Coatue co-founder Thomas Laffont on Hollywood mailrooms, Apple, Snapchat, and investing in the AI era

Summary

Thomas Laffont on Coatue, AI investing, and democratizing private markets

Thomas Laffont co-founded Coatue after a seven-year stint at CAA, where he went from the mailroom to representing talent including Jordana Brewster. That unlikely path — computer science at Yale, Hollywood agent, semiconductor analyst, then hedge fund investor — shapes how he thinks about pattern recognition, conviction, and staying in business across cycles.

From semis to AI

Laffont's entry into AI investing followed naturally from Coatue's semiconductor background. When Nvidia's data center momentum became visible, the team recognized the signal early because they had spent years covering chips through the mobile era. That foundation led Coatue to lead the Series B in Cerebras, which Laffont describes as a likely generational company. The infrastructure layer was the first call — he argues you didn't need to pick a model winner when the whole stack was going to benefit.

The harder question now is software. Laffont frames it plainly: you can construct a credible bull or bear case for almost any tech name today. Does AI make Workday's data more valuable, or does it allow a model to rebuild Workday from scratch? He doesn't have a confident answer, and says it's still too early to see clear separation between winners and losers.

What's different about this cycle

In prior tech shifts — cloud, mobile, the iPhone — there was meaningful institutional resistance. Carriers fought the App Store. Studios resisted streaming. Incumbents built hybrid clouds to avoid committing. This time, Laffont says, no one is head in the sand. Every board, every founder, every investor already accepts that AI will be transformative and is moving with urgency. The debate is execution and approach, not whether it matters.

He tracks ChatGPT's download share and user engagement almost daily as a proxy for OpenAI's momentum, treating it as a leading indicator for the broader AI spending cycle.

Investment philosophy

Coatue's analytical edge when it entered private markets was Excel-driven research in an industry that ran on Word. Laffont describes reverse-pitching Aaron Levy at Box with a deep public-market-style deck that Levy hadn't seen from a venture investor before. That differentiation eroded quickly as other firms replicated the approach or outsourced it to Bain.

The enduring principle is simpler: every thesis should compress into a few sentences. Laffont traces this to a meeting with Steven Spielberg during his CAA days, where Spielberg argued any great story can be pitched in three sentences. If you can't do it, you don't fully understand it. He applies the same test to investment theses, and says the best investors he's known — Stan Druckenmiller, Dan Loeb, his brother Philippe — all have that ability to cut to the essential pivot points.

Models are treated as a direct expression of the thesis. His Apple model, later passed to Coatue's current CIO Jaemean Rangwala, was built cell by cell with a specific color scheme. Where it held up well was unit volume. Where it was wrong: the team modeled iPhone ASP declining 5–10% annually because that's what consumer electronics always did. In reality, the price doubled from roughly $500–600 at launch to over $1,200 today.

Snapchat and private market expansion

Coatue's move into private markets started with later-stage deals — Evernote, Box — anchored by a rule of $100M-plus in revenue. The most successful private deal broke every rule: Laffont says Coatue led the Snapchat Series C at a valuation of around $1–1.5 billion. He called one of the fund's largest LPs to flag that it was off-brand, and the LP's response was simply: "Then do it. That's why we're investing in you."

Democratizing access

Laffont argues the cost of excluding retail investors from late-stage private companies is becoming socially and politically untenable. A generation that doesn't own homes, carries student debt, and feels economically threatened by AI has no stake in the upside of OpenAI or Anthropic. He's supportive of vehicles like Coatue's C Tech fund and structures like Trump accounts — he credits Brad Gerstner with spearheading that idea — as partial responses. Going public is not the only solution, but the transparency and democratized access it provides would, in his view, make a material difference.

The sharpest near-term tension he identifies: late-stage private valuations have run far ahead of what most retail-accessible secondaries funds are paying, meaning many investors are buying exposure at prices that have already priced in the upside. He doesn't resolve this cleanly, but the implicit message is that structure and timing matter as much as access.