truly not ready. I think big massive enterprises, they've got a big customer base, they've got a big brand. They're going to be OK. I think small companies who are more tech forward are going to be fine because they can pivot and adjust.

I've talked about this before. I think the middle is going to get squeezed real hard because they're stuck in the middle of they're not huge, but they're not small enough to move quickly and be nimble. And so if you're in the mid-market, you really need to pay attention and really start to think differently about how you work, how you deliver value to the market, how you solve problems. I think that's really important.

The second one that I would say is we've found that people often are the hardest part of this. The work is actually more about teaching people, enabling people, equipping the organization, and helping kind shift the paradigm. And that's a lot. That takes a lot of work. The building of the thing is easy now. It's really code is cheap, as we say. You can produce code at an exponential rate now. But the people part is the hardest part, and it will probably be the hardest part for a long time.

So I would invest a lot of thought and care into how do you bring people along. I think reskilling your workforce, upskilling your workforce, make those investments now because in 12 to 24 months, 36 months, it's going to be a different game. And I don't think you're going to be prepared. Your people won't be prepared. And frankly, you're not going to be doing them a service in the long run, because there's a lot of organizations that are just like, yeah, we turned on Copilot. And you're like, that doesn't quite going to get you there.

So anyway, that's number two. And then I think probably the third one, I'm a big believer in just play with the stuff. Get a paid account on Claude, really any of these tools, and just experiment. We're all learning. If anybody tells you they have this all figured out, they're lying to you. Spend some time. It doesn't matter what level of the organization you're in. You could be a frontline worker or a CEO, I think everybody should just be experimenting to really, full understand what it can do now.

So those are the three what I would suggest. And I think, hopefully, as I said just a minute ago, if you have an opportunity, if you're in a position of leadership that you have influence on regulators, tell them to think about this stuff. Call your Senator or whatever it is to help make this important, that it needs to be brought to the attention. Because some of the stuff, it is very Wild West right now.

KIMBERLY NEVALA: These days, if you call my Senator, the list is so long that I don't even know--

DREW BURDICK: I know.

KIMBERLY NEVALA: I'm not even sure a chat bot could summarize it well at this point. So as we move forward here, what are some of the key areas or questions that you're going to be really either pondering, or keeping your eye on, over the next, I don't know. You pick the time frame. I'll leave that one up to you.

DREW BURDICK: I think a lot about right now, there's a lot of people talking about SAS apocalypse. And it's kind of a funny buzzword that's getting thrown around media right now. But I actually think there's a lot of validity to it. Last week, I don't know when this is coming out, but last week, Anthropic launched Claude Security, the preview research version. And in the matter of hours, Palantir and all these big security companies lost like 12% of their market cap.

And they should be a little scared. There's stuff coming out that's pretty powerful. And so I've been thinking a lot about this concept of how does everyone start to move towards having apps on demand? And if you have structured data that's yours, and it's like living in a vector store. Without being too technical, it's like a database, essentially, that holds information. It's organized.

That you can have, essentially, a way to key it so that it's not getting taken by an LLM, it's not being trained by. But it gives you the ability to-- it's portability. And then you can on demand be like, I need this, I need this, I think more and more we're getting to that world. I said this on LinkedIn recently. SAS software seems silly sometimes now. If you really think about it, it's just data and the ability to do things with that data.

And because code is cheap, you can spin something up, and I'm mm obviously not at enterprise scale. I'm talking at consumer level. It just starts to make more and more sense to be like, if you can have whatever you want on demand, and your data is portable, that's an interesting thing. So I'm thinking a lot about that. I think a lot about in person events. I think that that's going to be extremely important.

It has always been important, but I think it's going to become even more important. And so we we're investing a lot of time and energy thinking about how do you get more out on the streets, if you will? Take it to the streets, go be with people, and build relationships, and be able to truly differentiate yourself compared to all the talking heads on the internet. I think being in your community and being the nexus of a community is really important.

So those are the things I think a lot about. I also find that as we've been hiring, I'm like, what is the right roles to hire now? Because all these roles are truncating. It used to be you'd have all these different specialized skills. And you're like, OK, it doesn't really make sense to hire this role. So those are things that we're grappling with right now. And I'd love your opinion. We're actively trying to figure that out. So, anyway.

KIMBERLY NEVALA: Yeah, well that just brings up an interesting point back to, again, it's a little bit of that tension between expertise is required and industry knowledge is required. If you want to be able to differentiate in there, you need to understand, to some extent, how it works, or at least understand the subject matter to know how it could truly work. A lot of tech problems, the reason that tech doesn't just solve the problem is we don't actually understand the problem. Which has a lot to do with us messy, wonderful people involved.

And so that even though what you just said there just highlights such an interesting tension right now, which is we're not developing some of that expertise. And then we're also questioning whether we need some of that expertise as well. And I don't know where we're going to net out. I'm not convinced we're on a net positive trajectory at the moment. But I am very heartened by the fact that folks like yourselves, and others, are attending to the question seriously and not just saying poo poo, it'll work itself out.

It's going to be hard for a while. Because I think when folks are like, it's hard for a while, but it'll work itself out in the long term, it's never hard for the person that's saying this it's going to be hard for the while. So I suppose it's the people in the middle that I worry about as well. But we will definitely link to StealthX and also to your podcast, which is a fantastic and looks at a lot of these types of issues in quite a great deal of depth.

But for now, I will thank you for your time, and your insights, and for being patient in getting this back on the books. For various reasons, a third time was in fact, the charm, so this has been fantastic.

DREW BURDICK: Thanks for having me on the show. It was a great conversation, I appreciate it.

KIMBERLY NEVALA: Awesome. And to continue learning from thinkers, doers, and advocates such as Drew, you can find us wherever you listen to podcasts and also on YouTube.