Drew D’Agostino on Crafting an AI Workplace
In this episode, Drew D'Agostino, founder and CEO of Crystal, dives deep into the realms of AI and Bitcoin. Intended for Tennesseans by Tennesseans, the discussion explores Drew's journey in the tech space, from creating a popular Chrome extension for LinkedIn to navigating ethical dilemmas in AI application. Drew also shares insights on leveraging AI for business growth, the challenges of entrepreneurship, and his unique perspective on managing company finances using Bitcoin. With a blend of relatable stories and expert analysis, this episode sheds light on the future of AI, the principles of building a tech company, and the intriguing dynamics of deflationary technology.
About Drew D’Agostino
Drew D’Agostino is the Founder & CEO of Crystal, the world’s leading personality data platform, helping professionals communicate more effectively through AI-driven behavioral insights. With a background in psychology, data science, and SaaS, he has spent the past decade pioneering technology that enhances human connection in business.
Named to Forbes 30 under 30 and co-author of Predicting Personality, Drew’s work has been featured in Fortune, CNN, MIT Technology Review, Wired among others. A lifelong entrepreneur, he embraces a long-term mindset—building Crystal with a culture that values craftsmanship, intentional growth, and personal connection.
-
Spencer 00:35
Drew D'Agostino, founder and CEO of crystal. Welcome to Signature required. Thanks. I'm happy to be here so you're in this cool space of AI, which I feel like, that's kind of all you have to say when you introduce yourself. It's kind of like the internet in 1999 when people were like, what do you do? It's like, I'm part of the internet. And not everybody really even knew what that meant in 1999 but all you had to say, it was just like, you know, I got a website, and that just immediately made you cool. So you're doing stuff in cryptocurrency, Bitcoin, AI, so I feel like you've kind of checked off all the boxes. So what is it that you tell people that you do? Unless you just say, you know, just cool tech stuff.
Drew D'Agostino 01:28
Yeah, I've been trying to figure that out for the last 10 years. People say, What do I do now? I usually just say, out of business. And then if they
Spencer 01:38
ask more questions right away,
Drew D'Agostino 01:39
If they actually want to know, then we'll go deeper. But what I my actual answer the question has changed quite a bit, because the definition of AI has changed. So I never said AI before, and there's a short window of time when I can legitimately say AI. And now if I say AI, it's like, oh, cool. You know, another one, another one of these guys that adjusted their LinkedIn profile over the last year. You know. So it depends on the audience, I guess.
Spencer 02:07
All right, well, for an audience that, let's say, hasn't had the opportunity to hear from someone that really is doing some cool stuff, break it down into a basic level, what are you doing? Because I think it's pretty cool.
Drew D'Agostino 02:22
Yeah, that's so when we started 10 years ago, really, our main product was a Chrome extension that you could add to LinkedIn. So, and a lot of people still use that. Still our most popular product, where I'm looking around LinkedIn, there's a chrome sidebar on there, and it just says, view this like view personality, and it'll analyze on the page all these little they're called personality signals. Oh, wow. And they'll tell you about the style. So if I'm going through a process of my daily, you know, meetings, whatever, I quickly look up a person. That's kind of where we started. So is that very accessible data set? And then we realized, okay, this works really well for resumes, so let's make it so you can analyze resumes for like, the recruiting use case, and then realized you could start building this data set, as long as you could link people's identities, and it works best for like executives and people with significant online presence, but as long as you can find that structured or even unstructured content, like things you've said or historical like professional experience, or a lot of companies have these big executive bios, stuff like that. If you can find that, aggregate it and then analyze it, you can build a really accurate data set that is not trying to be perfect, but it's trying to directionally tell you these like behavioral patterns and trends.
Carli 03:46
Okay, I have to ask the question, because the elephant in the room for me is, when you were talking to Spence, you're like, man, it's super creepy that they look at all of your history and try to know that what you're shopping for. How do you creep proof yourself? I guess in this space where you're not just I would maybe argue that looking at somebody's behavior online is one thing, but then, like trying to get into the depths of their soul and how to best talk to them is another. So how do you, how do you manage that?
Drew D'Agostino 04:17
So having been wrestling with that for a while, I would say there's a couple levels to speak about it. The one, before you talk about any technology or data use, you have to think about at the core, we are two people, and we're interacting. Does it feel weird for me to be using this tool when I'm interacting with you, like, beyond whatever the technology is doing, or you get, like, really technical with it. I just have a gut check as far as, like, if I'm, I mean, I use this for for going into meetings, if I'm selling to this person, what I want them to know that I use this. And if the answer is no, it's probably something little bit creepy about little conscience check, yeah, yeah. But if the answer is. Sure, and in fact, yeah, I want them to, because then they'll know that I'm going out of my way to try to help them have a better experience. Then that that passes that test for me personally. So if I'm using any tool, and not just my own, not just crystal, if I'm using any sales or any software tool, I always want to be using a tool that I'm totally fine with the other person knowing I'm using that's kind of my own internal check. It's kind of like a golden rule of, like SAS tools only use, use the data about someone else that you would have them use something like that. So that's, like the the philosophical part of it, and then when you get into the technical parts, you just have to think about, all right, if I'm analyzing some information, does the person have a realistic expectation of me using that to about them? And generally, if someone's putting a public profile up there, you're you're expecting Google to be scraping that. You're expecting someone to be even not technology. But I'm looking at it and I'm making, you know, micro assessments about the person in my head, like, Okay, looks like you're kind of, you're this way. You're communicating. It seems like you're you've got a good sense of humor because you've written in this way. So it's kind of automating the perception, which, to me, is just the same thing we do anyway. So that's kind of where I where I land with it. I don't know if that I think that's helpful.
Carli 06:22
I think AI listen. I use it in my daily life. It helps me write, it helps me check my kids trig homework, because I haven't done trigonometry in a really long time. Like, I am an early adapter of AI tools, but I'm always curious, maybe on the surface right, like, on the very populist viewpoint of like, how I could use AI in my life, but as we're starting to really think through in our own ethics and our own thoughts, how far these can go, I do feel that tension. I think a lot of people do. And it's just, I like how you said it's a gut check. It's like, this is the future. How do we use it responsibly? How do I do that in my own life? How do I let my kids use it? Because, you know, today, it's like, can I ask AI about this before I write my paper and have it do research for me, it's like, no, but is that what they're gonna do someday? So all that to say I'm wrestling with that tension in my own life. So I'm grateful to know that you are too.
Drew D'Agostino 07:21
Have you run into any of those things where something where you try to use the AI and then you got that weird gut check? It's worth, I don't know,
Carli 07:29
Yeah, I don't know that. I want to say it on a podcast, like in this moment in time, but if I'm looking at you with authenticity, yes, I have had that moment. And I was like, this might be a lion that I don't love, and I think I worry too. I looked something up the other day I was working on a project, how it's getting better and better, but it's not always factual, and I like that. It sounds like your product is working with things that individual people have put up on purpose to be analyzed, to be known, and it seems like you're just aggregating that data to make it more usable for people going into a sales situation.
Drew D'Agostino 08:05
And a key thing there's this overall AI question. So when it's important to distinct what we're talking about too. So things like it's called hallucinating when the AI doesn't when it makes up something. And these large language models, like chat, GPT, they they will, more than any person, give you the wrong answer as confidently as possible. So, I mean, it happened to me the other day. I was one of these moments. Was for me, I was like, I've been trying to learn about a particular company, and is this stock undervalued, whatever. So I'm, like, trying to use the AI as my consultant, and I'm giving all these questions, like, alright, if my goal is to protect downside in this case, and stock prices like this, and this is what the assets are, you know, I'm giving it these questions as if it's a person, I'm just testing it out, seeing, and it like, gave me these, like, trading recommendations. I was like, wow, this is cool. I'm gonna be like, Rich. I'm gonna throw this chat at it. And then I actually tried to, like, execute this thing, and I realized, huh, you know, this thing just communicated to me in this way that was extremely confident, except if I actually made the decisions on here, it was gonna, it was, it was gonna, like, basically, immediately lose a ton of money, because it just it. It sounded so confident in what it was saying, but it factually was just wrong,
Carli 09:17
often wrong, but never in doubt, is what we say.
Drew D'Agostino 09:21
So these language models do that Crystal's a little different, because we're not based on language models. We are a little old fashioned in this world, in that we use what's called,
Carli 09:32
oh, we're at a place where AI could be old fashioned. Please enlighten me.
Drew D'Agostino 09:37
So our AI is not even really, I mean, if we were to show you, like the technical parts of it, it's not. There's nothing to do with current large language model AI, we use what's called Bayesian inference, which is basically statistics and probabilities. So based on, you know, the weather over the last. 10 years, we've seen that at this day it's usually cloudy. Therefore we're going to give you an 85% probability that this day is going to be cloudy this year, that kind of thing. So it's a lot more about just statistical modeling. In our case, it's like, all right. Well, I'm looking at Spencer's LinkedIn profile. It looks like he's interested in golf, and based on these 18,000 other people who we saw fill out our personality tests who are also interested in golf, it looks like this gives him a 71% chance of having a dominant personality trait, you know. So it's kind of like that, where we're just taking all these statistics and projecting them forward. AI today is a little different, where it's more of a black box where you can pass it information and it will just give you an answer, and you don't know how or where it came from. So that's a little bit of a distinction between what we do and what currently,
Carli 10:51
What I might be interfacing with more well. And I really appreciate that distinction, because AI feels like it's such an umbrella word like the World Wide Web, right? And understanding we're gonna all have to get much more comfortable with the nuances of what the various technologies are, and gosh, if it spits out good stats and data, that's like your heart language, babe.
Spencer 11:12
Yeah, I think it's a, you know, part of what I take away from it is we've always known that anything that goes out on the internet is there forever, and that has to be the guiding principle for AI and any interaction is that if it's out there, then it's out there, and it's fair game. And we may not like that, it's fair game, or we may not feel like that the boundary line is being respected through the development, but I think it's just an important reminder, as we, you know, are teaching our kids and just mindful of ourselves, it's like, you put that post out there, you post that amazing meme. That meme was funny in 2010 it wasn't funny in 2025 right? I mean, there's stuff like that that. It's like people have something that 20 years ago looked different, sounded different, didn't age well, and it's like, wow, you have to be really, really careful, because it was one time where manually, we might have to go scrape through someone's profile to say, Okay, let me look and see what they were talking about 10 years ago. But we do that just in a way that's not very technology forward, but now we've got really powerful AI that can go through and then half of one millisecond look at everything that you've ever posted and draw statistical conclusions about who you are. And that is, that's something to be reckoned with very much.
Drew D'Agostino 12:40
Yeah. And you see, I mean, you always see the top of that with famous people, people, yeah, because, and it feels like a celebrity problem, yeah, because. All right, that's got ton of information. So the AI is going to say this thing about Donald Trump or Elon Musk, people who have a ton of information out there, and it's used. But the crazy thing is, what you're saying is that all of us already have this public footprint big time, so way more than we think. Yeah. So to the AI, it doesn't matter who we are, it's training on all of it, sucking it all up. So that's, that's where you really need to be, you know, aware and careful, yeah? Because the, I mean, we're only in the first inning of all this stuff, yeah. So the the AI is only bound to the data that it's trained on. If it doesn't have access to that publicly available data, then it doesn't know anything about that. So we all have that opportunity to, yeah, inform it what we want to, yeah.
Spencer 13:41
We’re talking to the next generation of kids that are growing up in a world of tech AI, and have an interest in saying, hey, I want to be proficient in this. I see that this is the direction the economy is going. The amount of money that is being paid to people that are skilled in AI is similar to what engineers and programmers were making a generation ago. So I always look at somebody like you drew and say, I would have no freaking clue where to begin on anything related to AI. I'm not a programmer. I'm not really a tech guy too much. So if you're talking to people that see that AI is the future, like they're listening in and they can say, all right, 10 years from now, is AI going to be more a part of our life, or less a part of our life? I think the majority of people will say, whether I like it or not, it's going to be more a part of our life, and I want to know how I can participate in it, how I can benefit from it. How would you counsel people? Because I could make up stuff of like, you know, start going towards programming or start, but I wouldn't even really know where to begin. So teach a. How to follow in your footsteps if someone wants to go that direction.
Drew D'Agostino 15:06
Well, I'll start off by saying everybody I am figuring this out, just as much as anybody else is. I can tell you the principles that I'm building crystal on, and I think those those principles I've tried to align myself personally with them, and I think they also could apply to just building a career, especially if you're a young person trying to figure out what, where do I invest my time? Yeah, like, especially if you're looking all the disruption like is, are the these jobs even going to be jobs anymore? For surely as I saw a tweet about how Y Combinator is putting, you know, they put out this call for startups, and most of the things that they're trying to focus on building are these, like, $100,000 or more jobs just to be completely automated away. So it's always big questions like, will this actually be there?
Spencer 15:50
Yeah, because you think about like lawyers right now and saying, you know, there's a lot of Bill billable hours that are going to go away because AI is going to be a really proficient writer already is, and so you see all those spaces and AI, despite being very different each day, I think we still can get to the core principle that it's going to be more, not less in 10 years. So how do we get more educated? And how would you direct youth in saying, right, I want to be a part of this.
Drew D'Agostino 16:24
I'm building, I'm building my company on these three pillars, AI, EQ and BTC. That's kind of how I think about the the first principles of crystal. All right, so diagnose what those, all those acronyms are. So AI, we've been talking about it, and the reason why we are heavily investing in it is because I view AI as it's where our P and L growth comes from, so our profit and loss and where our business growth is going to come from. I mean, we play in this narrow space, but literally every industry you have the ability to take tons of manual or tons of work that's being done in an inefficient way right now and use AI to make it more efficient, and then probably to multiply what you're doing. So there's immense scale and efficiency and accuracy that can come from using AI well.
Spencer 17:16
So are you a programmer, like, can you program in it? Or is it like, I outsource, and I depend upon people that are experts in that space. I'm a business leader, but I don't take the time to know the guts of it. Well.
Drew D'Agostino 17:30
I am a programmer. I built the first like few versions of our product, but now you truly do not need to be a programmer to use AI. In fact, being a programmer, I'm finding might actually be a little bit, try to think of a good metaphor for it. You can get in your own way because you're used to something needing to do something from scratch. Therefore, I'm going to try to do this on my own. Yeah? Meanwhile, someone who's not a programmer stepping in can actually often move faster because they're not even they're not even tempted
Spencer 17:59
A football player trying to transition to a coach is that rarely have our best coaches of football ever played a down of football themselves, which is a really bizarre thing. You would think, yeah, the best coaches would come from the best players. Oftentimes, it's the opposite.
Drew D'Agostino 18:16
Yeah, exactly. That's kind of what it's like. So I'm finding that people building products right now, super fast. They're not programmers. They're just using AI and asking it questions, saying, How do I do this? And then they're just so I think if you're, if you're someone that that first leg of it, I've been trying to train myself to do that too, and like getting very, very good at asking questions, but not going in with answers. I think that's this key with AI. It's we are all now prompting these tools. So the way that applies in my business is that everything I do as like a CEO, it's, as you know, there's just a ton of laundry list of different jobs that come up, and some of those are people oriented things, some of them are communication, some of them are technical, whatever they are, I'm kind of bringing that to the AI, and I'm finding that some of sometimes this actually really helps me and allows it to scale. Sometimes it doesn't, then I realize, okay, this is probably not the best use for it, but I think that applies in every industry, especially obviously ours. So I think AI is that kind of first pillar getting very good at using it, and I think that that is going to be ultimately, as we're seeing even this weekend with the release of deep seek, and how these models have all become, they're starting to become commodities.
Spencer 19:30
And for those that haven't heard so, deep seek is a Chinese released product for artificial intelligence that basically has turned the entire technological world upside down. Nvidia, one of the chip manufacturers that makes chips for artificial intelligence to run kind of the computer guts. When deep seek was announced, they had the largest drop in their stock price in the history. Of the US stock market, yeah, not their stock price, any company, any company ever in the history of the stock market. They declined the most in a single day because of what this Chinese firm had released in deep sea, which was basically, we can do this artificial intelligence, but we can do it with a fraction of the computing chips that were thought to be required. And obviously that was not what Nvidia wants to hear. As a manufacturer of computer chips, they like AI to consume lots of computer chips. So anyway, I just wanted to give the context for Deepseek
Drew D'Agostino 20:33
That's it. And with that now, regardless of what happens with deep seek, specifically it's it's kind of showing you the pattern that this is going to follow, where, right now, these AI models are very expensive to train. They're really big, and they require these massive data centers. This is, this is all going to be a Race to Zero, where we're getting smarter and smarter, and everybody ultimately has access to these models. So as an individual, what do you do with that? You just, I believe, start understanding, I think you need to understand your world well and think about it a little bit more critically, like, all right, well, this is clearly a market need, because I'm, you know, doing this. Say, I Say, I'm a lawyer in this area. I have an opportunity and I'm not a lawyer, but I could freak out about it and think, oh gosh, just say I was going to take my job because I'm spending all my time doing this thing that an AI could do better than me, and get paralyzed by it, or I could be that forward thinking lawyer in that field Who's then basically working with it, and experimenting and just playing around, not fully, like delegating My Tasks immediately, but playing around and seeing, Where can this make me better? Where can I automate things? Where can I scale things do 10 or 100 times more work output? So I think that's how i That's how I'd move forward with it if I was in that role. So that's the AI part, and that, that's, that's our piano EQ is kind of this what I was talking about with our product. Now that's, that's where we actually bring differentiation and value to the market. By EQ, you mean emotional Yeah? Like, emotional intelligence, yeah. This is it's really interesting that as AI, swallows up more and more and more tasks, and we have these questions even think back in like that, that legal example, ultimately, most of the things that AI are going to be able to do and take up are things that humans are doing right now, and we're just doing it a little worse than computers, because we have a computer in our brain. But there are certain things that AI will never be able to do, and those, we kind of call them soft skills, but I think the soft skills are actually the ones that are hardest for AI to replicate. So if we we've always kind of thought, I was, like, that's a nice to have, but, like, you really want to go into a STEM field because you want to have this, like, hard skill. It's weird how it's almost flipping where the pendulum is coming back, yeah,
Spencer 23:14
So it's, it is interesting, because you do say, is it safer for me to be a plumber, right? Because, like, how much is AI going to be able to turn a wrench and fix a leak at 2am in the morning, versus how would ai do as a psychiatrist to sit down and connect with someone on a human level and an emotional level? It's like two opposite ends of the spectrum, and I think you could make a pretty compelling argument for either side about how it is. But like, just as another example that you know, Carli and I are in major danger of AI replacing us, because you can load up chat GPT, and say, here is a book I want you to condense the lessons that I should take this book into a 10 minute podcast. I want there to be a male voice and a female voice, and I want it to be kind of funny. It will, literally, in a couple seconds, create a podcast that you can listen to with an audible voice, male and female. That's a little bit funny. And it's like,
Carli 24:22
That’s not funny. I don't find that funny at all.
Spencer 24:26
That's exactly right. It's like, that was not very entertaining at all. GPT, so any case, I just wanted us to take us down that rabbit hole for a second that it is this, is it? Do we want to be that soft skill? Do we want to be the plumber?
Carli 24:39
And I think the answer is yes
Spencer 24:44
It might be like the end of the bell curve of saying we need to live out in the edges, versus some of the things where we can see pretty clearly, like, this is in the cross hairs of AI, yeah,
Drew D'Agostino 24:54
There’s, there's a lot of angles to it. I think what I would Yeah, maybe I would loosely say i. EQ, because it's kind of our that's our field. We're talking about, like understanding personality and nuance and these things that computers really have trouble with. But maybe more broadly, it's just the things that humans will always, or the foreseeable future, will be able to do better than computers. So there's a bucket of those. Some of them might be the soft skills and like interactions. Some of them might be the fact that nai is going to have a lot of trouble going into a crawl space and replacing a water heater, you know, yeah. So I think there's a, there's a there's a big group of that, of those things that can kind of fall into that, into that second bucket, or that second pillar. And I call it EQ, we could call it something else, but those things are going to Jeff Bezos has a really good, I don't know if it's quote, I would butcher it if it was a quote, but he was giving a he was in an interview, and he said that people spend a lot of time, or entrepreneurs think, spend a lot of time thinking about what's going to change. Everyone always wants to know, how's the future gonna be different. And he said, we actually, at Amazon, think about a lot about what's going to stay the same, because we want to build our business and anchor to that. So he always talks about that customers are always going to want lower prices. Customers are always gonna want, you know, the faster delivery
Spencer 26:24
Sounds like how Warren Buffett invests. You know, he's like, Warren Buffett has become one of the wealthiest human beings to ever walk the planet, because he's like, You know what? People liked Coca Cola 50 years ago, and they still like Coca Cola today. And, you know, they like cheeseburgers, then they like cheeseburgers now. He invested in Rails, right? I mean railway that that is from a long ago generation, and it is interesting, and that, I love that quote, because it really does turn conventional wisdom upside down, that rather than trying to be on the next rocket ship, it's like what's been actually very successful for decades and decades and decades. You just want to be more on the Coca Cola side and less on the pay phone side. Yeah, don't get it wrong, but you could see which ones have been producing over time.
Drew D'Agostino 27:15
And that's and that's like, I didn't do this on purpose, but that's a great transition to the third pillar, which is BTC, Bitcoin, because I think the result of all this is we are going to be living through a very deflationary period. And by that, I mean you will get more and more and more value for less and less input, especially as AI starts taking over other areas.
Spencer 27:41
So break that down, because I think if you would have asked anybody, they would say we are going to live through a very inflationary period. That's what people, I think, would say, is that cost of everything is going up, and my money feels worth less. So that is textbook inflation. So you're saying we're about to go through and live through a very deflationary period. And so to give that one that one more time.
Drew D'Agostino 28:05
So I didn't, I didn't make up this concept one, there's a, oh, you're in it now. So there's a guy named Miller. So this guy named Jeff, there's a guy named Jeff Booth who wrote a great book called The price of tomorrow, and it's all about deflation. And this bit in, the main idea here is that technology is in this very deflationary cycle, where, if you think about any, anything that has been that started analog and has been made digital, has gotten much, much, much cheaper. Like calculator examples, great when I was in high school, which was like, I don't know, 2000 3004 like, got this big text calculators, yeah, it's cost, like, 100 bucks, and you had to have that in math class, ti, 86 Yeah. And that was what calculators were.
Carli 28:51
That's what brought us together. Our origin story for another time, but our origin story revolves around ti 86 not kidding, continue. Oh man, this is where you gotta cut out super nerds. Yeah, continue.
Drew D'Agostino 29:04
So if you think about what that was and what it costs, and over a very short period of time, there's a bunch of steps in between, but now ultimately, that's just on your phone, everything that the TI 86 could do, and then more is just on your phone. So that was deflation, because we used to have to spend what it was 100 or $200 on that big calculator. And now everybody just hasn't included so it's essentially free with your phone, and you can look at any area that has been kind of like analog and then made digital. That is continuing, and AI is going to accelerate that drastically if we, if we apply it in something like healthcare, biggest costs in our country, we're starting to see AI being able to do better than doctors detecting certain conditions. So think about this requires an hour of a doctor's time versus a few seconds of an AI's time. Doesn't have that kind of cost. The cost of actually doing these things is coming down drastically. Yeah, now you're what you're talking about. There's monetary inflation, because our entire economic system is built on debt, which, and if you're a holder of debt, meaning that you have lent someone money, if that money gets less valuable, you're in really, you're in trouble, yeah? Because all of a sudden they're going to be paying you back with, they're gonna be paying you back with cheaper, cheaper money. Yeah? So that's a really, that's a really big problem in our entire system is built on making sure that that equation always works, so that, you know, credit could be created. People take out the loans. Economic growth can be created, and they can pay back the money, and then we can keep this cycle going. So you have these competing forces. You have technology deflation. Everything supposed to be getting cheaper, but a financial system which is based on, built on this, like, solid foundation of money, kind of constantly, the amount of money constantly increasing, because we can't have monetary deflation, because that that pokes the whole whole thing, and the whole thing collapses. So we're almost in these, like, it's beyond political it's beyond any real individuals decision. We're just in these forces where the money supply has to keep increasing because the technology is basically forcing it to these forces. These forces are like at odds with each other.
Spencer 31:33
So that's really well explained. That's really well explained articulated. I think that fries enough brain circuits for, you know, 10 minutes of time, of like, okay, I when you think about deflation, you see it's like, you can get a 70 inch television for what a 20 inch television cost 10 years ago. You don't even have to go to Black Friday to do it. That's right. You don't have to rush the doors, yeah, that's right. You can have it delivered right to your house. And that is a fascinating way of looking at deflation in one side that is pushing against inflation in the monetary side. And so bring that full circle for us, because you this was your third pillar building around Bitcoin. So how does Bitcoin play into the tension of the deflation that you've explained, and the monetary system that we have that is historically been inflationary.
Drew D'Agostino 32:30
Yeah, all right, so I'll just take it back to my business specifically. Okay, so the end result of this is that my business runs on a Bitcoin standard, meaning that we pretty much have all of our excess reserves and stuff in Bitcoin, and we, whenever we have cash flow, we, instead of storing it in US dollars, we store in Bitcoin. And that's just how we run the business. That's kind of the end result of this. But I'll tell you 10 years ago when we started, here's where it really, really hit me and why I decided to do it this way. Was, let's even not 10 years ago, but we raised our first round of venture capital back in like 2015, so
Carli 33:11
Oh gosh, that is 10years ago. We're aging rapidly.
Drew D'Agostino 33:14
Around that time, a software engineer annual salary. It's about a call it $100,000 and that's a lot of the types of people you need to build a software company to hire software engineers. And I'll spare like the actual numbers. Just keep it in. Keeping easy numbers. Let's say you raised a million dollars in capital and had 100,000 you basically had you raised 10 software engineer salaries of money, and that's and that's how it worked in the whole game of building a software company in an environment with like venture capital, like that is you want to try to build a product, bring it to market, sell so that your revenue increases at a greater pace, to drive up your company valuation, even if you're losing money, like spending the cash, burning the money, so that the next time when you're out of that money, you can then raise at a higher valuation, and then repeat the cycle until you can do that. That's over simple.
Spencer 34:14
It's like a bucket of water with a hole at the bottom, and as long as the water coming in the top is coming out, coming in faster than it's coming out the bottom, then it's okay, but as soon as that music stops.
Drew D'Agostino 34:24
So the reason, the reason to do that is because, if the money's cheap, which in that environment, it was close to 0% interest rates, it is not worth it to just hold, let's say you raise a million dollars of venture capital, 2015 to 2025 if I was to just hold that million dollars of venture capital, really, all I would have gotten was, I would have started with a million dollars, and now I would have wound up with probably somewhere around maybe $700,000 of purchasing power, if I'm lucky, might be closer to 50, you know, depending on what your expenses are, because of inflation. Question. So if I was to just say, You know what, we're not gonna hire anybody, we're not gonna try to grow this business. We're gonna hold a million dollars, and it's gonna go from a million to what would now be, technically still a million dollars, but probably only, like 500,000 in purchasing power. So it would have been completely irrational to hold that money. So my business was incentivized to spend it as fast as possible in order to generate the growth. And that's how the whole tech like venture capital model works. Because the money is not really worth much. It's worth a lot right now. It's worth as much right now as it's ever going to be. Therefore, figure out how to spend it and generate that generate that growth. So that works for some companies. Other companies, it you end up crashing and burning because you spend the money. You can't figure it out, and that's just how it so if you're any adventure investor, you're going to be investing like that, expecting one out of 10 companies going to be really good. Five of them are going to be complete failures, and this could be four that are somewhere in between, you know, they get bought, or they're, you know, some, some other kind of outcome. That's how your entire premise goes.
Spencer 35:59
But that one company, that one success makes it all pay for all the failure.
Drew D'Agostino 36:03
That's right. So that's that model. And we started with that model because we raised venture capital to this, but then we ran into this, like, it's often called, like, the trough of sorrow, or like, a where,
Carli 36:15
What a name, Yeah, nobody's putting that on a t shirt.
Spencer 36:19
That's a good podcast name.
Carli 36:22
Your podcast will come on and it'll be called the trough of sorrow. Yeah, can't wait for that one. Yeah. It needs to also be a song by Marianas Trench, yeah.
Spencer 36:31
A lot of potential
Drew D'Agostino 36:33
You know, you gave me, you gave me some ideas. Continue. But so the trough of sorrow is typically every startup. Well, most startups run into this, where you end up after the initial excitement, initial spike of growth, reality kind of settles in, and you realize like, oh, wait a minute, this is not we thought everything was going to be, you know, sunshine and rainbows, and then things like customer churn or other issues, they start hitting you. So if you're on a graph, you think the thing's going up into the right forever, and then all of a sudden you hit this, like, plateau or a decline. You're like, oh, oh, no. We got to actually figure out how to do this. You can't ride in the early excitement anymore. That can last months. It can last years. And every startup goes through that and really long term success figure. It depends on how you navigate that period, that trough. Some people can adapt, and they can figure out and get back onto sustainable growth. Others, that is the beginning of the end. So when you land in that trough, you have to make decisions that are really, really hard. And that's that's kind of where entrepreneurs are in their stripes. The beginning is very easy, because if you have something that works initially, it's very easy to keep something up for 12 months. Once that second year kicks in, it gets harder. So we entered that phase, and we were in that for a very, very long time. And frankly, I would argue that we haven't really escaped from it, because we have not, as a business, had that like rocket ship growth. We're a company that is profitable. We have about 25 employees. We've got good, mature product suite. We're very proud of what we built, but we're still only, you know, a fraction of what we thought we were going to be when we originally raised venture capital. And I have these growth ideas. So it's been a grind, like going back and forth between profitability, trying to figure out, all right, I had three times I've tried to scale the team to get ready for that hyper growth, and three times I've had to cut it down. So we have to start over, because I just burned a million dollars and this anywhere because I had this like expectation of this is the venture model we need to be. I have this cash, therefore I need to figure out how to spend it to generate growth. Meanwhile, I know what's going on in Bitcoin world because I'm just like a nerd, and have been, like obsessed with that for a long time, and I kind of always kept that separate from the business until 2020 rolls around. Everything happens, and then there was a big sequence of events where, like micro strategy, big publicly traded company, traded company starts buying Bitcoin. It's hard to ignore. And that's a whole other tangent that I won't go down right now, but that's really it's very interesting. The most interesting thing happening in corporate finance right now, and I was faced a couple years ago with the realization like, Okay, if I was just now back to the original point, if I was just to sit there with the million dollars raised in 2015 and rather than say like, Okay, this is going to be valueless in a few years, or worth less in a few years, I need to spend it as fast as I can. If I was just to stick that in the time at in into Bitcoin, into this sound unchangeable money. I mean that literally would have, if we're measuring by today, it would have gone up about 200 times. It would be $200 million now, there's a lot of reasons for that, and a lot of reasons you can't expect that for the 10 years out in the future, but. But the point is that it actually was the rational decision at that time to store that capital in Bitcoin and to store it in something that wasn't dollars. And any decision I would have made, I don't I don't know what those decisions would have been, but rather than making the decision on the premise that this is getting basically losing half its value in 10 years. Therefore, I need to figure out how to spend it now. So I'm going to spend it not recklessly, but I'm going to be spending it aggressively, and because I need to take these home run swings, rather than doing that if I knew that every dollar I had was going to be worth 200 or, let's say even $100 in those 10 years. You, I mean, but I would be making decisions very differently, sure, because, I mean, think about that now it's or if every, every dollar I spend now it's 2025, in 2035 there's a reasonable chance that this dollar is worth 10 or maybe even $100 I really have a high bar to make that decision, and I might make the decision. I might do it, but it's gonna be a lot harder to so that was the realization for me. It was, it was not about like, oh, shoot, I missed out on this opportunity to, you know, for these gains, which sure that that's ultimately what it is. But it was really about how I made decisions with business, and I observed myself burn so much money because the alternative was just leaving it in $1 that was depreciating anyway. Ever since moving my business to the ever since, so ever since moving to this Bitcoin standard, yes, we've had like, an increase in value of our treasury. But the real value in it is that every decision I make as an entrepreneur now is not against $1 that's getting less valuable, it's against Bitcoin that is getting more valuable. So I still make investments as a in my business, because that's what the business needs to do, but every investment has to meet this bar that's a lot higher, so I find myself making a lot less decisions, and it's a lot more disciplined.
Spencer 42:08
It's like self imposed. But I think that's what's fun about technology, of just hearing you talk through it, is it is a language in and of itself, and a mindset that is really different than how most people look at the world. You know, when you're showing up and working in a job or doing something that is just routine, it's just a tirely different frame of mind versus thinking about these kind of skating to where the puck is going to be. You know, Wayne Gretzky says you don't skate to where the puck is. You skate to where the puck is going to be. And I hear you searching through that as you're looking at, how do I allocate the resources? How do I take care of my employees? How do I do these very real things that as one entrepreneur to another, that both of us, all of us, have spent a lot of time in the trough of sorrow. Carli and I have spent a lot of time in the trough of sorrow. Epitaph, yeah, you know, it really is like, if I could sing, I feel like it's a hybrid between like, hello, darkness, my old friend. I mean, it's just like, right there, that you're, you're just one step away from being right back into the trap sorrow. And I think you're, you do a really good job of authentically talking through that this is hard and it changes, and you can get it right now, but it doesn't mean you're gonna have it right tomorrow. And that's what scares me the most about technology, is that you can get something right in some other spaces, and you've got a five or 10 or 15 year runway of like, okay, I'm pretty safe here. That is not the case, right? That is not the case
Carli 43:56
Five or 10 weeks or months instead of years. With how rapidly this growth is.
Spencer 44:01
So Drew one way that we wrap up each podcast is we give you a short fill in the blank exercise here. So you haven't seen these ahead of time, I give you a very short sentence, and you tell me the first thought that comes to your head that fills in the blank. So just read the question back to me. Dangerous.
Carli 44:18
Read a sentence back to me. What makes it fun with it filled in?
Spencer 44:21
Okay, the future of AI in one word is blank way
Carli 44:32
to start with an easy one. Just kidding,
Drew D'Agostino 44:37
fast,
Spencer 44:39
so read it back to me. So that way we'll cut it. So that's so the future of AI in one word is read it back to me.
Drew D'Agostino 44:44
The future of AI in one word is fast.
Spencer 44:49
The biggest mistake companies make when approaching AI is blank.
Drew D'Agostino 44:59
Okay? Okay, the biggest mistake companies make when approaching AI is assuming that AI should only be replacing what people do and forgetting that AI is going to be able to do things that we don't even fully understand need to be done.
Spencer 45:27
right? I like that. That's stretching the paradigms again, of just this can do way more than what you're expecting and what you even realize today. People have learned today that they can recreate our podcast and maybe do it better than what we're doing. We were gonna cut that. That's exactly right. Edit that out, all right. Last one, if I had to give one piece of advice to an AI skeptic, it would be blank.
Drew D'Agostino 45:54
If I had to give one piece of advice to an AI skeptic, I yeah, it would be Buy Bitcoin.
Carli 46:10
Ah, yeah, okay. That is not what I expected you to say, yeah.
Drew D'Agostino 46:14
Well, because either way this goes, it's going to be, it's going to be, either way this goes, it's going to require a lot of money printing because of the whole deflation part of it that I was talking about before, or we're just going to have a lot of economic displacement and crazy things happening, where you're going to see companies like Nvidia, which are it went down 17% in one day because of a small, little spook. I mean, we have no idea where to store value. Ultimately, this stuff's going to converge into Bitcoin. So it's the it's the safest place. So if you're a skeptic about AI, or if you're a super AI bull, that's the best place to store value.
Carli 46:53
Okay, can I add one? Go ahead, and I don't know that. I'll say it in the like Jeopardy. I don't know that I'm going to form the question in the correct way. But what would you then say to somebody who is a skeptic of Bitcoin?
Drew D'Agostino 47:08
For someone who is a skeptic of Bitcoin, I would say, good. I was too, and most people in Bitcoin were, so learn about it. Well.
Spencer 47:21
There’s some really, really good, straightforward, simple books. I'll link my favorite Bitcoin book in the show notes. So this is where you know, drew you and I met a number of years ago, and I'm really glad to get to sit down with you now, because we've been able to flash forward in our original conversations were around some chips and salsa at a Mexican restaurant, talking about Bitcoin in absolutely what were the trough of sorrows back then? I mean, just really dark days of that, and you were clairvoyant then in having a vision and saying, see it through to the future, and you painted a picture, then that is remarkably accurate to what has transpired in the cryptocurrency space today, and so I really appreciate getting to be in your orbit. Thank you for being here today and helping us stretch our minds a little bit and turn us outside in on, you know, deflation and inflation. The mark of someone that understands something well is their ability to be able to take a very complex topic and simplify it. And I think you did an outstanding job of that today, you could have been just a cool guy to show up and be like, I'm an AI. Don't need to say anymore. Figure it out. I'll see you in five years. But we can clearly see your heart posture is one that you have a heart of serving through AI and a respect of we need to be mindful of what this technology is doing, but please, please, please, get educated about it. Don't ignore it, because it's coming for you either way. Thank you for being here. DREW You did a really good job. Thank you.
Drew D'Agostino 49:12
Thank both you. It's been really fun.
Spencer 49:20
Drew D’Agostino, CEO of crystal, but better understood as AI statistical model, Bitcoin guy, technology guy. I mean all of it,
Carli 49:34
yeah, I just appreciate y'all letting me get a word in edgewise that is like all of your heart language in one podcast conversation that was really fun.
Spencer 49:44
I've known Drew for a number of years, and consistent with a lot of people in the tech space, you just kind of have to give them a little bit of room to roam in the fields. They will oftentimes talk over you. Or they see things that you don't see, and to dismiss them is really to miss it. You don't have to understand 100% of what you hear, but if you can understand 5% or 10% something connects with you and you dig with them there, then you really can have some great takeaways from these guys, because they're also quick to admit. He admitted on the podcast like he's tried and failed a lot of times. I mean, if you're in technology, that's just how it goes. You blow up a lot of rocket ships trying to find one.
Carli 50:34
Well, in entrepreneurship in general, when he started talking about the trough of sorrows, I just we didn't go down that rabbit hole. I was like, golly, it's so relatable to anyone who has started a small business and, gosh, anyone that's raised kids, anyone that's been married, it's like the trough of sorrows is a real psychological dip from anything that you are working on growth towards.
Spencer 51:00
It's a season you have to expect in anything and you don't.
Carli 51:03
And it's amazing how it happens in every area of life, from business to all of it, you name it, yet it's you never expect it, and it happens everywhere. And I just, I really felt like that was relatable and something that I will take away from this conversation, even if I don't think about AI in 30 minutes, I will think of the trough of Soros
Spencer 51:22
yeah, he's running fast and moving at 100 miles an hour. And you could see that in the business. And you have to with technology, like entrepreneurship is already a high speed stake, and technology just makes that 10 times faster. And you get on one wheel for a second, and you can't pull it back out of the ditch. I mean, you're gone. And so I also thought that he did a really good job of explaining some remarkably complex things in simple ways. One of my very favorite quotes is Albert Einstein that says, If you can't explain something to a five year old, then you don't yet understand it well enough yourself. And I really like that mentality, that there's plenty of complexity there, that he could have just talked and lost the whole audience, right? I mean, we wouldn't have understood it. No one else would have either. And instead, he was intentional to give us the depth that we wanted to go to, but he was still able to bring it to a place that was digestible, so that anyone listening to it could learn something. And that's the ultimate objective. Is you don't have to get it all, but get something well.
Carli 52:36
And I think even as somebody that uses AI technology in different pockets of my life, and I know you do too, and we're trying to raise the next generation of kids to be smart and think about it. I appreciated the distinctions he made between his company and perhaps other products that are on the market, and I really appreciated his authenticity in talking about the gut check he has to do when he thinks about the ethical application of some of this tech, because, you know, it's just sometimes feels downright creepy. It feels like we are living in Orwellian novels. And it's hard to reconcile, as a human what to do about that. And there is this distinct desire to kind of put your head in the sand and be like, I don't want to read it. Don't want to read it. I see people do that a lot around elections. I'm just not going to read the news. I'm not going to pay attention. But like politics and technology, it's like, it's happening, yeah, and it's going to impact you. And we don't have the privilege of putting our heads in the sand. So that's why I was really grateful he came on, he made it digestible. I learned something certainly today, and it just makes me want to continue to look for sources that can digest this material and help me feel like I'm educated and I can at least make choices with some baseline instead of feeling like I'm chasing the car.
