All podcasts / Invest Like the Best / Summary

Josh Kushner - Concentration and Conviction

2026-02-18 - 65 min - source - Read full transcript

Patrick O'Shaughnessy (host)Josh Kushner

Key insights

Concentration, not diversification, is Thrive's central edge.

Kushner argues the only way to develop real context on a business is to spend a disproportionate amount of time on it, which is only possible with a small number of positions. This is the throughline from GitHub in 2014 through Stripe, OpenAI, and Databricks.

concentrated-conviction

The GitHub deal set the template: buy conviction quietly through chaos.

Thrive's initial $20M GitHub check (2014, Fund IV) came right before a leadership shakeup made the company look troubled from the outside. Because Thrive was the only party willing to buy out departing early holders' shares, it ended up owning roughly 10% of the company without anyone noticing.

concentrated-conviction

Druckenmiller validated Thrive's contrarian 2021 buying spree with a line that became a mantra.

After Thrive led a discounted Stripe round, backed Ramp, and led OpenAI's investment while much of the market was selling, Kushner sought out Stan Druckenmiller, who told him leaning in during a downturn is 'always the right decision, you just better fucking pick right' - conviction paired with real accountability, not blind bravado.

concentrated-conviction

Small teams are a deliberate structural choice, not a resource constraint.

Kushner says Thrive stays small because a large investment team creates a hierarchy where 'the investment team matters and other people matter less.' Instead, every function (legal, compliance, engineering, portfolio support) is staffed with people who feel ownership over decisions, which only works at small scale built on mutual respect.

venture-firm-culture

Insecurity is fuel to start, but it cannot be the destination.

Kushner observes that many people at Thrive are first-generation or come from backgrounds that gave them a chip on their shoulder, which drives early ambition. But he warns that operating purely to prove something to others is 'a never-winning game' - the goal has to shift toward passion or joy, or the drive becomes hollow and endless.

personal-resilience

'Never believe your own bullshit' became a firm-wide operating principle after Instagram.

After Thrive's early Instagram investment doubled in value within days, John Winkelried delivered that line to a 26-year-old Kushner expecting congratulations. Kushner had it printed on sticky notes for every desk at Thrive (then ~10 people) as a permanent check against overconfidence in a business built on predicting the future.

personal-resilience

Thrive Holdings inverts the usual AI investment thesis: disruption from the inside out.

After OpenAI's API reached $50M in revenue, Thrive pitched private equity firms on using it for internal efficiency and got little traction, so it started buying operating businesses itself. The thesis is that proprietary company data plus in-house domain experts to fine-tune models lets Thrive transform businesses from inside rather than compete with them from outside.

ai-investment-thesis

Thrive organizes its current AI bets into three buckets.

AI-native businesses (led by frontier labs like OpenAI, expected to consolidate around an oligopoly at the hyperscale level, plus domain-specific models in areas like robotics and drug discovery); critical infrastructure that benefits from AI rather than being disrupted by it (e.g., Stripe for agentic commerce, identity-layer companies); and 'holdings' businesses being transformed through applied AI internally.

ai-investment-thesis

The OpenAI investment decision crystallized at Kushner's kitchen table, not in a pitch meeting.

After first approaching Sam Altman in March 2022 and getting early access to a ChatGPT preview, Kushner describes staying up until 1-2am testing it, showing his wife Carly, and telling her it would change the world - a personal, unscripted moment he cites as the real turning point behind Thrive's $29B-valuation lead investment.

ai-investment-thesis

Thrive treats founder loyalty as close to inviolable, even at a cost to the firm.

Kushner says Thrive will not invest in competitive businesses once committed to a founder, describing the firm's word as more important than any individual trade. His summary line: 'kindness without toughness falters and toughness without kindness corrodes.'

founder-partnership

A24's model of enabling artists directly shaped how Thrive frames its role with founders.

Kushner draws the analogy explicitly: 'we're not Da Vinci, we're Medici.' Just as A24 focuses on supporting directors and actors rather than directing itself, Thrive sees its job as making founders' best work possible, not steering the company itself.

founder-partnership

His grandmother's Holocaust survival story is the explicit source of his perspective on hardship.

Kushner recounts his grandmother, whose name his daughter Ray is named after, digging a 600-foot tunnel with a spoon to escape a Belarusian ghetto after witnessing her mother and sister killed, then living with the Bielski partisans and in a refugee camp for four years before rebuilding in America. He says no professional stress will ever compare, and that nothing he accomplishes will equal what she survived.

personal-resilience

Books referenced

Meditations - Marcus Aurelius - Kushner cites the concept of being 'the same man' in good times and bad as a guide for staying level-headed as Thrive's reputation grew
The Fountainhead - Ayn Rand - A copy sits in Kushner's office; he points to Howard Roark's refusal to compromise his values, especially at trial, as something to aspire to

Media referenced

Zero to One - book - Referenced via a Peter Thiel comment that the chapter on the four types of monopoly should give 'scale' its own separate chapter
American Gangster - movie - Kushner quotes Denzel Washington's line that the loudest person in the room is always the weakest, on the idea that true power doesn't shout

Companies

Thrive Capital - Kushner's venture firm, founded 2011, now managing roughly $50 billion with a deliberately small investment team
GitHub - Thrive's early, formative concentrated bet (2014, Fund IV); Thrive kept buying secondary shares from departing early holders during a leadership shakeup and ended up owning about 10% unnoticed
Stripe - Thrive led a ~$1.8B check at a $50B valuation (50% down round) when Stripe needed to raise $6B or go public; used as the defining example of leaning in against consensus
OpenAI - Kushner approached Sam Altman in March 2022, got early ChatGPT preview access, and led Thrive's first investment at a $29B valuation; now a core, ongoing partnership
Instagram - Early Thrive investment (at a $500M valuation) that doubled in days when Facebook acquired the company; source of the 'never believe your own bullshit' lesson
Databricks - Cited alongside Stripe and OpenAI as a company Thrive helped raise large amounts of capital for
Isomorphic Labs - Kushner helped Demis Hassabis spin the AI drug-discovery company out of DeepMind as a standalone investable entity
A24 - Kushner draws a direct parallel between A24's Dan Katz enabling filmmakers and Thrive's approach to enabling founders
Long Lake - Alex Howman (a close friend of Kushner's, per the transcript) helped design the permanent-capital structure behind Thrive Holdings

Techniques and frameworks

Concentration over diversification - Thrive runs a deliberately small number of large, high-conviction positions rather than a broad portfolio, believing deep context on fewer companies compounds better than shallow coverage of many
Input orientation over output orientation - Thrive judges itself on process and culture (are we doing the work right) rather than chasing a specific fund-size or return target, on the belief that fixating on outcomes distorts decisions
Founders as heroes, firm as enabler ('Medici, not Da Vinci') - Borrowed explicitly from A24's model of enabling filmmakers; Thrive frames its job as making founders' best work possible rather than directing it

Summary

This is Patrick O'Shaughnessy's second conversation with Josh Kushner, founder and Managing Partner of Thrive Capital, recorded roughly two years after their first and released with deliberate distance from Colossus's cover story on Thrive. In the intervening period Thrive scaled to roughly $50 billion in assets with a very small team, and the conversation's spine is the tension between that growth and Kushner's insistence that concentration, not scale, is the firm's real strategy. He repeatedly returns to the idea that Thrive's edge comes from spending disproportionate time on very few companies rather than diversifying, tracing that philosophy from the 2014 GitHub investment (where Thrive quietly bought its way to roughly 10% ownership from departing early holders during a leadership shakeup) through the much larger, more public bets on Stripe, OpenAI, and Databricks.

A recurring thread is how Thrive holds conviction under public criticism. Kushner describes leading a discounted Stripe round, backing Ramp, and leading OpenAI's investment in 2021 while much of the market was selling, and being "eviscerated" for it. He credits a conversation with Stan Druckenmiller for validating that instinct with a line that doubles as the episode's title concept: leaning in during a downturn is always the right decision, but you have to actually be right. The OpenAI story gets particular attention, from Kushner's initial approach to Sam Altman in March 2022 through a late-night moment at his kitchen table testing an early ChatGPT preview and telling his wife it would change the world, to the firm's current three-bucket AI thesis (frontier labs, AI-resilient infrastructure, and internally transformed "holdings" businesses).

Kushner is explicit about why Thrive stays structurally small: he wants every function in the firm, not just the investment team, to feel ownership over outcomes, which he says only works with a small group built on mutual respect. He frames the firm's relationship to founders through an A24 analogy, "we're not Da Vinci, we're Medici," and describes near-absolute loyalty commitments, including a refusal to invest in a founder's competitors once committed. Several formative lessons recur as guardrails against the firm's own success: a line from John Winkelried after the Instagram sale ("never believe your own bullshit") that Kushner had printed on sticky notes for every desk at the firm, and a broader argument that insecurity can fuel ambition early on but becomes "a never-winning game" if it never matures into something more durable like joy or passion.

The conversation closes on more personal territory: Kushner's newborn daughter Ray is named for his grandmother, whose escape from a Belarusian ghetto during the Holocaust, including digging a 600-foot tunnel with a spoon and years in the forest with the Bielski partisans before rebuilding in America, he cites as the source of his perspective on professional stress. He also discusses his affinity for artists (Frank Ocean, the Beatles, James Turrell, Jony Ive) as models for intentional, non-output-driven work, and closes with the personal story behind the watch he has worn since a robbery in Guatemala during college.

Notable Quotes

"That's always the right decision. You just better fucking pick right." - Josh Kushner, quoting Stan Druckenmiller

"We're not Da Vinci, we're Medici." - Josh Kushner

"Never believe your own bullshit." - Josh Kushner, quoting John Winkelried

"In order to become king, God didn't give David a crown. He gave him Goliath." - Josh Kushner

"Kindness without toughness falters and toughness without kindness corrodes." - Josh Kushner