Tomasz Tunguz moved to California because of a book, and he got into tech because of raspberries.
Tunguz grew up in a Polish military family—his last name, he says with mirth, "comes from the region in Siberia where [his] ancestors were exiled by Nicholas II." But while living in D.C. in his early 20s, Tunguz was entranced by Jack Kerouac’s On the Road and stories of Carmel, Big Sur, and San Francisco.
"I have a machine learning background, and I really wanted to work at Google," said Tunguz. "I remember when I flew out to go interview at Google. They had these cafes where the food was free, and what stood out to me was this huge chafing dish, full of raspberries—and you could have as many of them as you wanted. I could not get over this plate of raspberries."
There’s sincerity in how Tunguz talks about it—they weren’t just raspberries, they were an idea about what life could be that he wanted to investigate. It tracks: Tunguz has seemingly always been someone entranced by ideas, and then inclined to test them. He’d go on to work at Google, and later became a VC at Redpoint, where he made his name as an investor in Looker and Kustomer (acquired by Alphabet and Meta, respectively). Tunguz also developed another distinction, as one of the most memorable early VC bloggers, known for his deep research and, well, his theories. Appropriately then, he named his new firm Theory Ventures when he struck out on his own in 2023. Last year, Theory announced its first fund at $238 million. And though that first fund isn't entirely deployed, recent LP demand sent Tunguz back to the marketplace: Theory recently raised its second fund at $450 million.
Tunguz, Theory founder and general partner, kindly spoke to me about the firm's new fund after I scooped the news from a filing earlier this month. What surprised him in this first year? It hasn’t been the marketplace he planned for, for one thing.
"Round sizes are up 42% since we started the firm," he said, adding that his financial model had suggested round sizes would decrease after the Federal Reserve raised rates in 2022.
Naturally, it all goes back to the AI boom, which has been bolstering round sizes. It’s been a double-edged sword—even though round sizes are bigger than anticipated, Tunguz and Theory have also been using AI themselves. Theory has built internal software that Tunguz says "has accomplished what a team of three or four might have before."
What's his theory on where the AI boom is right now? Tunguz suggests we’re at a plateau, with major changes on the way, especially for software giants. His case is this: AI is poised to disrupt entrenched workflows that major software incumbents like Salesforce and Zendesk are built upon, rendering them vulnerable. He likens these companies to "ossified," bone-like structures—rigid and unable to adapt quickly to the rapid changes AI will drive in white-collar work.
"Salesforce, Zendesk, Workday, ServiceNow, all those big companies, I think, are finally attackable in a way that they haven't been in 20 years," Tunguz told Fortune.
Tunguz is constantly calibrating for what change could look like. Another theory: "I’ve been wondering, In five years, will keyboards still have letters?" He thinks it's reasonably likely that, as computers increasingly listen and respond to voice commands for tasks like navigation and summarizing, manual typing will fade away.
Venture capital is a world filled with theories, and in some sense, having a theory is a VC's job. But Tunguz has made me think about why the industry needs its theories in the first place: In order to invest early at all, you need to have a view about the future. Tunguz also knows that he can't rely on theory alone, and says there's a necessarily balanced relationship between theory and practice. The goal isn’t perfection, but to be "mostly right" by constantly testing and refining ideas in the mold of the scientific method, he said.
"It's easy to pontificate about the future and end up in the wrong place if you aren’t trying the tools, talking to buyers, and testing it with people who criticize your ideas, who point out the flaws," Tunguz told Fortune.
So, what's unique about Tunguz is not that he has theories, but the public, data-driven forthrightness with which he presents them. It's a risk, really. He could be wrong, he could be copied. But in Tunguz's approach, there's thoughtful transparency—and confidence.
See you Monday,
Allie Garfinkle
Twitter: @agarfinks
Email: alexandra.garfinkle@fortune.com
Submit a deal for the Term Sheet newsletter here.
Nina Ajemian curated the deals section of today’s newsletter. Subscribe here.
