AI infrastructure spending is at a peak and your pilots are not paying for it. Jason Finkelstein's answer is that the data center buildout is a bubble, and that most AI projects fail on implementation rather than on the technology. Finkelstein is a startup advisor and VC and a guest on The Artificial Intelligence Podcast.
On the data centers, his assessment is blunt: "There is a bubble with the data centers." He is working in a $300 million data center in North Carolina and calls it a new wave, but the wave breaks on miniaturization. A supercomputer once filled a room, and the same power now sits in a phone. AI chips have roughly an 18-month lifespan, so the hardware bought today ages out fast. Data centers in space or under the ocean get the same verdict. By the time one could be built, the compute will fit in a watch and there will be no need for it.
Why do AI pilots fail to deliver ROI?
Because the buying process skipped the questions that used to come first. Finkelstein describes companies signing large contracts after a three-month due diligence period instead of the year, 18 months or three years it used to take, on the reasoning that it seemed really cool. The old sequence was: what problem does it solve, what is the long-term ROI, what is the benefit. Skip those and the pilot fails, and the AI takes the blame for a planning failure.
The definition problem makes it worse. Clients tell Finkelstein they want to implement AI and then describe five things they want to do, none of which are AI. The term now covers video generators, image generators and word processors. OpenAI has been making the same promise for five years, that the tool will save time and solve problems, and the commercial has not changed even as the tool improved. Companies that cut staff on that promise had to hire them back, and the trust did not come back with them.
Which AI companies survive the bubble?
The ones with a fan base and a second revenue stream. Finkelstein's view is that Google and Meta have the advantage because AI is an add-on to businesses that already make money, while pure-play AI companies have nothing else to fall back on. His summary: "I think fan base wins out to be honest with you."
OpenAI and xAI make their money from AI alone. Anthropic has the best product right now and no secondary product, and Finkelstein's read is that it is building ill will by cutting what customers get rather than raising the price. The xAI prospectus puts 15 percent of the money on rockets and the rest on AI. He places the market at the beginning of the cycle rather than the end, with the cream rising and the pretenders falling, and he compares the moment to 1999, when a pitch on a napkin could get funded.
What is the next wave after the data centers?
Cybersecurity. Finkelstein's pick for the next big confrontation is AI against AI: "I think AI versus AI is going to be the next big confrontation." Deepfakes and voice and video impersonation are already convincing enough to fool people who should know better, and he expects automated attacks to be met by automated defenses.
The exposure is self-inflicted. He has seen projects where a company gave an AI tool access to everything: customer credit cards, social security numbers, medical data. He has also seen chief executives pick the tool they like over the one they have a contract with, even when it is the same tool behind a different login. Startups rarely hire a security officer while chasing growth, which makes a newly funded company an obvious target. On education, his answer is human intervention rather than an AI, because the lesson has to come from a person.
What to do before you sign another AI contract
Answer the three questions first. What problem does it solve, what is the long-term ROI, what is the benefit. If the honest answer is that it seemed really cool, that is your answer.
Keep the change to your staff as small as possible. Finkelstein's rule when he implements a system is to ask how it can be done with the least amount of change to how employees already work, because the more behavior you ask people to change, the longer it takes and the less likely it is to succeed.
Do not hand over all your data, and do not move to a tool outside your contract. He has seen companies give a tool access to customer credit cards, social security numbers and medical data, and he has seen leaders choose a preferred login over the contracted one.
Bring in someone who recognizes the problem and can implement it. Finkelstein's point is that the problem is not with the AI, it is with the implementation.
Work on the boring part. Infrastructure, numbers, consistency, a tweak to ROI or conversion percentage. It is the least exciting work and it makes the biggest difference.
The tool is not the variable. The implementation is.
Original show notes
Are data centers the future of AI — or just the next bubble? Jonathan Green talks with startup advisor and VC Jason Finkelstein about the infrastructure boom behind AI: why the data center buildout mirrors the dot-com era, where the real money in xAI is hiding, and why today's billion-dollar server farms may be obsolete before we finish building them.
Key Takeaways: • The data center boom is a bubble — miniaturization will make massive AI infrastructure unnecessary within a decade • Fan base beats technology: companies with loyal ecosystems (Meta, Google) will outlast pure-play AI providers with no secondary revenue • Cybersecurity is the next wave — deepfakes and AI-powered fraud are already outpacing most people's defenses • xAI's IPO reveals where the real AI value lies: 85% of SpaceX's prospectus valuation rests on AI, not rockets • The subscription model is coming for personal computing — the same shift Spotify pulled off with music is heading to AI hardware
Notable Quotes: "I'm working right now in a $300 million data center in North Carolina. That's a new wave. But as you see with the iPhone and the technology and the computer, you're spot on. It's going to miniaturize. There's going to be no need for it. There is a bubble with the data centers." — Jason Finkelstein "I think fan base wins out, to be honest with you. I think you get by as long as you have a strong fan base." — Jason Finkelstein
Connect with Jason Finkelstein: LinkedIn: https://www.linkedin.com/in/jasonjfinkelstein
Enjoyed this? Follow The Artificial Intelligence Podcast and share it with anyone tracking the AI infrastructure race.
Auto-generated transcript, 5,189 words. Timestamps link to the moment in the episode.
Our data centers, the future of AI. We're going to find out with today's amazing special guest, Jason Fingelstein. Now, to have you here because this such a hot topic. Everyone is really interested in data centers. We're talking about building data centers in space, but we're also seeing can fit inside your phone. Yeah, underwater, all these different ideas. And it's this huge topic. And we've now seen with the XAI IPO that. The biggest people keep thinking of ~ it as a rocket ship company,
but even in their documents they said, well, fifteen percent of the money comes from the rockets and the rest of the money's from the AI. So even the whole prospectus in the future really is seeing this value in AI. And everyone keeps saying, We're in a bubble, we're in a bubble, we're in a bubble. And we all know that things eventually come to an end, but ~ how do we know which part of the bubble we're in? Is it still the beginning where it's time
to jump in? Is it the middle we need to watch out, or is it the end we need to jump ship? And what are the warning signs as someone who saw the dot com bubble and kind of lived through I was in 1999, the height of the dot com bubble. I saw, I mean, you could literally write a pitch on a napkin and get funding for your idea. you had all these companies that were overvalued. And I was with Shopping.com. was the
biggest app building here in 1999. It was about... Brought up by eBay for $900 million. And I think it's parallel to what we see today. I mean, you'll see the creme de la creme, like Elon Musk and OpenAI rise to the top, and the other pretenders will fall. I think we're in the beginning. think it's not time to jump ship yet. But I do see parallels to the 1999 dot com bubble, absolutely. One of the things that I look at and think
about is that Google has been kind of the furthest behind. Like they had those pretty embarrassing presentations the James Webb telescope at the beginning, which cost them a lot of market share. You only have to get one question wrong, right? they're the one company that AI is like an accessory. So OpenAI only makes money from AI. now x.ai is planning on making most, if not all, of their money from AI. But Google has this other revenue stream in the same way that Microsoft does.
They have all these other services. And so they're almost doing AI as an add-on. And I feel like even though they seem like they're behind, they have this advantage, which is that They have these other revenue streams. Like they can say, hey, here's an ad onto your Google workspace. In the same way that the only people I know who use grok are people who already have a paid Twitter account. They go, I get grok for free. There's nobody who goes and chooses grok of Thank you.
all the AI models. It just doesn't have that same cachet. So do you think that's gonna affect the long-term winner, the fact that there are these other revenue pipelines? What do mean we have this podcast? You're an expert, You're an expert. You got this on bat. 100%. I think like Meta and Google have the advantage because they have the other things that fuel other practices that will feed into the AI business. 100%. You talk about Anthropic and how the costs skyrocket and what happened with
it when they stopped. with the, I think, with no KPIs. Yeah, it's interesting you being up anthropic because They're a company that they have the best product right now, but they seem to be on this mission to make their customers hate them. So everyone uses them begrudgingly. Hahaha! Every single person I know who uses them, and I'm a power user, is like, I would love for any model to be one percent better so I could jump ship I don't they're they've built up a massive
amount of ill will and they're like bragging that they've just gone profitable. It's like, yeah, you just doubled your prices. Like so, and they ~ or like you pay a subscription, Right, right. they half your deliverables, right? It's you have Netflix and half the screen is black suddenly and they hope you don't notice. So they rather than raise the price, decrease what they give you, the value, but it creates this kind of ill will in the marketplace and they're extremely vulnerable to the
market because they don't have another revenue stream. They don't have a secondary product. They don't have a word processor or a hardware option or anything else. And yet we're seeing them get away with it for now. Do you think that's something that a company can get away with long term? No, I think that's part of the dot com bubble we talked about. A company like that, as AI increases and improves and user experience gets better, anthropic is going to fall by the wayside.
It might be short-term gain, but long-term loss. What do you think is more important? Is is it the infrastructure or is it the like fan base? Cause like Elon Musk has built this c cult of personality and we see people that love open AI and they're like, they're open AI till the day they die. Do you think that's as important? I love OpenAI. Okay. I think that's a great question. I think fan base wins out to be honest with you.
I think you get by as long as you have a strong fan base. So going back to the dot com pub bubble issue, we've seen so many startups get massive amount of fundings. And I've worked with several startups that they just add AI to their name and s and people go, Well, I wanna fund them just in case it's a good idea. We kind of saw this like in the like around 2020 with any company that was doing something in the metaverse or web three.
It's nobody knows what web three means. Still nobody knows what web three means. No I've never been to a website and been like, this is web three. We keep hearing these terms. We keep we still have people talk talking about the blockchain, but most people and like tons of companies if you put blockchain at the end of your name, like, we're a real estate company but blockchain. Yeah. Real world assets. Web3. And now they're like, We're a real estate company, but AI. Yeah. Yeah.
It's funny. It's funny you should say that because for a while, I mean, I hit a sore spot with VC for a while. was the money dried up. The money dried up and all of sudden AI came out and people started putting AI and you could find anything blockchain, anything Web3, anything. It's crazy. It's open wall. It's typically closed before. One of the things that AI has really changed is like buying cycles. So we see companies signing these massive contracts and their
due diligence period is three months instead of one year, 18 months, three years. I've seen a lot of these and I remember I've worked with some CEOs and like, we signed this huge contract. And I go, Why? And they're like, Well, it seemed really cool. And it used to be What problem does it solve? What's the long-term ROI? What's the benefit? We're starting to see now a lot of these pilot programs, they go, the ROI isn't there. And I wonder if that's because the
AI failed or because at the beginning of the project, they didn't plan it the way they normally do, which here's the problem. This is what the tool solves. I see the same thing with a lot of startups when I go, They didn't plan it. what problem do you solve? They didn't plan it They just jumped on the, oh, it's AI, it can solve my issues. Let me jump in on it, it seems pretty cool. And they didn't plan it properly and it failed them.
There seems to be this inverse correlation between how exciting an AI tool is and how useful it is. Like the most used tools are like video generators and image generators, which don't have any actual real world use. Like they don't actually make the world a better place. They're just funny. They're good for making TikToks and not anything that benefits real company growth. The things that are most useful are like a tool that's more efficient with handling your infrastructure or your shipments of your products between
here and China, if you can increase efficiency 1%, makes a huge difference, but it's boring. But no one talks about that. And I see this with a lot of startups. They really chase excitement. So many startups I talk to do something cool, but they don't solve a problem. And I know in my business, whenever I work on the boring part, the infrastructure, the numbers, the consistency, tweaking an ROI, tweaking conversion percentage, that's the most boring thing, but it makes the biggest difference.
During the beginning of a bubble, it seems like it's all the focus on the hype. When do you think we'll start shifting to the well, does it actually work? We go from the promise to the results testing. I think when enough customers get fed up and waste too much money and they realize that it's not working, right now it's just in the innovative one pool type stage. Once they realize they wasted all this money and fired all these people and they're behind the wall, that's when
people start waking up and saying, okay, get rid of the bad, go with the good. I find that The definition of AI has changed dramatically in my lifetime. It used to mean sentient computer or sentient robot. And now it means tool that makes funny images or tool that makes fake videos or a good word processor. And I find that the lack of definition is one of the biggest problems. Whenever I talk to a client or potential client, they go, We want to implement AI.
And I go, What do you mean? And they describe the five things they want to do, none of which are AI. So we've it's kind of like IT used to mean Yeah. the person that fixes your computer, and now it means anything to do. Do with computers at all. It can be a software engineer, someone working in a server farm. Because we've broken the definition of AI, and now we try to play this game of like, we need it used to be strong and weak AI,
and then now they say AGI or quantum AI or like what is it? Silicon intelligence. Like they use these different names, and it's like all it does is murky the waters when you change definitions. So I find that the biggest problem is we have all these commercials and promises. OpenAI is making the same promise for five years, right? This tool. Will save you a bunch of time and solve all your problems. The problem is hasn't changed, even though the tool's gotten better.
It's the same commercial. And I think that's why a lot of people are getting It's like, well, you said three years ago it could do this. It could replace people. So we fired our HR staff and it didn't work. And we had to hire them back. And then now you're saying we can do it now. I feel like a lot of these companies hyped too early. So that now they've broken a lot of that trust and like the tools
can now better meet the promises they made three years ago, but it's like once bitten twice shy. How does a market equalize or how do companies regain that trust and people figure out here's what AI can actually do and here's what it actually can't do when there's so much murkiness around terms and so much like marketing hype? That's why you need a guy like you and me who understand the products, understand the benefits, and you hire a consultant that actually recognizes
the problem and can implement it. The problem is not with the AI. The problem is with the implementation to the companies. We're seeing a lot of knock on effects from infrastructure. And one of those areas is that computer parts, the pricing is going through the roof. Like basically the AI companies have said we want all the RAM. And so RAM prices of five X, seven X, 10X in different markets, which means people can no longer afford a computer because the pricing has jumped up so much.
Pri computer prices have doubled for a lot systems. And so now people are like buying older computers and goes, it actually works fine. And we're seeing infrastructure as AI data centers. And the thing that I wonder about with AI data centers is like computers when I was a child, like they filled a room, right? A supercomputer was an entire room. And now you have all that power in a phone. So I feel like data centers are a sh short
For n they make sense now, but in 10 years when the AI is 50 times smarter and can fit inside your watch, because we continually succeed with miniaturization that will move past the point where you need them. Cause for most people in an ideal world, the AI lives inside the, they don't have to wait for it to send the signal back and forth. It's an immediate response. Like that's a real personal assistant, as opposed to one that you have to call and it answers.
So do you think the data center boom is another type of bubble? 100%. I mean, I'm working right now in a $300 million data center in North That's a new wave. But as you see with the iPhone and the technology and the computer, you're spot on. It's going to miniaturize. There's going to no need for it. There is a bubble with the data centers. What find the most fascinating is the people that are really excited by data centers in space. And as someone,
my very first job was IT, fixing computers when they broke down. And it's like the idea that an a data center would need no human to do any maintenance, which is what you'd have in space, right? Unless you're going to send up a rocket or an astronaut every time to do a repair, let alone the heat issues. It's like ever something always happens. So I I feel like that's a really good hype and it's a marketing story, but it seems like the cost and because
data centers and because technology is changing so fast, like the chips for AI have like an 18-month lifespan. So you shoot up all these rockets and 18 months later, you have to shoot someone up to replace all the parts. It seems like really crazy. Plus, when I was a kid, the last place you want AI is in outer space where It's us. you can't get it in case it turns bad. So it seems like a crazy idea. Do you think
But you also mentioned that we're thinking about doing data centers like under the ocean. Do you think these things will really happen or happen or they're just a way of like boosting IPO values? I'll be honest with you, that's one of the reasons why I'd invest in SpaceX because overhyped. When talk about Mars exploration, is that really going happen? Maybe, maybe not. But I'm saying he's putting his valuation on that. I think it's hype. think by the time we get to the point of data centers
in space or water, as you said, it's going to be miniaturized and there's going be no need for data centers. I think that what we're seeing in this AI cycle, at least from my perspective, is that we keep chasing the next excitement, but we never stop to think about human behavior. Like whenever I work on a project, I think, how can I implement AI or this automation or this system with the least amount of change of the company employees of the staff?
Because the more you ask people to change your behavior, the less the longer it takes and the less likely a system is to succeed. And now we're trying to shift because to this mindset of. Well, most of my computer is somewhere else. We're so used to the personal computer, which we transition to, right? The computer is in my laptop. It goes with me wherever I go. My data is right here. And we're already seeing people move from the cloud back to local infrastructure again.
So we are all local infrastructure, then cloud and then back. This idea that we want to switch back to remote to where my computer doesn't have anything in it. It's just a connection tool. It's just like a television. So a television doesn't work unless it has a signal. We're I don't imagine people wanna go back to that. That's a very hard story to sell unless you can convince people, well, now your computer will be made ch way cheaper, but there's nothing in it. It's kind of
like, They managed to do it with Spotify because for a long time when they called it renting music, no one wanted to rent music. I want to own a CD, I want to own the liner notes, I want to have a sense of ownership. They managed to create that transition where people pay a subscription because stopped calling it renting. Like they got that word doesn't work. They started talking about, well, you have a million songs and you you can have it for free.
And if you want no commercials, you can pay an upgrade. They changed the language. Good, good, good. Do you think they're gonna be able to make that work with AI and personal computers where you go, well, instead of paying three thousand dollars per computer now, you pay twenty dollars a month and when you don't want just give it back. Do you think people will kind of get pulled into I do, I do, 100%. Trends, just follow the trends. See what happens in other industries.
It's gonna happen in United. As Sullivan says, there's nothing new under the sun. So what do you think kind of is the next thing coming? Because we now we have the data center evolution. What do you think is the next big opportunity or the next big kind of wavefront coming in AI? It's funny, I was all about blockchain. I think cyber security is going to be the big one. You know, with all the deep fakes and all the videos out there.
I think we'll see with Anthropic. I see it was a cyber security threat with the supercomputer. I think cyber security is the next big area. Glad you brought that up because I feel like people's security consciousness has really lowered over the past few years because they're so I've worked there's four. on projects where people are so excited about AI, they go, we gave it access to all of our data. And I go, What do you mean all of your data? Like your customers' credit cards,
your customers' social security numbers, your customers' medical data. Why do you trust this machine? And there's this ~ I worked I've took this work security kill swore, and they say it's we A big part of it is like how much do you trust this company, right? Like how much do you trust OpenAI with your data? They haven't been around that long, like long enough for you to have a strong sense of are they good or bad with data? We already know what Facebook does with
their data, they sell it, right? We already know what they're gonna do. We already know what Google does with their data. They look at everything and they sell it. And they're probably listening to you through your cell phone. So like you know what you're getting. And I think that some people have kind of accepted that we lived in live in a post-privacy society. Like they've just accepted that they're under surveillance by their phone all the time and that all their data's out there.
Well, look at how people post online with Instagram and Facebook and TikTok. Their whole lives are open block. Yeah. Forget their security with technology. People, I went out to dinner. I just did this. mean, people have no shame anymore. Yeah, I mean, you're exactly right. So I feel like that's starting to drift into enterprise where they go. I've worked with people and CEOs who go, Well, I know we have a secure AI, but like I like this one more.
And I am like, Yeah, but we have a contr we have a contract with this one and it's the same tool. It's just a different login. That you just get this, well, I like it more, or it there's this mindset that's either I'm too small, it will never happen to Right, this is like most people, their first blog. No one would hack my blog. I'm too small. I usually think that way. And then you go, and you kind of like I see a lot of startups,
they don't hire a security officer because they just chase growth. They SISO, that's something you hire after two years of growth. And it's like, yeah, every if everyone knows that, then that's what they're looking for. Like you just watch TechCrunch, any company that gets funded, you go, Okay, they just got funded. Let's trick them. And there's so many The vectors you can use now to trick people. Like you could do a video call and seem like someone.
You can do a voice call. You can be so convincing. I think that we are going to see a shift towards we have to meet in person, face to face, and have to touch you. Then that's the only way I'll know you're a real person. I've been scammed Like I feel like we're gonna have that shift. before with the whole face call, video call, and I've been fooled. And it's it it's so good that it can that's the thing is that even everyone's
is out there. Just like you said, if you've posted one video anywhere, like my voice is out there. Anyone can get my voice, there's nothing I can do. There's AIs trained to talk like me and sound like me. It's like you can be very, very tricky. And we I've seen it where if you think it's your boss or you think it's you're kind lower your defenses and you go, my boss wants me to go buy some gift cards. I better do it. He's in an emergency.
So I've seen that work. And what do you think is the wavefront of cyber technology because with any new technology, right? We get so excited until bad things start to happen. We saw this happen with computers in the 90s and early 2000s and websites where people weren't paying attention to security, then a bunch of bad stuff happened. And then we saw it with the first wave of chat GPT where everyone was using it and then like lawyers were getting caught with cases that didn't exist.
So they were using it with no sense of security or like cautiousness, right? No trust but verify, just pure let's use Well, how can cyber like we're gonna have to switch to back to that? Like we go to open to close, open to close. We're gonna to shift to the cybersecurity policy. What do you think is the wavefront? Or what do you think will be the future of it? Is it gonna be AI cybersecurity versus AI attacks? Is that really the future? AI versus AI?
Or is there gonna be like what do you just what do you think is the future? Maybe I'm guessing we're wrong. I think, I mean, I don't know, but I think AI versus AI is going to be the next big confrontation. I do. So it makes me think about this ~ dead internet theory that like most comments on the internet are AIs posted to other AIs, You which I I more and more think that's true. And I wonder if that's we're gonna see,
because I've seen these companies like phone companies have started to make ~ fake AIs that will answer the phone if they think it's an AI calling. And so it's like more how much of our internet bandwidth and electricity and phone bandwidth is going to be used just by AIs talking to each other, right? So it's like becoming more and more electricity and usage for this negative usage is Well that's why data centers are so popular right now. Because the bandwidth is needed until we
figure out a way to shrink it and miniaturize it. Do you think it's possible? Because one of the big problems is like education, because people don't realize all these different vectors. Like, I know people who do spam emails for a living, and I was people still click on a link in an email from someone they don't know who it's from. And it's like all the time. Not even the bad end of it. Like it's a spam email selling stuff
to like lose weight, let alone the stuff where they're trying to st phishing, Right. which people click on those two. So he wasn't even in misleading part of it. And I was like, what do you mean? He's like, for every number of emails, a certain number of people will click on a link no matter what, which blew my mind. So we still have people clicking on that like African print from like 1992. It seems like it's impossible to keep education up to pace
the different number of ways to trick people. As much as we try education, most people don't educate it until after something happens. So do you think that the wave is in developing like these shields that don't human involvement, or is there gonna be a huge push to provide education and people are gonna start implementing their own defensive measures? I saw on TV, I Which direction do you think is gonna be? think it's going be human intervention because I saw on TV how people would
track all these scams and call them out and video them. And think education has to come from a human perspective, not from an AI perspective. So with how many much of your experience, how much of your business has shifted from I need to meet you in person? It's a handshake, I need to see you, let's meet in person. Like do you think the value of the interpersonal meeting is gonna go up and is going up now because of AI? What's funny,
people remember with COVID, people were like, I need a remote job, need a remote job. And I was like, if you start your career remote, you're to miss out because I am 30 years into finance and venture capital. All the meetings, the CEOs, the partners, it was all because of interpersonal contact. If you don't have that, you lose out and you're not going to be successful. So do you think we'll see a shift to back to the office? It is as someone who works remote,
~ absolutely. I like I have to do. work for mine as well, but I have one of my, I'm a board member of a company now. I met them online. They were in Miami. I spent a day with them at their office and now it changed the dynamic of the relationship. You can't, you can't beat that. Yeah, think that's so important in that what's interesting is we're seeing people use apps as a way to like decrease the need like we have dating apps
and ordering apps, and then it's like now you could order food and you can tell the person drop it off and don't look me directly in the eye. It's like it used to be only prints that that was the only person who had that And now it's like I could order McDonald's without having to talk to someone or look anyone in the eye, and I can just pick it up after they're gone. And now with AI, we can we're like creating more and more parasocial
less and less ability to communicate. I actually and this is something I try to teach my kids. I think the ability to talk to people and have a normal conversation is going to become a rare skill. Like if you can just walk in a room and talk to people. I think the problem was media with Instagram and Facebook. Kids are on their phones now. What would we do? We'd go outside, we'd play, we'd sports talk, interact. We had it different.
mean, the ability to have a conversation is a life-changing lesson. So I think this is really powerful for people who are really interested. Can you tell us a little bit about more about what you do, the things you're doing with data centers and kind of what you think the future is and where people can find you online and What you're doing now, what you're focused on now, kind of what you're working I'm a startup advisor VC, so I help
companies from MVP all the way to ~ &A, sale, with business development and marketing. I'm pretty much, if you look at my LinkedIn, I'm a connector. I know everybody, finance, founders. I just put the right people together and help businesses grow. Perfect. Keep it simple. I love that. Thank you so much for being here today, Jay. So I'll put the link to your LinkedIn in the show notes and everyone who's watching this on LinkedIn, of course, you're attached in the video as well.
Thank you so much for being here, giving us your time for an amazing episode. You didn't need me here. You're an expert. You can ask all these questions yourself. No, you were great. Thank you. Thank you, John. It was a pleasure. Take care.
