Thrive Capital’s Kushner breaks ranks with Silicon Valley on AI investing

Joshua Kushner has a message for his fellow venture capitalists: slow down and stop chasing every AI opportunity that crosses your desk. In Thrive Capital’s inaugural investor letter, which Bloomberg obtained, the founder of the New York-based firm breaks publicly with the conventional wisdom that has long defined how Silicon Valley VCs deploy capital in hot markets.

The disagreement centers on what constitutes prudent investment discipline when technology appears poised to reshape entire industries. While acknowledging that artificial intelligence represents “a magnitude of opportunity” that “is difficult to overstate,” Kushner warns that excitement can become a liability. “It would also be a grave error in our minds to let excitement weaken our investment discipline,” he wrote. “Within Silicon Valley in particular, the industry can become fixated on hyperincremental technological turns rather than where the technology ultimately leads.”

A Different Bet on AI

Thrive’s approach stands apart from the dominant venture model that emerged from Sand Hill Road and now extends across the technology ecosystem. Rather than spreading bets across dozens or hundreds of companies in hopes that a few mega-winners offset numerous failures—the “spray and pray” strategy Kushner implicitly criticizes—Thrive concentrates its firepower. According to Bloomberg, approximately 90 percent of capital deployed in each fund flows to just the top 15 investments. That degree of concentration requires conviction and comes with commensurate risk if judgments prove wrong.

The model inverts Marc Andreessen’s foundational venture thesis that VC is fundamentally a business of “outliers”—the rare, transformative companies that generate returns so outsized they dwarf a portfolio’s many losses. Under that framework, the occasional OpenAI or Stripe justifies backing dozens of companies destined to fail. Kushner rejects this as dogma. “We are independent because markets move between fear and enthusiasm, and neither is a substitute for judgment,” he wrote, positioning Thrive’s restraint as intellectual independence rather than conservative timidity.

The Insider-Outsider Thesis

Kushner also challenges Silicon Valley’s preferred narrative about how disruption unfolds. Most VCs assume new technologies overthrow incumbents from the outside, with nimble startups outmaneuvering established players. Thrive, by contrast, believed “many would be transformed from the inside out.” This distinction shapes which companies the firm backs and how it expects value to compound. It also explains why Thrive has been willing to work with existing large enterprises through its operating subsidiary rather than placing all bets on insurgent startups.

The Thrive Holdings Model

Read also: Fermi Appoints TerraPower Veteran as New CEO, Stock Surges 23% · AI Code-Testing Startup Blacksmith Raises $45M at $550M Valuation · TechCrunch Disrupt 2026 targets the challenge of building in the AI era · Fusion Startups Attract Record Private Capital as Timelines Accelerate

That philosophy crystallized in December 2025 when OpenAI took an ownership stake in Thrive Holdings, a spinout designed to acquire operating companies and retrofit them with AI capabilities. The arrangement goes beyond typical venture investment: OpenAI commits dedicated employees to work alongside Thrive’s teams, embedding the AI lab’s expertise directly into the companies being transformed.

Thrive Holdings has acquired more than 70 businesses staffed by a 35-person engineering team. The unit has generated tangible applications demonstrating how AI agents can alter business operations. An accounting platform developed by the division can produce tax returns 30 percent faster while maintaining 98 percent accuracy. A separate IT services firm deploys AI agents that independently resolve half of incoming help desk tickets without human intervention. These aren’t theoretical demonstrations but operational systems already handling real workflows.

Portfolio Depth Over Breadth

The concentrated strategy has delivered results in part because Thrive identified and secured early positions in some of technology’s best-performing companies. Its 2022 early-stage fund, capitalized at $516 million, made foundational investments in OpenAI, Anduril, and SpaceX. As of the end of June, that single fund’s value had appreciated to $3.7 billion, Bloomberg reported. Over 15 years of operation, Thrive has consistently increased its stakes in these companies rather than taking profits and rotating capital elsewhere.

Beyond the obvious mega-winners, Thrive’s portfolio encompasses Wiz, the cloud security company; Ramp, the corporate card platform; and Stripe, the payments infrastructure business. Kushner also highlighted a recent seed investment in Essential AI, a new research lab founded by Ashish Vaswani, who led the research behind the foundational “Transformers” paper that underpins most of today’s AI systems. The ability to back Vaswani’s lab reflects the kind of concentrated focus Kushner describes: deploying capital where conviction is highest rather than spreading dry powder across every promising cohort.

Returns and Scale

The financial picture Kushner disclosed in his letter demonstrates that the concentrated approach has generated substantial returns. Thrive manages $60 billion in assets across its various funds and investment vehicles. He reported a gross internal rate of return across all funds of 41 percent, with a net IRR of 33 percent after fees. Over the past 12 months alone, Thrive returned more than $1 billion of liquidity to its investors, signaling active harvesting of earlier positions.

Kushner also indicated that additional liquidity events are likely forthcoming. “There may be an opportunity for billions of dollars in additional liquidity in the coming quarters,” he wrote without naming specific candidates for exits. However, the trajectory is evident: SpaceX is preparing for a public offering, and OpenAI has publicly stated its intention to pursue a path to going public. Both represent potential multibillion-dollar liquidity events for early shareholders like Thrive.

Comparative Performance

These figures deserve context. The opposing philosophy—Andreessen Horowitz’s willingness to cast a wider net and accept higher failure rates in pursuit of outsized outlier returns—has also proven extraordinarily lucrative. According to returns reported through 2025, Andreessen Horowitz returned $25 billion to its investors between 2009 and that date. Both philosophies, in other words, work when executed by elite operators with access to the industry’s best deal flow.

The Access Question

That caveat contains an implicit critique of Kushner’s positioning. Thrive’s ability to concentrate capital and achieve outsized returns depends partly on access that most emerging VCs simply cannot replicate. Kushner is the son of a prominent New York real estate family with inherited wealth and networks that enabled him to build and sustain Thrive over 15 years. Smaller seed funds and scrappy emerging managers cannot realistically adopt the concentrated model without the cushion of vast capital pools and established relationships with founders and limited partners.

His broader point—that Silicon Valley has become intoxicated by AI’s potential and is overlooking fundamental distinctions between fast growth and exceptional quality—nonetheless carries weight. As Kushner framed it: “Not every fast-growing business is exceptional. And not every exceptional company is a great investment at every price. Our responsibility is to maintain those distinctions.” That maxim applies whether a firm is deploying concentrated capital or spreading bets across a portfolio, and it stands as a useful reminder during periods of sector-wide exuberance.