Loading summary
A
Welcome to the New Books Network.
B
Hello, I'm Dan Hill.
C
And I'm Julie Annixter. And this is Real Transformations, where we talk about change from the inside out.
B
Because real transformations aren't just business strategies, they're human journeys too, filled with hope that can either be leveraged or lost.
C
And we're happy today to have Craig Hatkoff with us. Craig is an author, a serial disruptor and and the founder of Dragon Camp. So welcome Craig.
A
Thank you for having me, Julie and Dan, it's a pleasure to be here.
C
So let's start in the present. Craig. Dragoncamp. I know it's a methodology, it's leading to a lot of really interesting big ideas in AI. Can you give us the backstory? How did you found it and what are you going to do with it?
A
Well, the backstory is it's a very organic, rooted in real life situations and my experience engaging with AI. And I'll give a little bit of the lineage AI, not so much as a tool, but as a collaborator and actually at this point almost a full thought partner. And I've been doing this since November 30, 2022, the day that ChatGPT was launched. And you know, the sort of the origin story is the first time I used ChatGPT, I had two reactions. One is, oh my God, this is magic. And I know it's not magic in the traditional sense, but in the Arthur C. Clarke, any sufficiently advanced technology should be indistinguishable from magic. And I immediately said this is unbelievable. And I also had the other reaction is that I could see immediately the power and the potential and that's both. Let's keep it value neutral for the moment, you know, because there's always the good, the bad and the ugly on any of these new technologies. But I said this is like a fire breathing dragon. And the problem though is you gotta learn to train your dragons. And that's been sort of the thematic. And so that's the origin of my first experience was oh my God, you know, what have, what have we unleashed both for good and for bad? And it evolved organically as Claude, Deepseek, Grok, you know, even I guess it's meta AI now. Everybody keeps changing. And I found myself working very regularly with nine of the what called large, large language models. And even the definition, what's a large language model is quite interesting. But you know, I've got the whole gamut, Perplexity, Copilot, Gemini. And when you put them together as sort of your teammates and it doesn't matter what the domain is, what I saw was the cross model interaction was remarkable. And given my lineage, having worked with Clay Christensen from the Innovator's Dilemma going back, I guess I started working with Clay in 2000 just as innovators dilemma and disruptive innovation theory was catching cold. It wasn't when I met him, I didn't know who he was, what he was doing, but just as at that inflection point where he was about to, you know, through the Andy Grove connection for those people who follow Clay's work. But what really emerged was a pattern. As you talked about it's Dragon Camp became part of a larger ecosystem of how can we trust AI? And everything I would say at this point, trustworthy AI should be. It's on the, I'll call on the tips of the tongue, but no one that I've seen so far has come up with a methodology that could really scale and take. One of the greatest fears are the errors, hallucinations, fabrications. And I'd say given. And I'm doing a lot in, I'll call it the legal field. I'm in some interesting litigations that I can, you know, at least talk about. I think I mentioned it in my consensus speech, which I think you've taken a look at, but it, it evolved into a new kind of. I don't want to overstate the philosophical metaphysical, but it's kind of a new epistem. It's an epistemological change in how we view and try to confirm what's real, what's not real, what's true, what's partially true, and engage with multiple AIs in what I will call fierce iteration back and forth. And then the key is how do you track who's saying what to whom? But I found it nothing short of remarkable. But it is not going to work.
C
Oh, sorry, No, I was just going to say I've observed you, I've worked with you in this approach. And the one thing I can say up front is it's very different from the way that most of the people I see and know are working with AI where they're using AI as prompts or they're trying to put agentic AI into workflows and processes. You are collaborating. So could you talk a little bit about what that experience of just co creating with these dragons or AIs has been like?
A
Yes. Nothing short of an experience. I never even could imagine that I would be having. You know, my interest in innovation kind of goes back to my high school days. Even though I didn't have any of the language. My father was a classic disruptor before, you know, going back to 1954. He gets credited for kind of having the first big box retail. We took over back in 1954, the year I was born, an old fire station. And my father created discount toy retailing long before Toys R Us. And so it's kind of been, as we say, you know, it's born this way. This is the bloodstream.
B
Yeah.
A
So that's a little bit of the context. But what it led up to was, okay, we now have this, you know, what is Dragon Camp? Is one of the questions. Little hard because we're still soft around the edges of definitions, actually on almost all the terms you get, not a lot of standardized terms on all of the jargon and lingo, but it felt like it belonged as part of a bigger ecosystem. And it led to something called verimatics. And verimatics is a combination of the, let's call it a portmanteau of VERIFY and Maddox's testodia. That's in mathematics. And so what we're seeing is a whole ecosystem that revolves around the protocol and then how can we create a methodology that actually verifies? So it's kind of Dragon Camp now goes hand in hand with verimatics, even though they could be used separately. But it's the ecosystem where I'm getting a lot of engagement from people and it's just been pretty wild.
B
So I have a question. If I go past Dragon Camp and I take your adjudicated approach to AI, it strikes me suddenly, when I was watching your Vinna speech, that this could radically change the annual performance review. Because one of the critiques, of course, is that the boss is biased. Over time, you get a lot of emotions going on in that worker, staff member relationship. So what happens if AI adjudicated AI using your kind of console of nine elders from AI world, are suddenly getting input and then making co evaluations of the annual performance review? Is that one application potentially, I think
A
that falls into a continuum. And I'm going to give you kind of the continuum up front because it'll make more sense. There will be autonomous. Yeah, there will be autonomous, which is everybody was concerned about. These are the, let's call it Waymo, driving around with no customers. I can't remember what city it was introduced where they have waymos going around in cul de sacs. Yeah, there's a great deal of fear. Let's kind of get that on the table. And fear and Humanism. I'm a humanist at heart and very little of what I see going on starts with the human not in the loop, human in the driver's seat. So we start with autonomous. There's no car and there's no seat for the human. You just call it up, you take whatever it says and it's a black box. The next level, as we move towards the human side, would be agentic, which involves some level of human interaction. And that might fall into your question about, you know, what's going to happen in the future. But it's largely driven by the AI component. But the human kind of has a role. And then one that we're sort of focused on is the next level, which is human genic, human plus agent, working together hand in hand. And that's the sweet spot for Dragon Camp Adventure. Because if you take a look at what Sachin Nardala from Microsoft is doing, he's got a different version of using the AIs, the multi, you know, multimodal AI. You're seeing a lot of people embracing that. But what he does and you know, is he puts his questions, his inputs into his own sort of desktop. I guess I call it agent. I don't know if we technically call it an agent, but he's a coder and he sticks all his questions into his model, lets them fight it out, appoints one of the models as the Chief justice and then comes back at the end of the day and what did they tell me? And then he's a busy guy. You know, he uses it for coding. He uses it for who's the greatest cricket player of all time. So he's getting close to, but he's hasn't really injected the human layer in yet. And to me, we haven't yet seen anybody doing it at this level. And when I say we haven't seen, I am constantly asking who else is doing this? And so doesn't mean we're covering everything. But if you have nine AI saying, well, here's the closest thing we can see, but it's really different. That's kind of where know we're, we're headed. But this is not human in the loop. This is how do we keep the human in the driver seat?
B
And so that sounds like that's good. This could upgrade the U.S. supreme Court. But we'll leave that aside for now. So.
C
So you mentioned, you know, let's get fear out of the way right up front. And you know, Dan, being an expert in emotions, was saying today that the
A
opposite of fear Is anger that you
B
are in control of the situation, anger and feel like you can make progress on your own terms or at least want to make progress on your own terms.
C
And of the core emotions, probably the most positive is surprise, which is, could be the good form of surprise is curiosity, which you have. You're steeped in Craig, and happiness. And my observation is that you've been enjoying working with AI and AI and the dragons has fundamentally shifted a lot of the kind of ways you're working. Most of the people I know are afraid of AI or I see fear being the predominant emotion out there. And I think one of the values of Dragon Camp and the way you're working is to shift that. So how could Dragon Camp transform the way people feel about AI?
A
Yeah, I was watching an interview, I can't remember someone, one of the, I think it was the CFO, Sarah Fields from OpenAI, and she talked about joy and curiosity. That's what they hope to achieve. And as Julie, you may know from just my background as a children's author, we used in picking our subjects what we called the two things the goosebump test and then the AWE principle. Goosebumps, we would say, no goosebumps, no interest. And that triggers one of the, you know, the EQ elements of curiosity. And then we have the AWE principle, Awe, wonder and enchantment. And that's what we lost since I'm going to go back to the 1600s when they separated the arts and the sciences. So how can we reintegrate the intuition? You know, I actually call it the romantic element versus the purely empirical.
C
Well, Craig, since we're on the New Books Network, I think it would be very interesting for you to share one or both of the experiences that you've had as an author this year. You know, you've written, I don't know, at least over a dozen children's books. You've sold millions, 10 million, 12 million into the education system. And then you had a chance to renegotiate your contract.
B
Yes, Tori,
A
change the name to protect the guilty. I found myself inheriting. I had reversionary rights. Since this is a book podcast. My books have been out there now for 20 years. I wrote them with my young daughters at the time who were 5, 6. They're now 27 and 31. And the market has changed. But as the, you know, and the book business is a long tail business and the 20 year run is a long time. And I had what are called reversionary rights. And so from my publisher, all of A sudden I had the educational publishers trying to, you know, come in and renew for the educational market, which I knew very little bit about the economics and how do you come up with pricing? And to make it a simple story, I dug in with my nine models and what they offered me. By the time I was done doing the Dragon Camp protocols, It was a 10x and the publisher agreed to it. And so, but it's not like I asked one question. I did a deep dive and really began to understand how the educational channel, not the retail channel, but the educational channel works. You have to get approved by Texas, you got to get approved by California, by New York, or you're not going to get published. And so the AIs, that, that one transaction accounted for a lot of my $20 a month payments for the 9a. But. It was a. But I, and I said to the, you know, in the email exchanges with the publisher, I said, look, I don't know, I can't understand this. Here's what nine AIs had to say. And I sent them the analyses, they came back, they agreed to it. So it's. Yeah, but that's what I mean by I'm not using it at the theoretic, I'm using it in my real life instances.
B
Sure. And sticking with real life. Looking at the workplace, I mean the role of innovation, a lot of work is now team oriented, but it strikes me that if it's a team of five, you can add in nine AI powers to be and you have a completely different kind of team. Can you? Surely you must, you're such an amazingly creative person. You must have thought of all of the ways in which this is going to permeate the workplace and change the processes, the limits, the possibilities, everything.
A
Yeah, well, let's say very hard to predict what's going to happen. Every week there's a new amazing unveiling of a new feature, a new image generator, video maker. I think we are at the stage where there's going to be the early adopters who are not afraid and who are excited. There'll be the middle, you know, kind of the early, but not the earliest, you know, the classic curve and then the laggards at the end. And I think what we're going to see in the workplace and actually starting to see, it's fascinating at the most senior levels. Let's start at the senior level. The head of, I believe it was kpmg, the chairman sent out a note to all the partners that said if you do not embrace AI, you should leave the firm. And we have to completely change our business model from throwing bodies at it and, you know, per hour charges to creating effectively what I'm going to call SAS chat bots that are very sophisticated that they can then sell to the clients on a subscription model, because that's just the better model, you know, as long as it's sticky. Um, so I think what we're going to see in the workplace, the earliest versions where one, how do you identify, you know, your true early adopters who can really make a difference? And it may not be coming, it's not likely to be coming from the C suite. It may be a young kid who, you know, is a summer associate who's incredibly conversant in all the AIs. So it's going to be a great leveler and it's going to change the hierarchies. And the question is, how do the corporations, A what can you confirm that that's what the model is going to be be? How do you sort of integrate that dynamic, but for critical thinking, what's this called? Innovation and creativity, I'm going to put that writ large. Creativity writ large. Using this collaborative process where the, the human element, and let's call us the semantic layer with the syntax layer of the AIs, we'll create a whole new model of. What Clay Christensen kind of inspired me with is look for the anomalies. That's the most interesting thing you can find are the anomalies. The anomalies lead to insights and the insights lead to emergence. An emergency.
B
You know, I, I, I love that. I got a PhD in English and it's always in the story you're looking for, where's the tension, where's the Frisian, to use the French term? And that, that's where the energy, that's where the inside, that's where everything that's
A
going to happen and is okay. And to that, to that point, nine models, I would say, well, they're all trained the same way. Well, not exactly. They have the core, you know, training component. Um, but the different models, even if they're similar, create enough friction to generate sparks. To your point. And those are the sparks that like the anomalies. And adding, you don't necessarily need nine dragons at this point on every problem. I don't need nine dragons to tell me the date of the JFK assassination. That will be an autonomous exercise. If we're, you know, asking about how the Enlightenment impacted the, you know, the use of the metaphor, that's a very mixed element. And you know, I can now pinpoint the death of the metaphor. I know when it ended was 1667 at the Royal Academy they banned the use of metaphor. I think his name was John Spratt or James Spratt, I think. And so I. How long would it take me to find that? Forever. And I may never find. But they wouldn't say, let's see what happened to the death of the metaphor. But I'm finding in bridging the gap, humans think emote, the whole empathy, the whole eq. That's how we really process the world. We have, you know, just so much brain capacity that I'm complete, I think in metaphors. I. If I can't find the metaphor to explain something, it, I'll come back. Maybe I'll. In a month or a year I'll have a different view. We need these new kind of call them the bridges.
C
Speaking of M4, I've got to jump in and say, you know, your dad was the big box toy retail innovator. Dan's father ran post it notes for 3M, the print operations for 3M for post it notes. So that's how he grew up. My dad innovated in the wire and cable distribution business and you innovated in the mortgage backed securities business. And speaking of metaphors, you've got one about a certain Ron Popeil story that I would love for you to tell here because. And how it relates to creating verimatics. Because clearly on this call we've got people who are thinkers in very unusual ways. A lot of curiosity. What was that?
A
So let me start with a premise and then work a little bit backwards. What we have right now, it doesn't matter how sophisticated the models are, it's still a black box. You either accept the outputs or you don't accept the outputs. What Veramatics is trying to do is create a bulletproof glass house where every claim can be run through this verimatics process to see if it's correct, if it's a fabrication, if it's an error. But we will have the provenance and the lineage. And it looks like this would be a great application of blockchain technology. So using block and that's different for your readers. It's not Bitcoin is not blockchain. And the, you know, the use cases for blockchain is really a perfect model for verifying with dates, with tags, with hashes. So you know, who said what, when, what the sources were. And you can then follow it all the way back. So that's the setup. Now we get to Ron Hope heel, who in the 60s and 70s, I would say popularized. I can't say he was the only or the first, but the one that really gained traction was Ron Popeil and his Vegematic. And okay, it was a, you know, a vegetable slicer and the phrase that he used along the way or the, you know, when the commentators would come in, you know, you can slice a potato, you can slice tomatoes, you can slice onions, but not only can you slice them, you can also dice them. So the tagline was it slices, it dices. And that's from the 60s and 70s. And I loved, I almost stay up to watch the Ron Popeil commercials. They were performative. And there's a great article by Malcolm Gladwell on Ron Popiel's kind of the world's greatest salesman, who we think probably Steve Jobs had seen the infomercial just on how he presents it. But do you want to ask a question in the meanwhile? Because, you know.
B
Oh, well, I was just going to say I remember the Saturday Night Live
A
parodies of Rob Appiatic, which was probably one of the most famous. Look it up on YouTube.
B
I don't know what I'll remember distinctly.
A
Well, for the audience. Look it up on YouTube. Yes, that was the parody. But when I was a young banker in the late 70s, I was a terrible commercial banker. So as I was about to get fired from commercial banking at the then Chemical bank, now JP Morgan, I was devastated and they kind of put me in the, they gave me a timeout and said, look, considered career possibilities in alternate industries, but, or certainly not at our company. But you're very smart. We understand what the problem is. And the answer is I was a mismatch. I'm not really a manager or that kind of process, detailed oriented person. I'm a little bit more on the invention, innovation, entrepreneurial side of the spectrum. And for traditional banking, that was, that wasn't the right kind of personality. So you asked me the question, what would the human resources interview be? Very interesting to see what they said. But anyway, I'm sitting in the corner about to be fired and I had noticed some anomalies in the business that we were in and we would just. Banks were just making home loans, putting them on their balance sheet and you know, you lend and you hope you get paid back. And that's. I didn't really love that model. But what emerged was the slice and dice inspiration. Ron from Ron Popeil. What if we took loans and sliced them and diced them and turned them into securities. And the technique, which is what we now would call the earliest roots of commercial mortgage backed securities, one of the villains of the 2008 Great Financial Crisis. I'm basically one of the earliest, if not the earliest identified creators of commercial mortgage backed securities. And I started noticing patterns from AI fabrications, these whole loans, so to speak, and said, I think we need to slice them and dice them to create the securitization of artificial intelligence. So applying Ron Popeil's slice and dice to commercial first, it was actually residential, but we don't have enough time to get into that. But I did the inverse of residential mortgage backed securities and commercial mortgage backed securities became the insight for what we think the future of Dragon Camp and verimatics will be.
C
Well, I know that we're probably only sharing the on the new books network, but since we're seeing you on video and Dan is a facial coder like expert, world, world guru expert. What would you say about Craig's face during this interview? What has it told you about him?
B
That he's delighted by all the surprises that he's come across that have surprised others, but first surprised him in like, this is really what I'm at. This is my career. This is what just occurred to me. You're an amazingly curious guy and that's wonderful to behold.
C
And you don't even know. Next. The last tranche of Craig's creativity was first the Tribeca Film Festival, which he co founded, but then the Tribeca Disruptor Awards, the Disruptor foundation, now the Disruptor Awards. And for anybody who is interested in seeing some examples of people who are out there in every walk of life doing amazing things, they can go look up the Disruptor Awards. And Craig, I know you're thinking about reinstating the Disruptor Awards with a. An AI flavor. Can you just give us a little peek into that thinking. Peek into that tent?
A
Well, I just, I just spoke at the consensus conference and it was a remarkable experience. You know, I, I'm a crypto healthy skeptic, which means crypto writ large, which is all the new technologies, whether it's stablecoins, NFTs, Bitcoin, Solana, Ethereum. I'm still in that. Let's watch and see what happens. Because there's no lender of last resort in the crypto community and we have financial crises, which is sort of my. I've been through 14 financial crises and the only way we get out of these crises is typically the Federal Reserve steps in and floods the market with liquidity and, but when I went to consensus this year, if I went last year, it would have been as they, you know, the commentary as these are merged, a lot of Lamborghinis and tech Bros, you know, and crypto czars. This year it was all what we would call suits and ties. J.P. morgan, Blackrock, Blackstone, Carlisle, State street, the institutions, once they pour in, that's when it becomes legitimized. So I see a lot of the innovation, almost everything is innovation driven. And I was more interested not so much in just the crypto side, but what are all the things flowing from this kind of new system? We've got some real challenges for the traditional tools of the Federal Reserve. The dollar's under assault, we're in these long wars, inflation's back up. So it's really, I see the new generation, there'll still be some of the traditional disruptive innovators, but at every booth I went to, in every speech I heard, I said, oh, my God, this just falls into the bucket. So I'm looking forward to keeping it, but sort of maybe using this conference as a platform.
C
Very cool. Well, it has been a delight and just so great to have this conversation with the two of you. Thank you, Craig. And I know people can find out more about what you're doing at Dragon Camp AI and at the Disruptor foundation and on Substack, we have real hyphen transformations we're just launching, but this has been really fun. Thank you so much.
B
Yeah, such a pleasure, such a pleasure for having me.
A
And it's, it's hard to pull it all together in one. That's why it's got to be in the form of a narrative. And that's, you know, Ron Popeil. Thank you, Ron Popeil.
B
We, we'll, we'll do the 12 hour version next time.
Date: June 4, 2026
Hosts: Dan Hill & Julie Annixter
Guest: Craig Hatkoff, Author, Innovator & Founder of Dragon Camp
This episode dives deep into the evolving concept of “trustworthy AI” through the unique lens of Craig Hatkoff, an innovator, author, and founder of Dragon Camp. Hatkoff shares his methodology of collaborating with multiple AI models (the "dragons"), discusses the humanistic side of AI adoption, and explores the need for rigorous verification systems (verimatics) to build trust in AI-powered processes. The hosts and guest explore how these frameworks impact industries, individual creativity, organizational structure, and even cultural attitudes toward technology.
“I said this is like a fire breathing dragon. And the problem though is you gotta learn to train your dragons.” (Craig Hatkoff, 02:12)
“I did a deep dive and really began to understand how the educational channel, not the retail channel, but the educational channel works. … By the time I was done doing the Dragon Camp protocols, It was a 10x and the publisher agreed to it.” (Hatkoff, 13:45)
“This is not human in the loop. This is how do we keep the human in the driver seat?” (Hatkoff, 10:18)
“Goosebumps, we would say, no goosebumps, no interest. And that triggers one of the, you know, the EQ elements of curiosity. And then we have the AWE principle, Awe, wonder and enchantment. And that's what we lost since ... the 1600s when they separated the arts and the sciences.” (Hatkoff, 11:55)
“It's going to be a great leveler and it's going to change the hierarchies. … For critical thinking, what's this called? Innovation and creativity … using this collaborative process … it will create a whole new model.” (Hatkoff, 15:46 & 17:00)
“What Veramatics is trying to do is create a bulletproof glass house where every claim can be run through this verimatics process to see if it's correct ... We will have the provenance and the lineage.” (Hatkoff, 20:55)
On the AI moment of awakening:
“The first time I used ChatGPT, I had two reactions. One is, oh my God, this is magic … And I also had the other reaction … this is like a fire breathing dragon. And the problem though is you gotta learn to train your dragons.” (Craig Hatkoff, 02:05–02:26)
On the real-world application of multi-AI negotiation:
“By the time I was done doing the Dragon Camp protocols, it was a 10x and the publisher agreed to it.” (Hatkoff, 13:53)
On the value of anomalies:
“Look for the anomalies. That's the most interesting thing you can find … The anomalies lead to insights and the insights lead to emergence and emergency.” (Hatkoff, 17:20)
On bridging the human and AI worlds:
“Humans think emote, the whole empathy, the whole EQ. That's how we really process the world … I think in metaphors. If I can't find the metaphor to explain something, I'll come back … We need these new … bridges.” (Hatkoff, 19:42–20:12)
On future disruption:
“I see a lot of the innovation, almost everything is innovation driven. … I was more interested not so much in just the crypto side, but what are all the things flowing from this kind of new system?” (Hatkoff, 27:00)
A moment of camaraderie:
“You're an amazingly curious guy and that's wonderful to behold.” (Dan Hill, 26:00)
Hatkoff’s approach reframes the role of AI from that of a tool or threat to a collaborative partner, advocating for a “human-in-the-driver-seat” model that leverages the strengths of both AI and human intuition. Through frameworks like Dragon Camp and verimatics, he advocates for transparent, verifiable systems that make AI more trustworthy, while urging a cultural shift from fear to curiosity, awe, and inventive exploration in the workplace and beyond.
Further information on Craig Hatkoff:
This episode is essential listening for anyone interested in the philosophy and practice of building trust in artificial intelligence, as well as the human side of technological disruption.