🎧🍌 Inside the AI Sprint, Understanding Anthropic's Strategy | Tomasz Tunguz, Theory Ventures
How Anthropic is commoditizing its compliments, why AI models will resemble pharma more than software, the three layers of AI business models, and where to invest in AI today
This latest episode of The Peel dissects Anthropic’s strategy, and the “all out sprint” happening right now in AI.
We talk through how it compares to prior technologies, how companies are actually buying AI products today, the three layers of AI business models, where to build and invest in AI today, and what Theory looks for in new investments.
Support this Episode’s Sponsors
Numeral: The end-to-end platform for sales tax and compliance.
Flex: The all-in-one bank for business owners. Apply here.
Amplitude: AI analytics. All you have to do is ask.
Inquire about sponsoring future episodes here.
Timestamps to jump in:
0:42 The “all out sprint” in AI today
1:40 Why GPU prices are up 116% in six weeks
6:34 AI infra end-state: “We’ll over build”
9:12 Tokenmaxxing, and why AI needs to get more efficient
15:48 AI models will resemble pharma more than software
19:52 Why Anthropic still trades at a discount
25:42 Anthropic’s strategy: commoditize the compliments
30:29 Why OpenClaw is so strategic for OpenAI
34:08 The three layers of AI business models
38:18 Where to invest in AI today
45:49 Who will survive SaaSpocalypse?
52:15 Comparing AI’s impact to historical technology cycles
57:34 How new technology historically impacts jobs
1:05:58 Where AI is underrated today
1:10:41 How people are actually buying AI products
1:14:06 Why Theory’s investing in ads, inference, and email
1:16:24 2026 IPO pipeline, how VC has changed over 20 years
1:20:56 What Theory looks for in new investments
1:22:32 Starting Theory Ventures in 2022
1:25:39 Running a monte carlo analysis to determine portfolio construction
1:27:54 Tomasz personal AI projects
Referenced:
Tomasz Blog
Find Tomasz on X / Twitter and LinkedIn
👉 Stream on YouTube, Spotify, and Apple
Related Episodes
Transcript
Find transcripts of all prior episodes here.
Turner Novak:
Tomasz, welcome to the show.
Tomasz Tunguz:
Oh, pleasure to be here, Turner. Thanks for having me on.
Turner Novak:
It’s kind of funny. We had never met before last week, and then we were at Samil’s thing, and then we’re recording this podcast, and then next week we’re gonna be at the Beyond Summit. So it’s three times right in a row.
Tomasz Tunguz:
Three-peat. Let’s go.
Turner Novak:
Let’s do it. And really quick, for people who don’t know, what is Theory Ventures?
Tomasz Tunguz:
We are an early-stage AI-focused venture firm. We invest anywhere from one to $45 million, typically in B2B software and infrastructure companies.
Turner Novak:
How would you summarize the state of AI today? A little bit of an open-ended question, but how do you think about everything that’s going on?
Tomasz Tunguz:
All-out sprint. That’s the way it feels. Okay, so why do I say that? The first is there aren’t enough GPUs for anybody, so people are sprinting to buy GPUs or rent them.
The second thing is model improvements. A model only remains state-of-the-art for about 41 days, even though it’s several hundred million or a billion to train, maybe less.
And then there’s also an all-out sprint for customer acquisition. Buyers are the most open they’ve ever been to trying new things, and so if you can capture many of them, you’ll have a big business. And then the businesses themselves are growing at unprecedented rates. So I think everybody is sprinting.
Turner Novak:
Maybe the first one you mentioned, the GPU prices, what does that even mean? For somebody who is not super familiar with any of this, how would you just explain that to a smart person who is hearing this for the first time?
Tomasz Tunguz:
To run a machine learning or an AI model, you need a GPU, which is a particular kind of chip that does lots of calculations in parallel at the same time. Matrix math, it’s called. If you have a MacBook, you have one. In fact, you have an excellent one, maybe one of the best that you can buy for your computer. And you can run small models, which are effective there.
But if you’re a company that offers an AI product, you can’t just buy a whole bunch of MacBooks. You need to buy servers that have lots of these GPUs in them, and those GPUs are hundreds of thousands of dollars most of the time. And the prices increase every week because there aren’t enough of them. So you can either buy them and run them in your own data center, or you can rent them from other people.
And there aren’t enough because, one, the models that we’re building are much bigger than we thought. They’re now trillions of parameters. The demand for those models is much bigger than we thought, primarily because of things like OpenCode and agentic tool calling and coding. And there’s not enough memory, there aren’t enough CPUs, which is really important. And that’s mainly because most of these chips are produced by a company in Taiwan called Taiwan Semiconductor Manufacturing Company, or TSMC.
Turner Novak:
Very creative name.
Tomasz Tunguz:
Yeah. But it’s coming back, like Nabisco. Do you know Nabisco is an acronym?
Turner Novak:
No.
Tomasz Tunguz:
National Biscuit Company.
Turner Novak:
Really? Okay.
Tomasz Tunguz:
Yeah. And there was this wave of American Motors, right? I remember I met this gaming company, it’s a total tangent, but they were called the Brooklyn Packet Company. It’s like, that is an awesome name. Anyway, we’re starting to see some of these locality-based names come back again.
Turner Novak:
Well, it’s either that or you make up a word. We were at a point where people, their name would be like computify.io. You had to make something up to come up with a name.
Tomasz Tunguz:
Yeah. To buy the domain name at some reasonable price.
Turner Novak:
And now the same thing’s happening with GPUs. The prices are going up.
Tomasz Tunguz:
That’s right. And then the other dynamic is just power and land. Can you find a place to build a data center? It takes three to five years, maybe seven years to build a new power plant or to buy a jet turbine to power your data center. And then you need to build a data center itself, which takes 18 to 24 months. So there’s all these lead times. Atoms finally are starting to be really important in the world of software.
Turner Novak:
A lot of people are making these projections, like, we need 10X more next year, and then the following year we need whatever the number is, 10X more from that. Can we even keep up? What do you think is gonna happen?
Tomasz Tunguz:
No, we won’t keep up, and we will overbuild. Okay, so CapEx spending, those are dollars that are spent to build out data centers. We’ll be at like $1.2 to $1.4 trillion this year, and out of US GDP, that’s about low to mid twos as a percentage basis.
So it’ll be one of the largest infrastructure projects ever. Right now there’s World War I, World War II, just in terms of large infrastructure projects. The railroads peaked at 7.7%, and then there was the Eisenhower development of the National Highway system, and I think this year we will exceed that. So the question is, will we beat 7.7% and get to around $2.1 to $2.4 trillion?
In a year or two? It’s definitely very possible. And you can see Google is outspending Microsoft in GPUs even though GCP is significantly smaller than Microsoft. That tells me that there’s some very sophisticated math to justify those build-outs, and as long as that math works, we will continue to build. And then the electric grid will have to be reimagined because it was never conceived to handle these kinds of volumes.
Turner Novak:
Do you ever reach a point, though, where this plateaus, like with the railroads and with the National Highway? We built the highway. It’s there.
Tomasz Tunguz:
Yeah. But the highways continue to grow, right? I don’t know, 101 in San Francisco, they keep adding lanes, and LA, I-5. But yes, we will keep going, and very likely overbuild because it’s impossible to determine when the economics either change or the demand changes.
Why would the demand change? Well, every year, I think Google, for the last two years has said they generate 80% more tokens per GPU hour than they did the year before, which is doubling productivity. So that’s a really big deal.
Then you have segmentation. So you can say, “I don’t really need a super fancy model to update my CRM. I can use a model that’s running on my computer.” So you could actually have a lot of the workloads going on your MacBook, and that will happen.
But for now, I don’t know what AI penetration is as a percentage of ultimate penetration, but I have to admit it’s less than 2%. I have to estimate it’s probably something like that, which means we can grow 50 to 100X from here.
Turner Novak:
So do you think what’s gonna happen then, it sounds like, is we’re gonna hit a wall in how much we can expand the infrastructure capacity, so we’ll have to just get more efficient essentially with what we have?
Tomasz Tunguz:
I think that will happen. You have this token maxing era. Token maxing is putting a leaderboard in your company and seeing how many tokens you can use, and that’s a lot of fun. I hit 250 million one day. It literally did everything through an AI. And after burning a couple thousand dollars, you’re like, “Okay, that was fun once.”
Turner Novak:
What did you do? You were telling me about that before, the token max. What did you make it do, and then what did you accomplish from spending the thousand dollars?
Tomasz Tunguz:
You can’t do it just by querying ChatGPT. There’s no way. You might get to a million tokens that way. Parallelization is absolutely essential, so you have to create a plan for what you want an AI to do that particular day. Coding is huge because it’s reading large existing code bases.
And then I think the best technique is anytime you thought to do something with a computer, do not start with a browser, do not start with your email client. Go to the AI and try to figure out how to do it with the AI. And then you can ramp.
The challenge now is many of the clouds will stop you from doing this. They’ll hit you with a rate limiting error of 429, or 529, and they’ll say, “Too much, too much for you.”
Turner Novak:
So this could be like you could say, “Go and read the entirety of Twitter and just give me a summary of the best tweets,” or like all of Reddit or something. Just give it an insane task.
Tomasz Tunguz:
Read all my emails, listen to these 50 podcasts, transcribe all of them, download these 10 GitHub repositories, install them and see if they work. Benchmark these four local models, which one’s faster. Download a bunch of startup presentations and analyze them, extract the data. All kinds of stuff.
Find for me the 10 most important academic papers in the last week. Anything that comes to mind. For the last two weeks, download all of the earnings transcripts of every public technology company, draft 10 blog posts, and pick a best one, and then critique it like an editor.
The harder part is not getting the AI to do it. The harder part is creating a workflow where you anticipate multiple steps and forks. And if you can do that effectively, you can go through a tremendous amount.
For those who are coders, there’s this beta feature within OpenAI’s Codex called /goal, where you just tell it, “This is what I want you to achieve,” and then it will just continue going. I was chatting with a friend who said he had his going for 18 hours and it worked. I did this on Wednesday. I was frustrated with a dictation app, and so I just told Codex, “Just replicate this app so it works.” And then, whatever, 45 minutes later it said, “Here you go.” That cost $15, but I’ll only pay that $15 once. And now you have this dictation app that you can use whenever.
Turner Novak:
Yeah. Dictation’s another way of driving a lot of tokens.
Tomasz Tunguz:
Yeah.
Turner Novak:
Interesting. Because I do feel like we’ve all seen those headlines at this point where Meta actually had to get rid of the leaderboard, I think, was one of the most recent things, because people were probably doing something similar where you’re gaming the leaderboard.
Tomasz Tunguz:
I mean, you can use jet fuel in your car and it’ll go faster.
Turner Novak:
Will it actually go faster?
Tomasz Tunguz:
Yeah, yeah. Racing car fuel, octane is the... So this is a fun one. Hydrocarbons form all of propellants, right? And there’s hexane, which is I think C6H6, and then there’s heptane, and then there’s octane. When you get there, there’s 87, 89, 91, and that’s the amount of octane that exists within gas.
And then if you get racing fuel, the octanes are much higher. They’re 91, 95, 105. So the amount of energy per unit volume is significantly higher, and your car will produce a lot more horsepower if you put racing car fuel in it because the explosion is stronger.
Turner Novak:
So you actually will go faster?
Tomasz Tunguz:
Yeah, at the top end. Sure. Provided that everything in your engine holds together. This is not an endorsement of putting nitrous into your Prius and seeing if you can break 200 miles an hour.
Turner Novak:
Yeah, but I guess it’s kind of the same thing with AI. You may have a super-powered model or you’re running all these different parallel tasks, but are you even doing them properly? And have you set things up right to make that even worth it?
Tomasz Tunguz:
Exactly. So let’s ground this in some numbers. A very small model might be a few hundred million to two, three, four billion parameter models. Those are great for dictation. They’re great for grammar cleanup, transcription, those kinds of things.
Then you have the next range of models, which I would put at the 25 to 35 billion parameter models. Almost anything that you can do with a computer aside from coding, you can now achieve with one of those models, and they will be faster on your laptop than they will be talking to Claude or OpenAI. They’re just faster at it.
And then there’s another class of models that’s like 120 to 150 billion. Those are not that often used. And then you have the state-of-the-art models, which are trillions of parameters, and they can do architecture and implementation of very sophisticated code or novel math discoveries.
Turner Novak:
So basically, I think I’ve seen, if you follow some of Dario and Anthropic’s positioning, it’s like it costs a ton to train these models, and as the revenue starts ramping, you start getting profitable on these models, the older models. But then they’re training the new ones, which are even more expensive to make, which makes it so it looks like they’re losing money, but the revenue gets even bigger. And it’s like these stacking super expensive to train models work way better. So eventually we get to a point where they just start making a ton of money. Is that kind of how this is gonna go?
Tomasz Tunguz:
I think it resembles pharmaceuticals more than it resembles software. You might spend three years researching a drug, and then I’m not deep in pharma, but I think you have 17 or 20 years with an exclusive patent on whatever, the next statin to reduce cholesterol.
But with AI models, like we said, you have 41 days to be state-of-the-art. And you can see it. There’s a company called OpenRouter, which is an open source router of model calls. You can see the share shifting pretty significantly. In November of last year, Grok, which is the xAI model, had pretty significant share, above 15 percentage points. Today, it has a lot less.
And then you can see the share shift as a result of the subsidies from OpenAI or Anthropic on their different products or new models. Whatever, GPT-5, 6. So you don’t have 17 years to recoup your investment costs. You have to keep running faster.
The other dynamic that’s really important is as these models and the training data become larger and larger, there’s a great paper that talks about how the model performance will ultimately converge, and we’re seeing this. At the beginning, you could see GPT-4 was significantly better at agentic tool calling, and then I don’t remember exactly what the Claude model was that caught up. And then Gemini was really strong in this particular domain. Maybe it was math. And another one was great at humanity’s last exam, which is a knowledge retrieval benchmark. And now they’ve all added more and more benchmarks, and they’re all more complete, and the differences between them are increasingly subtle.
Turner Novak:
And so ultimately then the advantage is just, do you have people using it? It doesn’t even matter how good the model is because they’re all the same, and it’s more so is it the behavior, or have you captured the workflow in some capacity?
Tomasz Tunguz:
Yeah. I think we’re gonna get to a place where you reach a minimum viable intelligence, where if you work at any company with a computer, there’s, whatever, I don’t know what it will be, but some minimum, let’s say it’s a 30 billion parameter model in late 2026. And if you have a computer that you can run it on, that’s good enough.
It’s just like, we’re not giving everybody inside of a large company a state-of-the-art laptop. You have an IBM PC that’s pretty good. You might have a MacBook Air that’s pretty good. But you don’t have an M3 Ultra with 512 gigs of RAM for everybody. You have a minimum level of performance that’s good enough, and then you kind of upgrade every two years or three years, depending on your company’s policy.
You can imagine we get to a very similar place with models where you say, “Okay, I have the current Gemma model from Google. It’s 31 billion parameters, and I can do most of my things on my laptop, and that’s fine for me.” And then the frontier models push into the domains of high-performance computing, math research, materials research, chemistry, and really pushing PhD-level analysis further and further. You have some companies, Dow, Corning, pharmaceutical companies, who are willing to pay a huge premium for that, but everybody else will use a mid-range model.
Turner Novak:
You had a post recently, AI at Discount is the name of the post, and the premise is Anthropic, if you actually look at it, it looks like it’s actually trading at quite a bit of a discount considering how fast it’s growing. But on the other side, everyone’s like, “Oh, these AI companies are so overvalued.” So what is actually going on with how these companies are being valued by the markets?
Tomasz Tunguz:
Okay. When a company is growing really fast in the public markets, many people, not everyone, but many people value it on a forward revenue multiple basis, which is a fancy word to say estimate the revenue in the next 12 months, and then you take the market cap and divide it by that estimate of the revenue growth. It’s called an EV to forward revenue multiple.
Most software companies, and you can benchmark, the fastest growing software company today at scale, aside from a pure model company, is Palantir, and they’re growing at 68%, which is mind-blowing.
Turner Novak:
That’s pretty good for their size.
Tomasz Tunguz:
Yeah. That’s like a mid-size venture scale business five years ago. But this is a publicly traded, billion-dollar-plus revenue business growing at 80%. And they trade at a very elevated multiple.
And then if you look at Anthropic, Anthropic year over year grew 30X. Now it’s closer to 43X. They went from a billion in run rate to $43 billion in the year. That’s just absurd.
Turner Novak:
It breaks all laws of business ever. It’s just impossible for that to happen. You’d think they were committing fraud or scam or it’s fake.
Tomasz Tunguz:
Yeah. They added, in the month of April, they added all of Snowflake’s revenue plus all of Palantir’s revenue in a month. Just monstrous.
So anyway, let’s say they grew 43X. Okay. What do you think they’ll do next? It’s almost absurd. What do you think they’ll do next year? But even if they’re at 43 and then they get to 100 and they’re valued at 900, well, they’re kind of valued around single digit forward revenue multiples.
And then you look at Palantir, and it’s valued at 30X, 35X. So Anthropic’s actually trading at a discount, which is kind of wild because the growth rate’s 80% versus 4,300%.
Turner Novak:
And is that like, does the market not expect Anthropic to continue to grow that fast? Is it just people saying, “Okay, this is not sustainable. It’s still growing really fast, but we’re gonna assume that this slows down”? Or is it like a private market thing? It’s just because it’s harder to get access to it, and technically Anthropic can just price it whatever they want, really. Is it just not a fair price that’s just out of whack?
Tomasz Tunguz:
One, it’s very difficult to project forward revenue. The revenue is non-recurring. Some of it is contracted, but it’s unclear. The third part of it is at some point, the revenue growth will be limited by just total amount of GPUs.
Anthropic and SpaceX AI signed an agreement so that 25% of the Colossus data center, which is focused on training, will now be allocated to Anthropic. But at some point there just aren’t enough GPUs. So what happens to a business that’s growing 43X in a year that starts to grow at, say, 30%, which is still, a $43 billion revenue base. You’re talking about adding $12 to $15 billion of revenue a year.
Turner Novak:
It’s like they just added that in a month, and now they’re gonna add it in a year. It’s almost unrealistic to think it’s gonna slow down that much.
Tomasz Tunguz:
Right. But you don’t, so what are you underwriting? What do you think it’ll be? I don’t know. And then there’s also the capital intensity. You need to build out these data centers. Do they need to raise debt? What does that look like? How much dilution are you taking as an investor? So there’s a lot of unknowns.
And then there’s this trope, trees only grow so big. Have you ever heard that?
Turner Novak:
Uh, no, but it makes me think of like the law of large numbers, or just like this company could never get that big. If you look at textbooks, they’d say there’s no, it’s like the law of physics says you cannot go from one to $43 billion in a year. It’s just impossible.
Tomasz Tunguz:
Right. I remember when we had the first, I was growing up and, anyway, I remember when we had the first trillion market cap company, and that seemed staggering. And now we have four companies that are around the three, four trillion, maybe five.
One interesting question to ask, this’ll be fun with you, Turner, is when do we have the first $10 trillion company? It’s, I don’t know. It’s definitely within our lifetimes. Is it 2030? Is it 2035? You have the devaluation of the dollar, and then clearly these companies are growing really fast. From my perspective, it’s inevitable.
So will Anthropic be the first $10 trillion company? It’s kind of hard to imagine, right? Who’s going to take the other side of that bet? I don’t know.
Turner Novak:
Yeah. Well, especially when you consider two years ago, they arguably had no business. There wasn’t, the thing that exists right now is just not there, right? And now it’s suddenly the fastest growing of all time.
But I think the interesting thing, you also wrote about this publicly recently, the strategy that they’re taking is similar to what Google did, where you’re commoditizing the complements, I think is how you describe it. How do you think then about the strategy that Anthropic’s taken with all the products?
Tomasz Tunguz:
Yeah. So Jason, I think it was Jason from a Smart Bear, wrote this blog post in the early 2000s called Commoditizing the Complements. The idea is if you have a really good business, what you want to do is look at all the people who have businesses around you and make all of those products free so that more people end up using your product. That’s called commoditizing the complement. You commoditize everything that’s complementary to you.
Let’s make this concrete. If you’re Google and you make money when people click on search ads, you want to make it so that people click on as many ads as possible. And I was at Google from ‘05 to ‘08, so I saw a little bit of this from the outside.
What did they make free? Well, it used to be you paid for email. Okay, email was free. And then it used to be that you paid for video hosting because video hosting was really expensive. But then they bought YouTube and made that free. And it used to be that you would pay a license to have an operating system on a mobile phone. Then they bought Android, and then they made that free.
And then it used to be that you would buy a dedicated GPS device for you to navigate your car from one place to another, and then they ended up buying Keyhole and making Google Maps and Google Earth free. And then they bought all these books and chopped them up and scanned all of them and put them in the index. So it was just driving more and more searches.
Google Docs, same thing. So you’re just using the internet more. By virtue of the fact that you’re using the internet more, and it was free, so there was less friction, you would go to Google more, and then you would get more ads.
So if you’re Anthropic, you can run a very similar strategy. Anthropic, you are selling inference. You are selling a prediction of an AI system. And then what you wanna do is, well, there was all this workflow software the previous decade. Maybe it’s legal software or finance software or accounting. I’m just picking categories at random here. But you don’t really wanna charge per seat anymore.
That’s silly because the amount of money people will pay per seat, maybe it’s $500 a seat per month, compared to the amount of inference they’ll buy at $2,000 a month. Just give away the $500 seat and have them buy more inference. You’ll make a whole lot more money, and then you have less competition.
I don’t have any, I’m just observing from the outside, but that’s a very game theoretical optimal way of maximizing when you have a really phenomenal business. You just wanna make sure everything else is free, so there are as many queries as possible.
Turner Novak:
Yeah. And then why does that become so important than paying for the inference that you mentioned? What does that even mean for somebody who doesn’t know what inference is?
Tomasz Tunguz:
Inference is when you ask AI a question or the AI does something for you.
Turner Novak:
It’s the process of them doing the retrieval and doing whatever they do with the GPUs that they then give to you, essentially.
Tomasz Tunguz:
Yeah. That’s right. All these systems are basically word prediction machines. So when you ask, “What is the capital of Italy?” it’s then creating a sentence where it’s predicting, and Anthropic and the other companies charge by the word. It’s called a token, but it’s really effectively by the word.
The longer the answer or the greater the amount of information you give the model, the more expensive the query. So if you have a really large code base, or if you have a really large legal case, or if you have lots of PDFs, and you want the AI system to analyze it, that’s a very expensive query. Because it turns out that the input tokens, or the data that you give the model, is around 90% of the overall cost of asking that question most of the time, or more, 90 to 95%.
So anyway, inference is what the model is predicting to answer your question. If I can just get the system to ask more questions. And it’s not somebody sitting there and typing and asking about a particular case. It’s, let me create a workflow.
So to analyze a startup, let’s say it’s like, okay, find the backgrounds of the founders, create a bottoms-up sizing of the market map, help me understand the backgrounds of the team, compare this to other companies. And then all of a sudden, the tokens that you use, the amount of information you’re feeding to the model, the number of words that you’re analyzing, predicting, explodes.
Turner Novak:
So really Anthropic’s business model and their strategy is just get people to do as much as possible in Anthropic products, just use it for things.
Tomasz Tunguz:
Yes. And this is why OpenCode is so strategic. OpenCode is a little assistant that lives on your computer, and you can create a task list for an AI. You can say, “Find for me the best place to visit in Italy. Go and schedule this with this person.”
Turner Novak:
It’s kind of all the things you described earlier that you can do with Claude.
Tomasz Tunguz:
Yeah. But instead of doing them synchronously back and forth, you can create a huge long list, and then those tasks can take 30 seconds or they can take three hours. That’s how you token max when we were talking about, that’s how you jump from a million tokens a day to 100 million or 500 million tokens per day.
Turner Novak:
And Anthropic and OpenAI, it sounds like, want people token maxing, the highest margin version of token maxing, which is probably like a B2B workflow in some capacity.
Tomasz Tunguz:
Exactly. And you want people thinking that they no longer want to interact with a computer without AI, which I think many people in the Valley are already there. Because you can just do so much more, because I can just enumerate this list of tasks, and then Claude or some other model will just burn through that backlog.
Turner Novak:
Is there anything that you’re not using AI for right now, on a computer?
Tomasz Tunguz:
There are some tasks. I’m on an Android, and so it won’t answer SMS messages because that pipeline’s broken. But no, you really wanna stay, there’s this great book called Flow, right, which talked about how do you get into a place where when you’re working, you’re just directly connected.
There was a philosopher named Heidegger who talked about the design of tools. If you think about using a fork, once you learn how to use a fork, the fork becomes an extension of your hand, and you don’t feel a difference. And I think working with an AI is like that in the sense of, I can just tell it, right? I can use the most native, I don’t have to learn to type, which is probably a dying skill.
Turner Novak:
Yeah. You can literally voice dictate to it.
Tomasz Tunguz:
I can just dictate what I want it to do, and then if it has enough information about the way that I work, and it has access to my systems, and I’ve helped it fashion its own tools, then it can work as if it were me.
And why would I, I’ll give you an example. I was on a plane going to Atlanta, and they told us in the waiting area there’s no Wi-Fi. And so half of the people are relieved because they can watch a movie guilt-free, and the other half, the workaholics, are like, “Oh, gosh, what am I gonna do for three and a half hours?”
So I sat there, and I tried to find a really fast internet connection so I could download a local AI model. Because now I look at the laptop and I’m like, “What are you doing? You’re so dumb.” It’s the same feeling when you get into a self-driving car and you start operating it, and then you get into a regular car because you’re someplace, and it’s like, “Why won’t you drive yourself?”
Turner Novak:
Yeah. So I have one more question on this inference topic. I actually don’t know on a tech level how this works. Anthropic’s business, you could say they basically sit on top of a cloud provider, and they’re basically this layer on the cloud provider. How does that actually play out in the sense of how that business model works? Do they need to build their own cloud provider eventually? Because they’re just kind of like a GCP wrapper or an AWS wrapper really at the end of the day.
Tomasz Tunguz:
They can decide. So you can own the buildings and the chips inside, which are called data centers, which Google does. Let’s think about this three-layer cake. There’s the data center, and then there is the chip inside the data center, the GPU, the chip that’s analyzing, and then there’s the model.
Let’s look at those three layers. Google has all three. Google manages its own data centers. Google manufactures and designs its own chips called TPUs, tensor processing units, and then Google makes its own model called Gemini and Gemma. And that is a great business.
And then you can say, okay, Anthropic does not own the data center. It does not design its own chips. It just makes a really great model. And that looks a lot like Netflix. So Netflix competes with Amazon. Amazon has Prime Video, but Netflix runs a lot of their infrastructure on AWS. Both businesses can succeed. There are pros and cons to each.
A great segue is, let’s look at SpaceX AI. SpaceX AI has a data center. They don’t have chips, so they’re missing that middle layer, and they have a model. So there it’s an Oreo, where they’re kind of, well, an Oreo with nothing in the middle. Oreo with a vacuum.
Turner Novak:
It’s like the, it’s an Oreo when you take it apart and lick the icing, and then you stick it back together.
Tomasz Tunguz:
Yeah, there you go. That’s right. Or you put somebody else’s icing in it.
Turner Novak:
Yeah.
Tomasz Tunguz:
Yeah. So there’ll be different strategies, and you need different amounts of capital in order to do that, and you’ll have very different margin structures. If you can vertically integrate, which means own each layer, I think you will ultimately be significantly better off because you can design the chips and the data centers for your algorithms. Whereas if you’re a model company, you will definitely have a say in how those chips and those systems are designed, but you are not the only customer.
Turner Novak:
Fair. And you probably need to have enough scale to justify the investment into all your own stuff, because it’s not easy and it’s not fast and not cheap.
Tomasz Tunguz:
No, it takes, it might take you, I don’t know, Google has been developing the TPU since 2012. Amazon has been developing their own chips called Trainium and Inferentia, I think, for the last five or six years. And it probably takes seven to 10 years to get to a place where you are at state-of-the-art. You have executed enough cycles to really be there.
So at some level of scale, sure, if you’re one of the five most valuable companies in the world. Apple has its own chips, all the M1 to M5 silicon that you and I run on our computers. That’s proprietary, and it’s a big advantage.
Turner Novak:
So then the play is probably, if you’re Anthropic right now, maybe at some point you need to start doing that. But it’s really just get as much adoption as you can, get as much usage, get as much revenue to have cash to work with to now fund all this stuff. And to your point, it’s just a sprint. Go as fast as you can to get there.
Tomasz Tunguz:
Yeah. If you have a significantly better model, you will win share. And the opportunity cost is so huge, and the willingness to spend is enormous, because if your model is meaningfully better, you might add $100 billion to your market cap in a quarter.
Turner Novak:
So I think it begs the question, where do you think is a good place to be investing in AI today? Is it over because Google is vertically integrating and will win everything, except maybe Anthropic and OpenAI win on the edges? Is it wide open for startups? Obviously, you’re investing in startups, so maybe this is a loaded question, biased question, but what do you think the opportunity is today investing in AI?
Tomasz Tunguz:
There are certain markets that are uninvestable because they are on the direct roadmap for the large companies that are incredibly well-capitalized. So agentic coding, I think if you were to start an agentic coding company today, it’d be very difficult because it is probably the most important market, and you have so many businesses whose roadmaps are pointed in that direction.
Turner Novak:
Is it the most important because it is so tied to that inference thing that we talked about, where there’s just so much inference flowing through that?
Tomasz Tunguz:
Yeah. Okay, great question. Why is agentic coding such a phenomenal product market fit with AI? The first is there is a lot of spend in software, so the market today is really big. The second reason is software engineers are largely very expensive, so there’s a lot of labor spend as well. So there’s technology spend, and there’s labor spend. Both are very large.
The third is the demand for software, I would argue, is infinite. You and I, as we age, and all of us, will only use more software. We won’t use less. And it will become increasingly sophisticated, building on the previous software. So you have labor spend, software spend, and a very fast-growing market, and a market with infinite demand.
And then the last thing is, it is a set of tasks that an AI can test whether or not the AI’s answer is correct.
Turner Novak:
Because it’s so objective, rule-based, and you know if you got it correct or not.
Tomasz Tunguz:
Yeah. It’s like math. Either the equation resolves or it doesn’t. And if an AI system can test that itself, well, then sure, you can just let it spin overnight until it has satisfied all the different equations, or all the different parameters that you’ve defined for the piece of software. So that’s called a closed-loop problem. You can just have the machine spin faster and faster and faster. So the combination of all those four makes it really great. Makes these systems perform exceptionally well in software.
Where is that not the case? Well, let’s say we asked it to paint impressionist art. You and I can debate, are Monet’s Lilies the zenith of impressionist art? You gotta, you can say, “No, Pissarro is the bee’s knees.” So it’s subjective. It’s open-loop.
The blog post, how do you, when we summarize this great episode that we did together, there’s no objectively best blog post. So that’s not a closed-loop problem. The AI has a much harder time because you can’t just let it spin. You have to apply judgment as a person and say, “That’s enough.”
Turner Novak:
So the reason that all the biggest AI companies are going after agentic coding is because it’s ultimately the biggest TAM and the biggest opportunity. So then you’re almost accepting that you’re maybe settling, quote-unquote, for smaller, less interesting markets. But then there’s an opinion to be had of, well, these are actually still very big markets, or they may be very strategic for these other reasons.
Tomasz Tunguz:
Right. Yeah, it’s like, after Google in 2006, would you have started a search company? Probably not. Maybe, did DuckDuckGo? I don’t know. I think DuckDuckGo maybe started around then, and it’s still alive. But yeah, I don’t know if I would’ve invested in it.
Turner Novak:
No, it’s just really tough because you don’t attack your opponent in the area they are strongest.
Tomasz Tunguz:
So do you think that there may be some jockeying where, I don’t know, a company that’s not in agentic coding that we all know of and hear of every day just suddenly emerges and has created a position to ladder themselves in there or something?
Tomasz Tunguz:
Well, you know Cursor, right? There are all the dynamics around Cursor and the brilliant business that they have built. So that’s definitely an interesting one to watch. And you have Poolside, which is releasing US open source models. So now sovereign AI, AI that is limited to a particular country, has become a critical geopolitical issue.
You have companies that are building models for India and companies that are building models for Japan and United Arab Emirates. So maybe there’s a market segmentation. You say, “I wanna be the best agentic coding system for India.” There may be a market segmentation there that makes sense, just the way that you might have a vertical search engine to compete with Google that was focused on travel for a long time, and that was a standalone vertical.
So it’s not to say that you can’t segment and then compete within that segment. I don’t think you can just go and say, “Okay, I wanna win the United States agentic coding market as a model provider.” That’d be tough, unless you really have a meaningful scientific advance, a mathematical advance.
Turner Novak:
How do you think then about what are the opportunities that are interesting? How do you figure out, is this side market, this other market, this non-incumbent market that they’ve already kind of captured? How do you figure out what’s worth going after?
Tomasz Tunguz:
Well, let’s think about the markets where clearly they’ve demonstrated an interest, the incumbents. So agentic coding is one. The second one is health. OpenAI has a great team pushing health products. You have Anthropic launching a collection of skills on Monday of this week, tied to finance and the automation of finance. That’ll be important.
There’s legal work that’s associated, so the legal market is definitely in scope for them. Anything around infrastructure and software automation is definitely core. Those are some of the markets. I’m sure there are more. Security, clearly they will push. I don’t know if the model companies will dominate that market in its entirety. They will be a supplier more than an individual competitor.
But there you have six markets where the direct competitive dynamics of the largest AI companies you must consider. And you can either invest and say, “I’m going to, I believe a company is sufficiently far ahead that one of the incumbents must buy or partner with them.” Viable investment strategy.
Or you can say, “Okay, there are 10 markets they really care about, and I’m not investing in any of those. I’m going to go pursue markets 11 through 100.” And then I’m going to analyze each of those market dynamics. How many competitors are there? How many venture-backed competitors are there? How likely is it that the customer population adopts software?
If you’re a longshoreman, the odds you adopt AI, I think are pretty low. But if you are in the business of back office automation and you are like an insurance company or a third-party logistics company, pretty high. And then the question, do the model companies care about that market or not?
Turner Novak:
So then what’s your lens for thinking through this whole SaaS apocalypse? We’ve gone through these waves where people are like, “Oh, every software company’s dead.” And then I don’t know if now it’s flipped or it’s like they’re not all dead. I’m not sure where we’re at. It’s hard to keep track. In terms of that side, if you’re a mature software company, how do you think about the defensibility?
Tomasz Tunguz:
Yeah. Okay. The public markets value growth. It remains the most important factor as an input to valuation. It’s about 50 to 60% correlation.
Turner Novak:
You’re saying the growth rate of a public company, 50% of its valuation is just depending on how fast it’s growing?
Tomasz Tunguz:
Yeah, 50% is explained by it. Yeah. So which are the three fastest-growing segments in the public markets? The first is security, the second one is data, and then the third is core systems infrastructure. All of those have tailwinds from AI.
The slowest growing ones are vertical software companies, and then productivity apps where some of them are seeing negative growth. And then I forget the third. But there really is a distribution. It’s not, you can’t look at it as all publicly traded software companies. There’s a distribution. The faster growing ones are doing fine, and then the ones that are slowing or contracting will be punished.
One really interesting question, actually, this’ll be fun with you, Turner, is, imagine you are at 2001 and the dot-com crash has just happened, and you’re looking at all the venture-backed and publicly traded software companies. They were building on-prem software. So you would have a CD, and you would get a box of software at a store, and then you would install it, right? And you’re the head of IT for your company.
And then after 2002, some number of companies moved to the cloud. Which companies were big during the boxed software era that transitioned to the cloud, that survived, maybe even thrived?
Turner Novak:
Uh, so I was born in 1991, so I was about 10 or 11. So I’m trying to give you the perspective I would have as a public market investor in ‘21 or 2001/2002. At the time, I’m just trying to think of how it even ties up. I guess looking back in hindsight, maybe Adobe.
Tomasz Tunguz:
Yes. Great. Yes.
Turner Novak:
But that is not really what they did in 2001, right? Like, they slowly transitioned to the cloud over the past 25 years. But I mean, it probably didn’t start in 2001. It probably started in 2005 or something.
Tomasz Tunguz:
Yeah. No, that’s right. Okay, so Adobe is a great case.
Turner Novak:
Maybe Salesforce.
Tomasz Tunguz:
Salesforce is post-cloud. So they launched directly on the cloud, and then their banner was no software, which meant no on-prem software.
Turner Novak:
Okay. Maybe Oracle, but I don’t know how fair that would be to count.
Tomasz Tunguz:
Very fair.
Turner Novak:
Okay.
Tomasz Tunguz:
Yeah. So you’re on it. You have Adobe, clear market leader with Photoshop and InDesign, and all those things. You have Intuit.
Turner Novak:
Yeah, that’s a good one.
Tomasz Tunguz:
TurboTax and all that stuff. They made the transition dominant in their category. You have SAP, right? 50, 60-year-old software company.
Turner Novak:
Man, that is a common one. The AI stuff is all going hard at SAP now, I feel like.
Tomasz Tunguz:
Mm-hmm. Yeah. That’s right. So can they survive again? We’ll see. Anyway, you keep going through this exercise and we were able to name about seven to eight companies that navigated that transition.
Turner Novak:
Out of how many?
Tomasz Tunguz:
I have no idea. How, do you know how many there were?
Turner Novak:
I mean, hundreds, right? I think it’s order of hundreds.
Tomasz Tunguz:
So are there characteristics of some of these? Is it that they had a very specific customer that they served? And were they like, did they have management teams that took the cloud seriously maybe? That feels like a big component of it.
Tomasz Tunguz:
I think the characteristic is that they were near monopolists.
Turner Novak:
So it almost didn’t matter what they did, whether they made the change in three months or 10 years. They just would eventually manage it.
Tomasz Tunguz:
You think about Oracle, transactional databases inside of banks. Who’s ripping that out, right? It still hasn’t happened. Intuit, there’s nobody else even close. Adobe. Name, I mean, before Figma, name a competitor that mattered to Adobe. Didn’t matter. SAP. Can you name another enterprise ERP system?
I’m being a bit glib here, but I do think they just had tremendous control or tremendous presence within their markets, which bought them time, and they clearly had the resources to figure out how to make the transition. As a result, customers couldn’t leave to a better alternative because maybe there were or there weren’t.
But I think it really is a dominant market position that buys you the time and gives you the resources to learn how to transition. And maybe it affords you the opportunity to buy a market leader and then integrate that DNA plus the product into the next evolution of the business.
Turner Novak:
So it’d probably just be paying attention to, in pretty much all these categories, there’s probably a bunch of these AI native companies, and it’s just seeing these incumbent publicly traded, how does their product seem to be evolving relative to these new companies that were founded in the past couple years? And are they able to make these changes fast enough to continue to keep their dominant position? There’s probably a couple, and there’s also a lot more that won’t do it properly.
Tomasz Tunguz:
It’s very hard. Yeah. ServiceNow has about three or four different AI companies, right? They’ve definitely been aggressive. That would be an example. But there are many companies that really have not yet responded and will need to.
Turner Novak:
Yeah. In terms of maybe, I can’t remember if we were talking about this before we started recording or not, but just the impact of AI in the economy compared to some of these other economic cycles. Did we hit on this a little bit? I think railroads was the peak. I think you said it was 7.6% of GDP or something like that.
Tomasz Tunguz:
Mm-hmm. That’s right. Yeah. We’ll be about low to mid 2% of GDP within this year. And in Q1, 75% of GDP growth is AI.
Turner Novak:
And a lot of this is data center build-out.
Tomasz Tunguz:
Yeah. So there’s the construction, the manufacturing, the assembly, the chips, the networking associated with it. And then all the labor that’s associated with that, and then the revenue that’s generated from it, which, fastest-growing market.
So yeah, 75% of all US GDP growth, if it continues to grow at this rate, the US overall GDP will continue to grow much faster, and then it will go from 31 to whatever it is, 33 or 35. And then if we can get to seven or eight or 10% of that, you’re talking about $3.5 trillion a year of investment going into AI in the intermediate future. This is a big business. It’s a big industry.
Turner Novak:
Well, and you think about the scale of, I mean, cloud’s maybe an interesting example, mobile. Did they make the economy grow faster? I’m not actually sure. They had to have.
Tomasz Tunguz:
Oh, yeah, of course. The networking build-out, this was, you know, when you were 10 and I was 18. Before the internet was broadly adopted, everything needed to be connected. Every house needed to be connected, every building needed to be connected, fiber and copper.
So you had huge GDP, not nearly close to the scale, but significant GDP when you had Nortel Networks and Qwest and all the initial internet service providers who were then the telephone companies adding new telephone lines that were ultimately replaced by fiber. That drove a lot of the ‘99 boom. Juniper Networks and Cisco and all those businesses, they were explosive. Very similar to this era.
Turner Novak:
Yeah. Well, and then, I mean, that begs the question, it didn’t end that well, right, in 2001? Do you feel like, is there sort of a bear case to be made in terms of just being careful or being cognizant of where we’re at in the technological or economic cycle, or the capital, the debt cycle related to all this stuff? Is there any kind of thing that you keep top of mind when thinking through that?
Tomasz Tunguz:
Yeah. It is a lot different than 2001, because in 2001 revenue models of many of the businesses were not known, right? Amazon, okay, fine, in the fullness of time, but like Peapod, which was yesterday’s Instacart. We didn’t have phones. It was literally you’re placing your food delivery order on the computer or whatever.
Turner Novak:
Yes. Right.
Tomasz Tunguz:
With dial-up. Like it takes three minutes to load.
Turner Novak:
Yeah.
Tomasz Tunguz:
So it’s a different era. I think you can legitimately say AI converts electricity into work, just the way that gasoline is converted into work if you use a lawnmower. And it can meaningfully improve the productivity, right? You have like Boris Cherny from Anthropic who talks about he can ship 30 to 50 times as much code with AI as not. Okay, he’s turning electricity into real work.
Okay, so what are the things to worry about? The first is, yeah, the credit markets. Many of these data center build-outs are built with 80% credit. OpenAI, I think SoftBank limited the size of the debt. I think they dropped it by 40% this morning. So we will see what happens there.
When you borrow money, like you borrow money for a house to pay for a mortgage, you are providing the house as collateral to that mortgage, and the lender looks at the house and says, “Okay, what does the inspection say? How long will the roof last? How much investment?” In the very same way, people who are lending to data centers have to look at the GPUs. How long will those GPUs last? Are they productive? Will they fail at some level? And there’s a debate about how long those inference GPUs are productive. So that’s a big one. The credit market is definitely one.
I think the argument at some point, like the token maxing wave, I think in the back half of this year will wane, and everyone will say, “Yeah, you’re burning a lot of electricity and you’re buying a lot of intelligence, but what did it do for the company?” I think that’s definitely coming at some point. But overall, it’s hard to paint a negative picture.
Turner Novak:
I mean, part of the negative, just general perception, is there’s gonna be all this job loss or whatever. The other one is like water usage in data centers or something like that, and contaminating the land, noise pollution maybe. I don’t know, whatever the argument is for the data centers.
But the other side though, with jobs, is it’s actually not gonna cause job loss. If you look at every technological revolution, it always ends up actually creating more jobs. What do you think will be, or at least what are you seeing, maybe it’s still pretty early, but what kind of new jobs do you think we’ll see from a lot of the AI build-out?
Tomasz Tunguz:
Yes. Okay. So let’s talk about why there are more jobs. There’s not a finite amount of work to be done, right? There’s this thing called a lump of work fallacy, which is, there’s a total amount of work to be done every day across the globe, and there’s a certain number of workers, and they have to allocate their share, and then once they’re done, they go home.
Turner Novak:
I’ve never heard this before, but it makes sense, yeah.
Tomasz Tunguz:
Yeah. But if you’re a workaholic or you’re married to a workaholic, you know there’s always more work to be done.
Turner Novak:
My wife’s listening to this like, “Yes, let’s...”
Tomasz Tunguz:
Right? And so, okay, what ends up happening? Well, you used to write Java code, and then somebody used to review that Java code. Well, great. Now you no longer have to write or review that Java code. You have to architect that system, and then you have to make sure that system is now resilient.
And it turns out, in order to compete, you can no longer just offer a point solution. You need to offer six times the breadth of the product. Okay, get to work, right? And so I think that happens just across the board.
We looked at the automobile industry in the United States before interchangeable parts and Taylorism. You had 80,000 people who were artisans building different components of a combustion engine.
Turner Novak:
This is like before the assembly line too?
Tomasz Tunguz:
Yes. Right before the Model T.
Turner Novak:
So just some dudes sitting in a room in a circle banging parts together.
Tomasz Tunguz:
Yeah. Making a piston, right? Or camshaft. And it worked, and they sold cars. And then all of a sudden, the price of a Model T collapsed. Collapsed automobiles, and then everybody was driving one. And the number of people working in the US automobile industry within five years went from 80,000 to 500,000.
There were a few people working on the line, but there were people marketing the cars, there were people designing the cars, there were people building dealerships, there were people building roads. So the overall employment exploded.
And it wasn’t, you know, around that time we were looking, there were about a million manual dishwashers, people who washed dishes for a living.
Turner Novak:
Wow, that’s crazy. This is in the US.
Tomasz Tunguz:
In the US. Yeah, every restaurant in the United States needed three or four dishwashers.
Turner Novak:
That’s like 1% of the population, 2% of the population. Yeah. That’s crazy.
Tomasz Tunguz:
Yeah. Or farming, right? Think about the shift from agrarian farming, and people moved to the cities, and they found all kinds of new work, and now we have all these incredible industries.
Turner Novak:
I mean, we used to literally just send kids into the mines, and you might die, and you get lung disease, but, “Hey, we got some coal from it.” Or like, so it’s gotten a lot safer too.
Tomasz Tunguz:
It’s gotten a lot, although software engineering, I will say, is not that hazardous to your health.
Turner Novak:
Carpal tunnel.
Tomasz Tunguz:
Carpal tunnel will get ya. Carpal tunnel. Myopia maybe. But I agree with you. I think that’s right. And there are other benefits. You look at the Waymo statistics of how much safer these cars are. 50,000 people die in the US, unfortunately, on roads, and once we get to a place where we have significant volumes of cars, think about, the longevity of the average American will increase as a result of the safety.
Turner Novak:
Yeah. That’s a pretty big one where I hear that a lot is, there’s millions of people who drive, and this is a significant displacement. We were at dinner probably like last week, and I overheard the women beside me talking about this, and the massive concern with them was they don’t trust them, but then also like, “Oh, what about all these drivers that are gonna lose their jobs from these self-driving cars? I can’t support that.”
I think the argument, though, is there’s probably still gonna be people in these vehicles in a decent amount of cases. Like long-haul trucking, you may still need, maybe it’s flat or something. There’s just more trucks on the road that are unenabled by this, and you’ll still have people in the warehouses that are unloading them. Or maybe you sleep in the trailer or whatever. You have a nice bed, and you’re maintaining the car while it’s self-driving across the country or something like that.
Tomasz Tunguz:
So long-haul trucking, average age of a long-haul trucker, I was just looking at this, 46 to 47. It’s not an industry where lots of young people are gravitating to. And maybe the tastes of new job seekers have shifted, and they don’t love that lifestyle.
So I think there’s two parts to it where ideally we are automating the jobs where there’s not a tremendous amount of labor supply. One of the ways of looking at AI is, you really need, a great place for AI, while it’s not perfect, is you have a labor market shortage. You have somebody, the hiring manager, who needs that job to be done, and therefore, they’re willing to accept like a 70% solution.
Electric pole inspections, long-haul trucking, anything to do with sewer inspection. Those kinds of things, AI is phenomenal at. And it can be a very unappealing job. So maybe there’s this generational shift where people’s preferences for different kinds of work evolve and the machines take the work that is no longer interesting.
Working, tilling a farm. There’s some fraction of the population that likes that, great. But you don’t have 10% of the population who wants to go and yoke some ox and oxen and then plow behind them, right? The preferences change.
Turner Novak:
Yeah. And I think too about accounting or finance, right? Back in the day, an accounting department was just a big building. Maybe it was next to the factory with just people literally writing the debits and credits on paper or whatever, like manual invoices. And some people still do some of this kind of stuff, but now it’s literally a spreadsheet, and you type it in, and it automatically calculates. And like QuickBooks, literally, we were talking into it, the software just does it for you. It calculates the financials. You can literally press a button and get the final financials.
Tomasz Tunguz:
We all know what a calculator is, but when I say the word calculator, you imagine, I don’t know, like a TI-82, or an HP calculator. But before that was invented, there was a title.
Turner Novak:
Like a human person that was a calculator.
Tomasz Tunguz:
Yeah. The Apollo missions, all the math was done, much of it by women who, and their jobs were like senior calculator.
Turner Novak:
They literally had, I think I’ve seen those pictures where there’s a woman who was standing, and there’s a stack of papers that she had calculated that was literally taller than her or something.
Tomasz Tunguz:
Yeah. It was just calculations of the route or whatever they had to calculate for this thing.
Turner Novak:
Yeah.
Tomasz Tunguz:
Yes, the trajectories and the orbits. Yeah, that’s right. And so, okay, what happened to all those calculators? Well, we found other work for them at a higher level. They didn’t have to look up logarithms in big books.
Turner Novak:
Well, maybe instead of spending literally weeks just hand calculating the equations, it’s done by the computer. You’re like, “Oh, this was wrong. The calculation was wrong. Let’s see what we need to change about this route.” And you get into more strategic work around the calculation, the stuff that’s a rule-based thing really at the end of the day.
But yeah, just to the point of, with investment firms or finance, right? Back maybe in the ‘50s when they were doing, selling junk bonds or whatever they were doing, doing early stage LBOs and companies, you had this army of people that just had to punch out all the calculations. Versus now it’s like, a lot of them are doing more sales, more marketing, right? It enables more people to do LBOs, more people to take out credit, more people to raise venture capital, because we have this army, all the AI’s doing all the analysis on, this is a good investment. So it’s just all these VCs going out and giving money to founders and enabling them to start companies.
Maybe I’m exaggerating this a little bit, but the productive work shifts towards things that grow a business or add more sales, do more things for customers.
Tomasz Tunguz:
Yeah. Those calculators got into the business of aerodynamics, computational fluid dynamics, quantitative stress modeling on different elements. There’s always more work, and it’s increasingly sophisticated.
Turner Novak:
Yeah. Are there areas that you think AI is still underrated today? Or maybe you’re expecting it to get really good in the next couple years and people are maybe not thinking about it? I know you invested in an advertising company recently.
Tomasz Tunguz:
Yeah. We’re really keen on online ads. I think Google generates something like $120 per user in the United States in ads, and the online ad market in the US is about $450, $460 billion, global $460 billion. And if you think about what ads can do for offsetting the cost of GPUs and also helping consumers find things that they might like, I think it’s an absolutely huge market.
And so we’re very keen in that space. It’s been tough, I think, for startups as a whole within the online ads ecosystem, but AI is such a disruptive force that I think there’s an opportunity to build a great business, and we’re lucky to work with a fantastic team there.
Turner Novak:
Is there anything that’s not in the data that you are kind of waiting for or looking at? Maybe there is data, but it’s not well-known data or it’s not matured data, it’s just early signs of things.
Tomasz Tunguz:
Within the online ads ecosystem?
Turner Novak:
Or just in general, in AI adoption or in usage or...
Tomasz Tunguz:
I don’t think we’re seeing the productivity gains yet. Why haven’t we seen that? Well, before, say, October or November of last year, AI systems were great search engines. And then in November of last year, the models started to be really great at executing workflows, multi-step processes.
That’s where you really get time compression in work, because I can write up a workflow in English, and then I can say, “Here’s a list of 100 entries in a file, and I want you to run each one of these workflows in parallel.” And boom, in 15 minutes, I have the work that I could have done in four days.
And we’re not really seeing that in the productivity statistics yet or in the earnings per share of publicly traded companies, but it will be significant and sustained. I think it’ll be tremendous. So maybe early next year or mid to late next year we’ll see that.
Turner Novak:
So what’ll that show up as? Is it like, I think I saw, I didn’t actually look at this, I just saw Datadog, the day we recorded this, is up like 30%. I saw someone make a joke that, this is the AI productivity we are expecting. Maybe it is or maybe it isn’t related to it, but is it just companies are getting more efficient per employee essentially? Is that probably what shows up?
Tomasz Tunguz:
Mm-hmm. Yeah, they can do more work per hour, whatever that unit of work is, whether it’s lines of code for a software engineer or customer support cases solved for a customer support rep, or companies reviewed by a venture capitalist. It is the throughput of whatever factory you are operating, has just gone up, because the conveyor belts and the machines can now operate at twice the speed.
Turner Novak:
That’s a really great mental model for it.
Tomasz Tunguz:
There’s a whole discipline called operations research, which is, I have a factory with a factory line, and as I change different components to it, how many more chocolate boxes can I make? And I think with AI, the reality is, I think people will, could you see a 30%? We just talked about Boris, who’s at 50X. He’s clearly, I don’t know how many standard deviations out, but can you see a 3X to a 5X productivity gain for a software engineer on average? Maybe 3X. So all of a sudden your software factory is now operating at 3X the throughput.
Turner Novak:
Is this kind of related to, you put out a study, I think it was about a year ago, where you interviewed a bunch of, or you ran a survey with a bunch of, I think it was in go-to-market and with sales teams, and basically you found that using AI had zero impact on revenue growth or something like that. So what was the study, and maybe has this changed in the past year or two?
Tomasz Tunguz:
We’re just about to launch the new go-to-market survey, so we will know this year. And I think in retrospect, that’s the answer we should have expected, because, again, everyone had access to a fancy search engine instead of a system that could actually paralyze work. So I think even this year we will see modest positive response, and then next year I would expect to see very significant response.
Turner Novak:
Interesting. Okay. And then how do you think people are actually buying AI today? What are you seeing in terms of maybe companies you’ve invested in, surveys that you’ve done? What do people seem to be getting purchased? Maybe, what’s the obvious things? What’s the less obvious things? And just, what’s the general decision-making framework that you see people using?
Tomasz Tunguz:
So there are buying committees. There is the line of business owner, VP of products, VP marketing, VP customer support. There is the head of technology, VP of engineering or CIO, head of security, and then oftentimes general counsel because there are lots of different data information and security questions around AI.
I would say the sales cycles were extremely fast November until March. And now, as a result of some of these buying committees becoming more sophisticated, they’re slowing a little bit, but they’re still much faster than software sales cycles.
And one mental model, which is not universally true, but it is useful, is that every leader within an organization will pick a platform that they trust to deliver to them the vast majority of their agents. If you’re the head of data, you’ll pick a company like Monte Carlo and say, “Great, I trust you to deliver all of these data agents.”
If you’re a VP of engineering, you’ve kind of already done that either with OpenAI or Anthropic or Cursor, one of the three. And same for sales. And many of those categories still, it’s still TBD who that brand is, but that’s what will end up happening. As a leader, you’ll trust, you’ll make a career decision and say, “I trust this particular company to deliver for me all the different sales agents I could need.”
Turner Novak:
So then there might be a sort of jump ball type opportunity in some of these categories, like in sales, like Salesforce. Do you make a bet on Agentforce or whatever all the Salesforce AI stuff is? Or is there a new, more emerging product or company that’s out there that you maybe make that bet on?
My guess it would probably depend on who the actual decision maker is there, and if they use the product, and then if they probably make a bet on the slope of improvement. You may say the startup has added all these new features, they’ve got so much better. Probably making a career bet almost on, this roadmap seems like it’s actually gonna be super useful for us and will actually drive the needle versus maybe the existing option we’re using. Is that a fair way to think about it?
Tomasz Tunguz:
Yes. Right. And today you have general purpose tools. You have low-code and code workflow builders that are growing very fast because there’s been no specialization. And the most valuable tool now is a stem cell that I can play around with and then have it specialized until I see it germinate and blossom into a workflow that I will then crystallize, which is what happened in software, right?
In software, everyone was building a whole bunch of custom stuff, and then you had Salesforce that said, “This is the right way to run a modern sales organization with software.” And HubSpot did the same thing for the SMB, and then Marketo came around. And then the workflows, I don’t wanna say they calcified, but they definitely crystallized around best in class, and everybody copied that until there was a new platform shift and everything has to be reinvented.
Turner Novak:
Do you think a lot of those companies are probably already founded?
Tomasz Tunguz:
No, I think it’s wide open.
Turner Novak:
Really? Okay. Any areas that you’re most interested in, in Theory, for people listening, if they’re like, “Oh, I’m working on this”?
Tomasz Tunguz:
Yeah, we’re really interested in online ads. So if you’re doing anything in the online advertising ecosystem, please look us up. We’re very interested in inference. We think you can think about, inference will be the biggest market, and there are many different kinds of inference. There’s like really fast inference or real-time inference. There’s inference that’s for images or video. There’s inference that is for very long-running background tasks.
Just the way that if you had 1% of the database market, you can become a public company. If you have 1% of the inference market, you’ll be able to be a public company. So specialized inferencing is fascinating to us. And then another category we’re really keen on is email and the automation of email with AI.
Turner Novak:
Interesting. And is this because agents are gonna start reading most of the email? Is it...
Tomasz Tunguz:
Yeah. Turner, what are the odds you’re logging into Gmail five times a day in two years?
Turner Novak:
I’ve been thinking more and more about how much of my time is just deleting these stupid AI emails that I get that just, it’s always the same format where it’s like three follow-ups and whatever, and they’re pretty, I don’t know.
Tomasz Tunguz:
Yeah. There’s no way, there is no way you’re logging into an email account five or six times a day in two years.
Turner Novak:
So what do you think is gonna happen? Am I just sitting in Claude and it’s pinging me when I get the best ones or something, or am I only texting or Slack?
Tomasz Tunguz:
Yeah, it will learn what you care about, right? It will learn who you care about, and everything else it’ll either summarize and prioritize. But there’s just no way. Look at the volume of emails you and I both receive and millions of other people do. I don’t wanna spend my time, and neither, you know, please archiving this and archiving that. And then now a text message is about, I don’t know how much credit you were offered today, but I can tell you.
Turner Novak:
Yes, I get the calls every day. Every day I get a call on average about...
Tomasz Tunguz:
I know a guy if you need to borrow some money.
Turner Novak:
Oh yeah. Great. Part of me, I’ve thought, should I just do one of these and just get like a $100,000 loan or whatever? Should I just see what happens if I actually say yes to this? It’s kind of funny.
So you do these predictions every year. I think the ones you put out for 2026, I feel like we’ve actually kind of hit on some of them.
Tomasz Tunguz:
Oh my gosh, it’s depressing, isn’t it? Like half of them are already there.
Turner Novak:
But one of the ones that you predicted was a lot of liquidity in kind of the late-stage ecosystem. I forget which ones you said, but there’s SpaceX, OpenAI, Anthropic, Databricks. I don’t know if Anduril is considered if it’s big enough or close enough to IPO-ing, but there’s just a lot of these companies that are, I don’t know, a couple trillion dollars of liquidity.
Do you still think that’s gonna happen? We’re a couple months in now, and what do you think the impact of that’s gonna be?
Tomasz Tunguz:
Yeah. I think SpaceX, OpenAI, Anthropic definitely go public. Stripe and Databricks, I’m just looking at the blog post now, I don’t think that either one of those happens. If those three go public at $50 billion each, they will raise more money from the public markets than the sum total of all IPOs in the previous decade.
Turner Novak:
You’re saying if each of them, when they go public, if they raise $50 billion on average between the four or five of them, it will be, between the three of them, it’ll be more money than the last decade of IPOs that they’ve raised?
Tomasz Tunguz:
Yeah. When Facebook went public, it was a $15 billion IPO, and it was unconscionably large, and there was one, and now we’re talking about three $50 billion IPOs. Sure, inflation, okay, let’s say there are 40% more US dollars today than there were back then. You’re talking about $16, $18 billion compared to $150 billion. It’s still 10X larger.
So it is bending the public markets in a very real way. They will go public. I think there’s a real question of how people become liquid and sell those positions, but it, the only thing it can be is positive for the ecosystem.
Turner Novak:
What do you think happens with the late stage venture market? Because there’s a lot of people, their business model is just like asset management firms. The business model is getting their cut of these rounds when they happen. Do we basically just have new companies that grow into it and take their place, where they’re like new trillion dollar private companies that then IPO in another five or ten years?
Tomasz Tunguz:
Okay. When I started in venture in 2008, there was one billion dollar outcome in enterprise software.
Turner Novak:
Really? In a whole year?
Tomasz Tunguz:
Yeah. Up until that point, aside from Microsoft.
Turner Novak:
Oh, so there had only been one outcome of over a billion dollars?
Tomasz Tunguz:
Yeah. Venture backed. And I remember being in awe of the venture capitalists who had that billion dollar outcome, and everybody was like, “Wow.”
Turner Novak:
Yeah. Now you can start a company and raise over a billion.
Tomasz Tunguz:
Yeah. Well, this is exactly the point. And to go public, you needed about $50 to $75 million in revenue, and you would raise $35 to $50 million in an IPO. And there was a bank that would underwrite you and take you to market and charge a fee for it.
And today, a Series A, many Series As are larger than those IPOs, right? Every Series B of significant company is larger. So the private market has basically taken over that. And I think that’s fine. It’s because it’s so expensive to go public.
But there’s plenty of business there. My point is, the IPOs of 15 to 20 years ago are today’s mid-size Series Bs and Series Cs, and there’s plenty.
Turner Novak:
Yeah. Well, so then does the average Series B or Series C in 10 years, is it like a trillion-dollar valuation? I hope it doesn’t continue that direction.
Tomasz Tunguz:
No, that means we are in hyperinflation like pre-war Germany. No, no. I hope not. I think it’s cyclical, right? You have the oil industry went through a huge boom, and the railroad industry went through a huge boom, textile industry went through a huge boom, automobile industry. So we will have a cycle. And when that cycle or that downdraft happens, no one can predict, but it will happen, and then the levels of over-investment will be exposed. Excuse me, but that’s healthy. It’s really important for us to have recessions and corrections.
Turner Novak:
So then I guess one question then, when you’re, let’s say I’m a founder, I’m meeting you for the first time. Maybe you do or don’t know much about my business and the market that I’m in. What kind of things are you gonna be asking me and looking at when you’re making a decision of what you wanna invest in and be exposed to today as a fund? What are the things that are most important to you that you’re thinking about?
Tomasz Tunguz:
Yeah. What does the company look like in seven to ten years? I think is probably the hardest question, but the most germane question in this era.
Turner Novak:
And is it ultimately you’re thinking about inference, owning some inference spend? It sounds like you’re thinking about advertising that you mentioned. I’m trying to remember the other two. I feel like you mentioned two other things. Email is one. Oh yeah, email is one. So you think a lot about, how do you slide into the future of how AI continues to eat more software?
Tomasz Tunguz:
Yeah. And what does the business look like in seven years? You can say, “Well, we have a technology advantage, some awesome piece of kit that gives us 18 months, and we will sustain an 18-month advantage in our market.” Very valuable.
You can also say, “We can sell better than anybody else and build a brand.” And brand is probably the only enduring strategic advantage of any company at scale, and so that’s also a very viable strategy. But you need a booster rocket, a way of getting a head start relative to the market. So what is the answer there?
Turner Novak:
Yeah, I feel like that’s kind of related to, going back to Theory Ventures. I think the name is kind of related to the disconnect in technology. Can you explain the name?
Tomasz Tunguz:
Yeah. Theory, and the website says we craft theories about the future and then help them become a reality. So we research a lot of different categories and try to understand the history of the category, which we’ve talked a lot about. And then if we know the history and we can understand the technology innovations that are occurring within it, then maybe what does the future look like? We try to find founders where we are similarly aligned in that vision and then work really hard to help them achieve their dream.
Turner Novak:
I know you were at Redpoint for about 14 years before you started Theory. What were sort of the seeds of starting to do this? Did you always know you wanted to, or was there a moment where you’re like, “This is it. I’m doing my own thing”?
Tomasz Tunguz:
Yeah. I had a wonderful time at Redpoint. Many wonderful people there who taught me the business, and I’m extremely grateful for it. And then decided to launch our own adventure, and now we’re three years into Theory, and we’re, I think we’ll be 15 people here by the end of the year. So we’re off to the races.
Turner Novak:
And it was, was it just you when you started it, or did you team up with more people that have joined you?
Tomasz Tunguz:
Yeah. Yeah, the team is 10. It’s always been a we, and I’m really grateful. Anybody who starts a company, for the people who join and believe when we don’t have an office and we’re all building it together, I’m grateful for their confidence and all their hard work.
Turner Novak:
Yeah. And I think one thing, I was realizing this when I was asking Claude all my things, trying to prepare this episode, I have not had a lot of people that have spun out from existing pretty big funds on the podcast. I’ve had a lot of people who were like, “I started my own thing. I raised from founders that I invested in,” blah, blah, blah. So how did you go about, you probably got to meet a ton of LPs over a long period of time, just putting the first fund together? What was the process?
Tomasz Tunguz:
Yeah. It was, it’s enterprise sales, right? Raising a venture capital fund is, you’re talking about an 18 to a 36-month sales cycle, because you are selling a 10 to a 15-year contract.
Turner Novak:
That’s a pretty long contract.
Tomasz Tunguz:
Yeah. You’re asking somebody to entrust you with their capital for 10 to 15 years. And best in class enterprise sales is 15 to 25% conversion with that sort of sales cycle, and so you need to build a funnel and build relationships and run it. Understand, okay, how do we map the account? How do the decision makers feel? How do we get references and all those kinds of things?
Anyway, that’s the mental model we applied, and I’m grateful to the limited partners and our investors who took a bet on us when it was just a pink deck and a dream. But it really is just working a funnel and building trust.
Turner Novak:
And I think you did a Monte Carlo analysis to figure out what the portfolio was gonna look like. I don’t actually know exactly what you did, but why did you do that? What was the process like?
Tomasz Tunguz:
It’s really, somebody tweeted this recently, which is, your fund size is your strategy. What does that mean? Well, for us, the opposite is true, which is our strategy determines our fund size, and that’s true at every raise.
This is how many companies we want to invest in. These are the kinds of ownership targets we want. This is how many of them we want in a fund, and this is how much money we want to continue to support them over different rounds. And given what’s happening in the market and what 75th percentile Series As go for, you can kind of calculate what that fund size should be, and that’s the way that we think about fund size.
At Theory, you have to first pick your strategy and then capitalize the business to be able to execute that strategy, which used to be the case for startups and no longer the case. But I think that’s really essential. And then you wanna put the probabilities on the side of you winning. What is the 75th percentile exit and the 90th percentile exit for a startup? What is that worth? And then what does that mean for our expected value for fund multiples?
Putting together all that math is, I think it’s an absolutely essential function or essential task for early funds, because the greater the confidence you can have in that business model, the more confidence limited partners will have in your ability to execute it.
Turner Novak:
What is, I mean, if you’re willing to share, what do you guys assume is the average outcome look like for an investment that you’re making?
Tomasz Tunguz:
I don’t wanna be too public about some of those numbers, but many of those numbers are public, and you can pull them from PitchBook and those kinds of things. We have our own very special way of underwriting. And, you know what? I think a key part of running a fund is you have to make exceptions, and there are exceptional companies that don’t fit the mold. You can’t have a portfolio full of them, but you can have some.
Turner Novak:
Yeah, that’s fair. And I know you make a lot of things with AI, like personally, like you’re just always messing around. What would you say is the coolest or most interesting thing that you’ve built or done, whether it’s related to Theory or just for fun?
Tomasz Tunguz:
The thing that’s daily useful is a podcast processor. It listens to 50 podcasts and then pulls out all kinds of interesting statistics and facts. That’s really a lot of fun.
Turner Novak:
What do you get from that? Does it give, take this episode and it would give you the five most interesting bullet points that you mentioned or something?
Tomasz Tunguz:
What are some interesting statistics? What are some counterintuitive perspectives? What’s the overall narrative? And, you know, that’s again parallelization. I don’t have the time to listen to 50 podcasts in a day. I’d run out of hours about halfway through.
Turner Novak:
And not get any sleep.
Tomasz Tunguz:
Right. So that’s not possible, but that’s really useful. I think the most effective uses of AI all boil down to parallelization. You have the big long list of something to do, and you don’t have enough time, how can you parallelize it? And the crazy part is the GPU, which is the chip that powers all of AI, is amazing at parallelization.
Turner Novak:
Yeah. And you do a lot for your content, too. People probably have come across you. You have a blog that I think you write a couple times a week. I don’t think it’s quite daily, but actually sometimes you do post multiple times a day, if I’m remembering based on the timestamps. So you post quite a bit. Do you use AI in the process of creating those and coming up with ideas?
Tomasz Tunguz:
AI is an incredible editor. When I first started writing, I hired an editor. I had this AP English teacher who taught me to love to write, a guy named Mr. Dunn. And so when I started writing, I really wanted to be graded like an AP English student. So I hired a wonderful person who did that. And now AI will do it for you exceptionally well.
So the amount of revisions with AI is, most blog posts ten years ago might have had two revisions or three revisions. Blog posts today have ten or 15 or 25 revisions.
Turner Novak:
So do you write something that’s maybe long-winded or not fully fleshed out, and then you have AI edit it? Do you have a series of prompts or Claude skills that you made or something where you’re banging through, you read it, you’re like, “Fix this, fix this”?
Tomasz Tunguz:
Most important thing is to create an outline, figure out the lead, the real story, and then the data points or the supporting arguments. It takes multiple versions. Even after 20 years of writing, Claude Code or whatever, Kimi K2 will say, “You buried the lead. You buried the most important part in paragraph 14.”
Turner Novak:
You’re saying you’ll say that to the AI? Like, you guys...
Tomasz Tunguz:
No, no, it will tell me. I’m just like, “Hey, critique this post,” and it will say, “You buried the lead.” And I feel like a freshman in high school. It’s just so basic. But it’s that consistent discipline.
Turner Novak:
So thanks again for coming on the show. This was awesome. I know you have to run. Where can people follow you, like, Twitter, LinkedIn, blog?
Tomasz Tunguz:
All three. T-T-U-N-G-U-Z. You can find me on LinkedIn and Twitter, and then tomtunguz.com is the blog.
Turner Novak:
Cool. We’ll throw links in the show notes for people to check them out.
Tomasz Tunguz:
Thanks for the conversation, Turner. Really enjoyed it.
Stream the full episode on YouTube, Spotify, or Apple.
Find transcripts of all other episodes here.
Really interesting breakdown of how AI is shifting from products to infrastructure. The point about models becoming more like pharma than software really stood out to me