This is about investor Josh Kushner's philosophy: manage $50 billion but only bet on a few companies to go all-in and help founders. He thinks markets overreact to short-term noise—like when Stripe's valuation dropped, he invested $1.8 billion because people will keep buying more online. He also backs OpenAI, believing AI will be as transformative as the iPhone. Key holdings: OpenAI (he helped recruit an engineer at 2 AM), Stripe (bought the dip), and GitHub (early investment that returned 10x).
Josh Kushner in a deep-dive conversation with Thrive Capital, centered on the investment philosophy of extreme concentration and deep engagement. Thrive manages approximately $50 billion in assets with a very small team, renowned for its concentrated bets. Key investments include Instagram (doubling
Josh Kushner is the founder and managing partner of Thrive Capital, managing approximately $50 billion in assets with a remarkably small team. The core theme of this episode: how an investment philosophy of extreme concentration and deep engagement serves as Thrive’s differentiating weapon, and the fundamental paradigm shift in investing during the AI era. Josh Kushner argues that true differentiation stems from extreme conviction and focus, not diversification — "Concentration is central; it imposes a real burden on everyone, forcing us to be extremely disciplined and cautious."
Josh Kushner believes that concentrated investing is not a strategic choice, but a natural outcome of serving founders.
Thrive's concentration stems from a core belief: only by deeply engaging with a small number of companies can it deliver full value. Josh emphasizes: "Our desire to work with only a few people stems from our belief that to truly deliver full value, we can only commit ourselves to a limited number of things."
This philosophy was first validated in the GitHub investment. In 2014, Thrive led a $20 million investment in GitHub. A month later, CEO Chris Wanstrath replaced most of the leadership team, leading outsiders to believe the company was in trouble. But Thrive chose to engage deeply—the more they understood, the more confident they became. When other early founders sought to cash out, Thrive was the only investor willing to buy shares, ultimately holding about 10% of the company. Josh explains: "The only way to truly build contextual understanding of a business is to invest a lot of time. That's why we maintain such a concentrated portfolio."
Stripe's $50 billion round was the ultimate expression of concentrated betting. In 2022, Thrive led approximately $1.8 billion at a 50% discount to the previous round (a $50 billion valuation)—the largest check Thrive had ever written. Josh's underlying logic was remarkably simple: "If there's one thing I can predict until the day I die, it's that people buy more things online every year." He observed that the market at that moment was overly focused on post-pandemic short-term metrics (ARR multiples, margin structures), while ignoring long-term certainty. "When you have a product- and founder-oriented investment style, the ability to generate conviction largely depends on people's ability to solve problems. There will be bad quarters, bad years, but if you believe in the people running the business, everything will eventually be fine."
Josh learned a key lesson from Stan Druckenmiller. When Thrive faced industry criticism for its contrarian investments (Stripe, OpenAI, Ramp), Josh went to see Druckenmiller and said: "We've been doing the same thing—concentrating on what we believe are generational businesses. We sold a lot in 2021, and now we're heavily adding again. I feel like I'm running into a burning building carrying my most precious possessions, and everyone thinks I'm crazy." Druckenmiller's reply: "That's always the right decision. You'd better be fucking right."
Josh Kushner believes OpenAI is "the most important prize worth winning," and its pricing logic stems from a reimagining of the trillion-dollar market landscape.
Josh first met Sam Altman in March 2022. Prior to that, Thrive had been monitoring AI/ML but believed that only companies with existing distribution channels could build these models. A few months later, Sam gave him a preview of ChatGPT—"I couldn't ignore it." At 1-2 a.m., he showed it to his wife at the kitchen table: "I think this is going to change the world."
The pricing logic is Josh's most counterintuitive judgment. At the time, the industry consensus was that LLMs would become commoditized, open-source would prevail, and value would lie in the application layer. Josh countered: "We live in an era with companies worth $3-4 trillion, which could become $7-10 trillion by the end of this decade. So it's not inconceivable that a private company today could be valued at $500 billion or $1 trillion—especially the most defining company in this ecosystem." His core argument: Disproportionate value in public markets has already concentrated in fewer than 10 companies, and the same will happen in private markets.
Deep collaboration with OpenAI has reshaped Thrive's investment perspective. Josh admits: "This industry is not just about the deals you make, but also the ones you don't. Learning from inside the company has helped us avoid many things that ultimately won't be on the right side of history." Specifically, understanding what large labs will not do has pushed Thrive in specific directions—such as robotics, drug development, and other verticals.
A small story reflecting Thrive's service ethos: When OpenAI was fiercely competing with another lab for an engineer, Josh answered a call at 2 a.m., spoke directly with the engineer until 3:30 a.m., and ultimately convinced him to sign. Josh concluded: "If you have to choose between the most educated person, the most experienced person, and the person who wants it the most, always pick the one who wants it the most."
Josh Kushner categorizes Thrive's current investment focus into three types and introduces a new framework of "inside-out disruption."
Category One: AI-Native Enterprises. This includes general-purpose labs (with OpenAI as the leading consumer company), vertical-specific models (robotics/embodied intelligence, drug development/life sciences), and the application layer. Josh is extremely cautious about the application layer: "We are living in the moment when Apple launched the iPhone—they didn't know which apps they wanted to own themselves and which they were willing to let others own." The applications he favors must possess: reinforcement learning capabilities and memory systems to understand user preferences.
Category Two: Infrastructure. These companies will not only avoid disruption but will benefit from AI. Examples include Databricks, Stripe (powering agent commerce), and identity verification firms (as identity becomes more complex in an agent-driven world, companies like Wizz grow in importance).
Category Three: Holdings—Permanent Capital Platforms. This is Josh's most distinctive innovation. He admits: "One of my biggest insecurities is that Thrive looks fundamentally different from the companies we invest in. In financial history, most innovations eventually get arbitraged away." Holdings originated from a serendipitous event: after investing in OpenAI, the Thrive team pitched its API to private equity firms but received no response, so they decided to do it themselves.
Core logic of "inside-out disruption": Under the reinforcement learning paradigm, you need proprietary company data and internal company experts to fine-tune models. Therefore, disruption will occur from within, not from outside. Holdings, as a permanent capital vehicle, enables Thrive to acquire businesses and hold them indefinitely—"If you have a differentiated cost of capital perspective and can transform these businesses the way you want, you'll ultimately want to hold them forever."
Josh credits Alex Halman (Long Lake) for helping design this structure and notes that Holdings' North Star metric is "creating a differentiated, scalable cost of capital."
Josh Kushner believes that the core of Thrive's culture is "the people who want it the most" and "never believe your own bullshit."
The Power of Small Teams: Josh insists on keeping teams small because "the only way to have a team is to have a small team, and the only way to have a small team is to have a group of people who respect each other." Every function at Thrive (legal, finance, compliance, engineering) has "10x talent," and they feel accountable for every decision.
A Lesson from the Instagram Experience: When Instagram raised funds at a $50 million valuation (Thrive invested $12 million, a significant portion of its $40 million fund), it was sold to Facebook for $1 billion two days later. A 26-year-old Josh expected congratulations, but John Winkelried (now TPG CEO) called and said: "I'm going to give you the greatest lesson—never believe your own bullshit." Josh wrote it on a sticky note and placed it on every employee's computer.
His grandmother's story shaped Josh's perspective. She grew up in a small town in Belarus, was confined to a Jewish ghetto, dug a 600-foot tunnel with a spoon to escape, witnessed the murder of her mother and sister, fled into the forest to live with the Belsky partisans, and eventually made it to the United States. Josh says: "No pain I have ever experienced can compare to what she went through. More importantly, no achievement I have ever made in my life can surpass what she accomplished."
On Trust and Values: "I default to trusting people. But if someone breaks that trust, they're out." Josh emphasizes that Thrive does not invest in competing businesses and will not break promises for the sake of a deal. "Kindness without toughness is weak, and toughness without kindness is corrosive."
| Position | Analyst Stance | Key Data |
|---|---|---|
| OpenAI | Bullish (Core Holding) | First contact in March 2022, led the $29 billion valuation round; API revenue started from $50 million |
| Stripe | Bullish (Core Holding) | Led the $50 billion round in 2022 (50% discount from the previous round), invested approximately $1.8 billion |
| GitHub | Bullish (Exited) | Invested $20 million in 2014, ultimately held about 10% stake |
| Bullish (Exited) | Invested $12 million at a $500 million valuation, sold to Facebook for $1 billion two days later | |
| Databricks | Bullish | Invested as infrastructure, specific amount undisclosed |
| Ramp | Bullish | Led the round at a price lower than the 2021 round |
| Isomorphic Labs | Bullish | Co-founded with Demis Hassabis, focused on AI drug discovery |
| A24 | Bullish | Invested to learn its "artist-first" model |
| Wizz | Bullish | Invested as identity verification infrastructure |
1. "Concentration is the core; it places a genuine burden on everyone" (Josh Kushner) — Thrive's concentration is not a strategic choice but a natural outcome of serving founders; only by deeply binding with a few companies can it deliver full value.
2. "Predicting the long term is much easier than predicting the short term" (Josh Kushner) — When investing in Stripe, the market was overly focused on post-pandemic short-term metrics, while Josh only focused on the long-term certainty that "people buy more things online every year."
3. "Never believe your own bullshit" (John Winkelried) — After Instagram's success, Josh posted this quote from Winkelried on every employee's computer; Josh believes "intellectual honesty and self-awareness" are the greatest sources of correct decision-making.
4. "We are not Da Vinci; we are the Medici" (Josh Kushner) — Thrive's role is that of an enabler, not an artist; the core lesson learned from A24 is "input orientation" — how to create the best working environment, rather than pursuing a specific output.
5. "Inside-out disruption" framework (Josh Kushner) — In the era of reinforcement learning, disruption requires a company's proprietary data and internal experts, so it will occur from within rather than from outside; Holdings, as a permanent capital vehicle, enables Thrive to buy companies and hold them forever.
6. "If you have to choose between the most educated, the most experienced, and the most hungry, always pick the most hungry" (Josh Kushner) — Thrive's culture is one of "those who want it the most"; answering the phone at 2 a.m. to help recruit engineers for OpenAI is a microcosm of this culture.
7. "Before making him king, God did not give David a crown; He gave him Goliath" (Josh Kushner, quoting) — Every difficult experience prepares you for the next; Josh believes "post-traumatic growth" is more valuable than "post-traumatic stress."
8. "Real power does not shout; it acts quietly and decisively" (Josh Kushner) — Learned from the OpenAI experience: power corrupts, but "gratitude and humility" are the way to deal with it; quoting Denzel Washington: "The loudest person in the room is always the weakest person in the room."