The $575B AI Bet (Biggest Since the World Wars) and the New Two-Buyer Reality in Software | Tomasz Tunguz, Theory Ventures

It will be this year the fifth largest infrastructure project ever, including the two world wars. When would you anticipate that? By twenty thirty. And the demand for inference is infinite.

A lot of goods, general partner and theory vendors for every dollar hyperscalers make from AI. They're actually spending twelve dollars on infrastructure. That's a lot of $500. How does it happen?

In the short term, it is market care capture gains. Who wins the most care? Who's number one? And in the longer term, it's a market game.

Who makes the most profit? And today it's the game that's tickets. If you were to ask AI, what is the best laptop to buy? You may even go as far as what is the best cost I could buy.

$20 to $30,000 line if it's in the afternoon. But in the world of enterprise sales, it's much more complex. Otherwise, in a company that is building 82 and continues. And today's guest is Tomas Tunguz from Theory VC, formerly at Red Point.

We just did a deal together, which is exciting. So we'll tease that for now and have more on that soon. Paul, how are you? It looks uh nice and sunny in Vancouver today.

It's beautiful. I've been noticing at the beginning of each of these intros, we always talk about how busy and how much travel we've been doing lately. So I'm I made a mental note to not try and do that this time, even though I feel like that's true. And um excited for this one.

I think people are gonna love the episode. Tomas is one of the most thoughtful thesis investors, I think, in all of venture as a as an industry, and uh so many good nuggets from that episode. Yeah, uh love how he's investing, and a lot a lot of his writing really resonates, but he's so good, uh, so clear on kind of where we're at and where we're going. So definitely uh big fan of his writing.

If you're if you're out there following on X, uh LinkedIn, and Theory VC, definitely check him out. Uh, you know, before we get into this episode, you know, certainly uh public markets and and private markets feel like closer than ever before. Blurred lines, right? There's so much secondary liquidity available right now, and you're starting to see this common theme uh emerge where you know if you've got this AI value chain or where the value is crewing in AI, and of course, that's way more than just open AI and anthropic.

You've got kind of the bottom layer of that, which is energy, rare earth materials, uh, grid. You've got kind of the second layer, which is data centers and infrastructure there. You've got uh the next layer up, which is GPUs and chips, uh memory, et cetera, et cetera. You've got the layer above that, which is the AI infrastructure, which is anthropic as open AIs of the world.

Um, and then you've got the application layer beyond that. So in the public markets, you've seen kind of horizontal SAS, um, you know, have a little bit of a downturn. It's it's rallying back now, big pops from uh folks like Datadog, Zscaler, we're seeing seeing a bunch of others kind of continue to move back up finally. Um but if you look at the public markets and you look at a lot of the companies that have 10xed or uh two, three, four X uh recently, it's a lot of energy companies.

It's a lot of the microns and the sand discs and the folks in the um chips and memory space. It's um it you're certainly starting to see kind of where are the picks and shovels, and then even where are the picks and shovels for picks and shovels. Like everybody's trying to find that edge or that that part of the chain that is um about to break or has you know is is is something in the AI, you know, a pyramid that is foundational and is running on these crazy backlogs or is trying to fulfill some of these crazy backlogs.

So, you know, we see that in the public markets and we're starting to start starting to see it a reflection of that in the in the private markets too, which is a lot of investment is moving away from SaaS and applications and moving into, you're starting to see even um, you know, investors that have previously done mostly software move into companies that are software hardware or mostly hardware with a software component or services software, uh, right, to to kind of find a way to I guess put themselves in the middle of this token path that um it's been called or this kind of um you know AI value chain.

So it's interesting to hear from Tomash and his kind of view on the world in AI. Um you know, what are you seeing in our own portfolio portfolio, Paul? How are you looking at it? Yeah, it's uh it is very interesting just to see how quickly the intersection of the technology itself, so the performance and the benchmarking of the model companies and what's possible has very, very quickly entered the physical world.

So the the bottlenecks and I mean Jensen and Nvidia said they need a thousand decks, the the energy and power capabilities that they have today, if we're gonna continue scaling the inference and the usage of all these frontier model companies over the next, you know, uh two, three, four, five, ten years. Um all of this is is based around for the most part a text file. Like if you look, and I think Tomas talks a little bit about it in the episode, like image and video being another frontier, you know, a single image is a thousand to ten thousand larger than a text file uh from just a memory and and compute perspective.

And so when we bring robotics and computer vision into it, and we have a few companies in the portfolio that are, you know, software, hardware, computer vision model on the back end, really taking the physical world and and uh building you know software, hardware or AI hardware products uh that take a lot of the value that's being created at the frontier labs and brings it to areas where it wouldn't normally be possible. But you're also talking about data that's being accrued, training of robotics, where if we think our our data infrastructure or some of the infrastructure comp constraints of today are a challenge.

I mean it i it it's only accelerating. And I I think in a world where the revisions continue to go upward and the moment that compute is brought online or memory is brought online for any of these really fast growing, whether it's an application or an infrastructure company, um it it it just continues to be a signal that there has to be more build out, and that build-out ends up waterfalling far deeper than most people expect into different corners of the economy, to the point where, you know, you and I were talking about some of the, you know, caterpillar in some of the you know heavy industrial equipment companies that are growing rapidly just because if you're gonna be doing a lot of this data center data center build out, if you're gonna be doing a lot of this physical world infrastructure build out, well someone has to do that as well.

Um it's it's amazing how far the dominoes uh continue to fall and just how deeply it impacts so many corners of the economy. Yeah, I mean he had um some really good sound bites on the podcast. I'll let everybody take a listen, but I know we talked about um you know the next frontier in image and video data and how single image is uh a thousand to ten thousand times larger than a text file, and that you're gonna start seeing these kind of constraints, but um the data infrastructure will look like the constraints around this will look trivial uh compared to kind of what's going to exist and what exists today.

And I think we saw NVIDIA uh had portion of their revenue on their data center side of their business was flat for uh in you know up until 2022, and then GPT launched and three years later 17x growth. So I think you're starting to see kind of the infrastructure lag a little bit but catch up quickly, and then you know we'll come out the other side once a lot of this is built out in a way where it's like, wow, our minds are completely blown about what's possible. Another thing that Tomas talks about um which I thought would be interesting to touch on was just how from a first principles perspective, how differently you will build your organization today and and every company, but I think you'll see this at the forefront at startups and technology startups because they're you know deploying this, I think, on the cutting edge of what's possible.

Um but so many of our assumptions about how many people should be in control, how an organization should be structured. I think it's easy to look at the near-term implication of hiring freezes in certain back office categories or even layoffs, which are never easy in certain categories of the economy because you know, the same capacity can be met and exceeded if we're layering AI and technology into those organizations. But what we're seeing in in the startup ecosystem is a multiple layers deeper than that, which is not just, hey, what's the organizational structure that I can make more efficient, but how can I completely re-or-architect what a business looks like, how many people I need, what things can be done um from an AI native perspective, and then what things need to be done with people.

Yeah. Yeah, they were saying um Benny Off was saying Salesforce is spending $300 million on anthropic, and it's like, okay, when you do the math, they did the math on the uh 20 DC episode with Lemkin and Rory, and it was like, oh, it's only you know an incremental like 15, 20 grand per head. And if you're getting so much benefit out of that incremental spend per headcount, and then you don't have to hire humans to do more work, you probably are saving a lot of money. Like the spend sounds huge, but then when you actually like frame it like that, well, how much does a headcount cost, fully burdened at Salesforce, you know, from 150k to 500k, let's call it, for individual contributors, it's it's not that crazy.

And then you're also starting to see things. I mean, I think Tomas talked about agent-to-agent GTM. And you know, we haven't seen exactly that yet, but you know, we're investors in Avara, uh digital twins companies, we're investors in in um a lot of the kind of infrastructure layer here. Tavis is another one.

But it'll it'll be interesting to see when we're at a point where I don't think it I don't think buying and selling fully goes agent to agent, but it can certainly do the kind of fact-finding uh base case stuff. And we had uh Orin Hoffman on the pod that was talking about it on the VC side of the business, which is like qualifying uh a deal and an investor, um agent to agent before you know the the investor and the uh and the investee end up uh and the company end up talking for the first time.

Yeah, it's it's interesting, and I think maybe that's the first thing that ends up happening is if you want to call it agentic marketing or marketing to agents, um it's not that and maybe that will be the future, you know, if we start playing this a few hands in front of where it is today of agents fully completing a number of transactions from end-to-end, my agent, your agent, and a commercial transaction gets executed, it gets, you know, it discovers, it qualifies, it buys, and implements.

Um for now, I I I it's very curious to see how teams start doing this sort of parallel discovery or marketing. So um you know, there's a human decision maker, but then there's an AI agent that's researching vendors before any human call ends up happening. How do you package information to be more compelling for those agents? How do you find, you know, how is your discoverability improved?

And I know we we've been long-term investors in Noble and have been super excited about with what they're building uh you know underneath that umbrella. But just the degree where marketing teams and go-to-market teams are starting to build a parallel tract of you know discoverability for humans and then discoverability and information transfer for agents, as more and more of that initial like qualification and discovery ends up happening agent to agent and not you know human to agent or human to human.

All right, so without further delay, today's episode Tomash Tungus, general partner at Theory Ventures. Excited to have him on. Let's get to it. Tomash, welcome to GTune now.

Pleasure to be here. Thanks for having me so. Thank you for having me in this beautiful room that we were just talking about. Actually, it was named after Waterfall.

Can feel the vibes in here. It's great. Now, want to dive in because you've been talking about the decade of data for years, far beyond the AI wave. And now everything is catching up to what you've been talking about for a while.

And when you look at what Meta and Google are committing in terms of AI infrastructure spend, it's astronomical. Does anything surprise you about that? Like, did you anticipate this scale? I don't think anybody really appreciated the scale.

I mean, to compare the percentage of spend of data centers to US GDP relative to other projects, it's uh it will be this year the fifth largest infrastructure project ever, including the two world wars. Wow. Yeah. Um and so the next what rung up is railroads, and that's at about 5% of US GDP data centers today at 3.

5% of GDP. Um, and there's some signs that there's a little bit of weakness. I mean, you see Stargate, the team leaving OpenAI, you see some cancellation of the projects. But then on the other hand, you have uh Manthropic leasing 20 billion from Google and others, and Core Weave just announcing some pretty significant commitments.

So I think it'll continue to go up. You could I think you could see it three, four, five, maybe six, seven percent of US GDP. When would you anticipate that happening? By 2030.

The demand for inference is infinite. And then you look at the mythos model and uh you know, rumors are it's 10 trillion parameters in size, so it's five to ten times the size of the largest model deployed today. You need big machines to run them. And we're not even talking about video and images.

Yeah. Right? Sora was canceled. So I I think the scale is gargantuan.

I don't think anybody really appreciated how how much data we're processing. Yeah. Well, we'll sit down in 2030 again and and revisit this and see where we are there. Maybe in a data center.

Yeah, that's true, true. And you've broken down, you know, for for every kind of dollar hyperscalers make from AI, they're actually spending $12 on infrastructure. That's about a $575 billion bet. Yes.

How does this pie up? Well, it's okay, so in the short term, it is a market share capture game. Yeah. Who wins the most share?

Who's number one? And in the longer term, it's a margin gain, who makes the most profit. And today it's a big game of chicken because you see the most profitable companies in the world like Google, you know, they were generating $75 to $90 billion in free cash flow per year, and they're taking all of that and then borrowing to be able to fund out to data center CapEx. Meta's doing the same thing.

Oracle's levered seven to one on a cash flow basis. It's crazy. Uh and so people are really betting that they can win significant share. Over time, the dominant metric that really matters is how much intelligence can you drive per watt of electricity?

And that's increasing enormously and still has a tremendous amount to gain. So it characterized the first wave, and we're definitely in that first wave, is just how much share can Gemini win, how much share can OpenAI win, how much share can anthropic win, and then and then drive to profitability. And the rumors are that Anthropic is very high, gross margins, and some others maybe not so much. So we're starting to see bits of that, but uh but that'll be the sequence.

Very cool. And now with we think about how that implicates investing. You know, you've invested in data companies for for many years, you know, Demio and Monte Carlo and Hex. Has this changed how you're viewing investing in data companies?

Well, so historically, yes. The answer is yes. Historically, the data stack was separate from everything else. It was a separate world.

And you had AI and um it was called NLP in that era and classical machine learning models. And they they were largely separate. The data stack was built for people who wanted dashboards and analyses and uh just to understand how to operationalize the business. And AI was or NLP was mostly in research.

And I'm just drawing with broad brushstrokes, but yeah, it was used in ad targeting, which is where I was first exposed to it. And it was used in um sentiment analysis for surveys and those kinds of things. Now all of a sudden AI and data have fused because as we just talked about, AI is driven by huge volumes of data. So all the pipelines that we built for dashboards and analyses are now being used to train and run these machine learning models.

And so they fused in a very big and real way. Now they're the same thing. You can see that organizationally, where data teams are starting to report to the head of engineering. That's a really big change.

But overall, just like the scale of the data process. I mean, you um maybe make this concrete. So it um at a recent event, Jensen showed the latest um InfiniBand networking equipment, which is the equipment that connects GPUs in the data center. And those InfiniBands, they can transfer the entire size of the internet in less than a day.

Wow. Right. And so um you just have huge volumes of data. And so as a result, both of these ecosystems are fusing.

Fascinating. And when you actually are assessing startups and with that lens and understanding of the fusing, what do founders need to know about their own positioning? I think, okay, so there are a couple of different things. The first is uh the data world is fusing with the AI world.

That's very real. The second is um the no one really knows what the future looks like. We can all pretend that we know and predict, but if you're a buyer of software today, you are looking for a trusted partner who you believe will be the person to guide you through the future for the next three to five years. So you're looking for someone to develop, say, three to five to ten to fifty agents to solve your needs.

And that's true for head of engineering, head of data, head of sales, head of marketing. And so that's the sale that you need to make. It's not a point solution like it was in software or it was in the excess in the time of data. It's really about that trust for the future.

Yeah, incredible advice. Our LP base spans from individual operators to institutional allocators, and Angelist has been instrumental in supporting all of them. They handle everything from investor onboarding and accreditation to distribution and tax documentation, creating a seamless experience across geographies and fund types. Plus, all of this is available on a single modern platform.

For an LP base like ours with over 300 C-suite and VP level operators, this kind of white glove service and seamless workflows is so important. It's also instrumental that we support our institutional LPs that we're fortunate to work with, and Angelus is able to do so every step of the way. If you're looking for a platform that can support any type of LP investing in your fund, learn more at angelist.com slash GTM fund.

And then we think about the journey. You know, in October of 2025, you wrote a piece about how product market fit is no longer static. It used to be this binary thing. Yeah.

How should founders and operators and leaders be thinking about product market fit now? It's continuous. So, you know, in the era from 2010 to say like 2021, you would have product market fit, you'd establish a product, and then you would scale. And you'd have all the economics, all the ratios were known.

It's just a question of raising the capital and executing. Today, that product market fit. Okay, let's take the foundation model companies. Yeah.

A foundation model company will develop a state-of-the-art model. They have 35 days to commercialize it before somebody else competes them. Yeah. So it's not very long at all.

Oh, especially for uh you know, five or ten billion dollar investment. Yeah. The same is true for software. You if you develop something unique, it's very easily copy.

And so you have to keep pushing and you have to keep pushing, which is what we mean when we say that product market fit is continuous. You can have to continue, and then the buyer demands are also changing because the buyers are starting to understand what they want. Yes. And when we talk about the buyer needs, you know, recent things that you've been talking about are um, you know, how AI implicates sales quotas or marketing too and advertising.

You know, what what are some of those largest implications that you're seeing right now? Well, the in the marketing world, the first thing that's changed is um we were interviewing a woman named Lena Waters. So she was CMO at Notion, Grammarly, DocuSign, and she says the buying journey is very different. The buyer educates themselves much more using agents than they ever have in the past.

And so that's a big change. This as a result, you have now have two different constituencies to which you need to market. The first is the person, let's say the head of engineering, and then the second is the agent of the head of engineering, because the head of engineering will consult the agent before they ever pick up the phone and call the vendor. Um and then that agent is now part of a new buying committee within enterprises.

So if it's the head of engineering, there might be the head of AI, the head of legal, and then there's probably an agent also involved in that purchasing process. Now all of a sudden you have a different dynamic across those three or four people, and you need to figure out, particularly if you're selling large deals, how to navigate successfully through that buying committee. Yeah, it feels Like it's never been more layered before is almost a pipeline of the first layer being the agentic process and then the second being more the emotional human component.

And maybe it's not linear like that, but two different layers that get activated at different times. Yeah, two different personas. So that means the website. Oh, I was at a conference called Human X earlier this week.

And the head of uh Carta was saying they're no longer investing in their website or their mobile app. Wow. And no new product development. Yeah.

It'll all be for agents. Yeah. Yeah. Yeah.

I'm curious about that because I've had quite extensive conversations with, you know, uh this Linda, the CEO of Webflow and folks that are really doubling down on website space, but transforming it into a little bit more of like a revenue source and a truth for agents. Yes, that's where it's going. It's just transformative where we may not visit them, but they still serve a purpose and role, at least from their purview. Right.

Well, and then there's a question do you care about visualization or not? Yeah. Right. There are lots of marketing or positioning campaigns that appeal to human emotions as a way of engendering trust.

Agents don't respond to emotion, at least not yet. No. So how do you appeal to an agent? Yeah.

How do you think you appeal to an agent? Right. Now it's just pure text. You just want raw and markdown and statements of facts and clarity.

Yeah. Authoritative content. Okay. Interesting.

And you know, one thing I think everyone in the investor and operator and founder community really appreciates about all of your work and writing is that you emphasize both the need for technical innovation, but also go-to-market strategy, like we've been talking about. What are you seeing specifically in the data space around patterns and go to market right now at this inflection point in time? I think everyone's trying to understand what the implications are for agents. Agents are in your distribution channel.

So how can you leverage cloud skills or be involved in the decision process when an agent says, oh, we need to use this database, or we need to use that database? That's really important. I think the second is the re-imagination of the pricing model. So it used to be C-based.

Snowflake and Databricks clearly have consumption businesses, uh, but inference is a one or two orders of magnitude larger than say data warehouse compute as a market. So figuring out what is your pricing structure to drive exposure to inference growth is really critical. And then the last is just scale, which we talked about before. How do I position my product or technology to be able to handle the volumes like very, very, very large scale?

Mm-hmm. Yes, absolutely. And we talked about the implications of AI and different use cases, and you've been very vocal about Vercel, for example, replacing nine or so of their SDRs with one AI agent and a part-time engineer. That's one of many go-to-market use cases around how AI is impacting it.

What else are you seeing on the impacting the team perspective? Yeah, so in sales, the there's a transformation of the SDR and the BD BDR roles where there's big drive for full automation of those roles. And I think that's real and important. And it's a one-way change.

Other dynamic that's really important. Okay, so if you were to ask AI uh what is uh the best pair of running shoes to buy, you'd probably trust that recommendation. Yeah. If you were to ask AI what is the best laptop to buy, you'd probably trust that recommendation.

You may even go as far as what is the best car I should buy. And so you might, you know, uh $20,000 to $30,000 buying decision, you may outsource AI. Within the world of enterprise sales is much more complex. And uh so I think the need for humans to engender trust between each other uh will persist.

Although we just met a company that is building agents that actually influence you. Interesting. And can change your decisions. How so?

Well, there's a uh a benchmark, it's called giving for good. Okay. And there's an AI agent that engages with you and it tries to convince you to donate money to a charity. And so there are different AI systems that have different techniques and they are scored on what propensity does the person have to donate and what amount.

Based on historic data. No, no, just you know, you're typing. Yeah. And you're saying, tell me more about this charity.

Right. And so it tells you about the charity. Well, what do you care about? Well, I care about these particular things in the in the way that I donate my uh money.

And then it says, Well, you should really consider this one, and this is aligned, the more effective it is at convincing you. Well, you can take that dynamic and say, I really think you should use Omni as a BI platform. Right. Right.

And so at what point, like, is that a good thing? Is it a bad thing? Is it ethical? Is it non-ethical?

How do you use it? And so I think all of that will happen. In the marketing world, there's we're seeing tremendous automation of creative. So images and video, you're seeing the use of reinforcement learning with an ad targeting.

Meta's published many papers there. The effectiveness is quite significant. Um, and I think uh maybe just to bring it up one level, you have the reimagination of almost every role. And so many of the most forward-thinking leaders hired generalists as opposed to specialists.

And you're hiring yourself. And you're hiring in a very interesting capacity, very uh uh very much so on the technical side. Tell us a little bit more about how you're thinking about your own team composition and hiring as it pertains to AI. Half of our team are AI engineers, and I think that will be the case for a very long time.

Uh I think the leverage that many others are able to drive from AI should also come and will inevitably come to venture capital. And uh we'll begin to take or be part of that way. Yeah, exactly. I think we always talk about how you're an operator, you're operating business, just like a software company.

I mean, we actually have operator background. So I feel like inevitably, but a lot of the time in venture, you know, we talk about it in a different capacity, but it really is the same thing. And so if you're not adopting AI, if you're not taking the same kind of steps that software companies are forced to to compete, like you will inevitably hit the hit the ceiling where others are advancing. So you are are certainly kind of paving the way on that front.

And it's been incredible to see the developments of AI, and you've been very vocal about sharing those developments too, even around the way that you ingest podcasts and like personal use cases. So I'm curious, like, what are some of your favorite, most transformative use cases personally on AI beyond the firm specifically as a whole? Oh, uh personally, I guess it worked. And I know you have a lot.

Yeah. In work too, but less operationally at the firm level, more about you know, yourself as an investor. Yeah, I think um I think the most impactful use of AI or one of them is you always have to coming out of a meeting or in a meeting, you have a question about something that you would never normally have the time to answer. Yeah.

And so AI is great. It's like, oh, someone t told me yesterday about a book. Uh it's a book about the psychology of playing tennis and that it was excellent. And so, okay, great.

Like, I'm not going to have time to read that book, but I'll ask an AI to summarize that book and tell me how it can apply to mention capital. And so that's it broadens your knowledge in that way. So I think that's really powerful. Very cool.

And then at the firmwide level, just to go back to that, because you are investing so heavily in AI engineers, what are those AI engineers building? What are you doing at the firm level? Yeah, we are um I mean I think one of the key things is we're really trying to understand uh how these systems work, which informs our investment DCs. We do have a lot of fun.

So there was an event last week, Funeral for MCP, which is a technology. That's great. The technology uh anyway, it's gone back and forth on whether or not it will be a dominant technology. And so we we uh created a thing called RipGrep, which allows you to figure out like which technologies seem like they're dying, and everybody says they're dying, but they're not actually dying, which would be true in the case of MCPs.

We definitely have a lot of fun too. Okay. So it sounds like you're saying MCPs are not dying. No, no, no.

I think uh MCPs, so there's a role of calling a software directly through an API. Yes. And that's really useful in some circumstances. The benefit of MCPs, at least the way that we understand them, is in a large company, if you want to offer all the finance team a particular set of capabilities, that's the easiest way to distribute it to them and control it.

A place and a time for both. It sounds like it's not mutually exclusive. Brilliant. And then when you think about the landscape right now, like if you were starting a data company today, what would you be excited about building?

Images and video. Okay. Yeah, the data volumes there are so enormous. We're lucky to work with a company called LandsDB that's in that space.

Um, and you can just see, I mean, uh, you know, an image is probably a thousand times to ten thousand times bigger than a text file and a video two or three orders of magnitude larger than that. Well, if we're already struggling to move text around the way that we are, and we know that custom video is coming, we know that robotics is coming in a very big way. Well, we need much bigger infrastructure to be able to support those demands. Yeah, great.

Okay, well, it's a fun space. It's never been a more exciting time to build. That's right. And prior to this time right now, you know, you back some incredible companies like Customer and others that have had fantastic outcomes.

Are there any kind of patterns that you've seen across your, you know, visibly successful companies? I know you've got a ton of successful companies in the portfolio that are probably less visible in terms of outcome yet. What are the patterns that you think makes that successful that founders can apply for themselves? Understanding the history of a space is underappreciated.

That's really important. So deep understanding of the space. The reason second-time founders are so successful, especially the ones who start a business in the same domain as their first company, is they know. They know the people, they understand the history.

They've just learned about it. And so that level of specialization is incredibly powerful. Today with AI, you can understand a lot about your core domain. So that's one key ingredient.

And I can imagine the connections too is a big part of that. And the distribution if you built in the same space, because now it like in very much the way is at least what we're seeing around the go-to-market side, is people are leaning into ecosystems and partnerships more than than ever before. And that level of connectivity is a really unfair advantage. So I can imagine how that would apply also to second time vouchers.

Yeah, no, that's true. And then I I think the other change, I mean, one of the big changes has been PR is much more significant of a distribution channel that it has been the press is willing to write about AI in a way that they maybe, and I mean the mass media, uh, they weren't willing to write about in software. And then maybe the last is uh much an accelerated use of channel. So historically, channel was something that you might engage with outside of security, you might engage with like a 15 to 25, maybe 30 million in ARR.

But today we're seeking channel partners and min, even low single-digit ARR. Wow. Are there any kind of contrary views that you have or hot takes, if you will, about the AI space right now? I think the impact is still broadly understated.

It'll be so transformational. And I think we'll see it in organizational design. Companies today are structured in the way that they were for a time before the computer was invented. It's kind of wild.

It is wild. You know, like the idea of a product manager is maybe 40 years old. Yeah. Uh and I think AI will completely transform the way organ companies are structured.

Uh you can look at like uh if you think about cutting up a company with executive leadership, middle management, and then doers or individual contributors, the ratio is probably like 5%, 75%, 20%. In five years' time, it won't look anything like that. It's wild. It's wild to transform into properties.

And I mean, you've written about two the implications of pricing and pricing for AI agents. So it's just tremendous. It'll be very exciting to see. Any last kind of messages or advice to founders broadly?

I I think the only advice I'd have is nobody really knows the answer to anything. Yeah. And so we're we're in the tremendous period of experimentation. Yeah.

The best thing that you can do is just jump in with two feet and figure it out yourself. That is fantastic advice. Now, you've got an incredible blog or writing space. People can follow along with your writing.

Where else can people follow along if they want to keep in touch with you? Yes, uh, we're on Twitter, we're on LinkedIn. We just started um a series called uh Office Hours, where we host sessions with um executives. And uh the great part about that is you can dial in and ask a question and we'll leave it into the conversation.

Yeah, I really enjoyed your honor, Carol, personally. That was great. Well, thank you for joining us. It's been fantastic.

Appreciate the time to watch. Oh, a pleasure is mine.