Coatue’s Laffont Brothers. AI, Public & VC Mkts, Macro, US Debt, Crypto, IPO's, & more | BG2

This episode dives deep into Coatue's 10th-anniversary insights, exploring the AI supercycle's profound impact on public and private markets, the evolving role of crypto, and the strategic recalibrations demanded by this technological shift.

A Decade of Insights and the AI Supercycle

• Brad and Bill, the podcast hosts, congratulate Philippe and Thomas Laffont of Coatue on the 10th anniversary of their "East meets West" event, praising its durable impact on founders and the tech industry.
• Philippe Laffont expressed significant optimism, greater than in previous years, centered on the AI supercycle. He referenced Coatue's slide 4 on this theme and slide 6, which provocatively questions when AI might reach 75% of total US market cap, drawing parallels to historical dominance by industrials and transport.
• Laffont highlighted that major tech trends, like AI, build upon previous waves (mainframes, PCs, internet, SaaS), leading to progressively larger impacts. "One of the reasons these trends get bigger, they're built on all the on top of each other," Philippe noted, emphasizing the compounding nature of technological advancements.
• The discussion touched on the reclassification of sectors, suggesting that industries like utilities or semiconductor equipment might increasingly be viewed as part of the Technology, Media, and Telecom (TMT) sector due to their foundational role in delivering tech products.
• Strategic Implication: Investors should consider the expanding definition of "tech" and how AI's foundational nature could reshape sector classifications and valuations, potentially identifying undervalued assets in enabling industries.

Diversification Beyond Mag 7 and Private Market AI

• The conversation shifted to the performance of AI-related stocks beyond the "Mag 7" (a term for seven mega-cap tech stocks: Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, Tesla). While the Mag 7 has been relatively flat year-over-year, AI-powered companies, AI-related software, and AI semiconductor firms have seen gains.
• Thomas Laffont discussed the private market, noting the "tremendous value accretion to the top AI companies like OpenAI or Anthropic." He highlighted CoreWeave, a company providing specialized cloud infrastructure for AI, as an example of a pure-play AI company recently going public, which Coatue is a shareholder in.
• CoreWeave is a specialized cloud provider focused on delivering high-performance compute, particularly GPUs, essential for AI/ML workloads.
• Thomas Laffont observed that many public companies have legacy businesses alongside their AI initiatives, making pure-play AI investments like CoreWeave particularly interesting.
• Actionable Insight: Crypto AI researchers and investors should look beyond established tech giants for AI exposure, considering emerging pure-play AI infrastructure and application companies, both public and private, that offer more direct exposure to the trend.

Crypto's Evolving Landscape: Bitcoin and Stablecoins

• Philippe Laffont addressed Bitcoin, acknowledging his firm's current lack of institutional investment but also his inability to ignore its significance. He humorously mentioned losing sleep over it.
• He contextualized Bitcoin's market cap (around $2 trillion at the time of discussion) against global net worth (approx. $450-500 trillion) and gold (approx. $15-20 trillion), pondering its potential for growth. "Could it be four? Could it be five? It's a real asset class, right?" Philippe mused.
• The discussion highlighted slide 18, suggesting Bitcoin's volatility might be decreasing, potentially making it more attractive to institutional investors. Laffont stressed the importance of distinguishing between Bitcoin, stablecoins (digital currencies pegged to stable assets like fiat), and "outcoins" (a likely reference to altcoins or alternative cryptocurrencies) and meme coins.
• Stablecoins are cryptocurrencies designed to minimize price volatility, typically by being pegged to a reserve asset like the US dollar or gold.
• The recent passage of stablecoin legislation was noted as a major positive step for the regulatory framework in US finance.
• Strategic Implication: The decreasing volatility of Bitcoin and positive regulatory moves for stablecoins could signal broader institutional adoption. Crypto AI investors should monitor these trends, as increased stability and regulatory clarity can de-risk investments and unlock new use cases at the intersection of AI and decentralized finance.

The Power of Mental Flexibility in Investing

• Brad emphasized the importance of mental flexibility for investors, contrasting it with dogmatism, especially relevant in volatile and rapidly evolving sectors like crypto. He noted that public market investors might develop this flexibility more readily due to the ability to act on changing views, unlike private market investors with longer lock-in periods.
• Philippe Laffont cited Stanley Druckenmiller, a renowned investor, who reportedly said, "I've made 120% of my money on obvious ideas and I've lost 20% elsewhere," underscoring the value of reconsidering seemingly obvious or established positions.
• Bill Gurley pointed out two new, significant factors for re-evaluating crypto: the shift in government stance from antagonistic to more supportive, and the high utility of stablecoins in business workflows.
• Thomas Laffont added an interesting perspective: "What if actually the alternative [to the US dollar] is Bitcoin?" This ties into discussions about government overspending and the search for alternative stores of value.
• Actionable Insight: Crypto AI investors must cultivate mental flexibility, continuously re-evaluating their theses as new data, regulatory shifts (like stablecoin legislation), and use cases emerge. The ability to adapt is crucial in a field where "obvious" truths can rapidly change.

Consumer AI: ChatGPT's Impact on Google

• The discussion turned to consumer AI, referencing slides 22, 24, and 26, which analyzed ChatGPT's impact on Google search. Thomas Laffont noted the anecdotal evidence of ChatGPT affecting Google search usage, with queries becoming increasingly commercial.
• Philippe Laffont explained Coatue's data science approach, which involved joining credit card receipt data (100 million per day) with email receipt data to track user behavior. "Data is useless unless you can join data sets that don't speak to each other. That is the unlock," he stated.
• The analysis showed that for users subscribing to ChatGPT (a large language model developed by OpenAI capable of generating human-like text), their Google page views, which were previously growing at 4% per year, declined by 8% to 11% peak-to-trough after subscription.
• Despite the entry of competitors like Google's Gemini, Elon Musk's Grok, and Anthropic's Claude, ChatGPT has shown remarkable resilience and continued engagement growth, both in the US and internationally (slide 24).
• Strategic Implication: The sustained engagement and user shift towards models like ChatGPT, even with increased competition, indicate a fundamental change in information consumption. Crypto AI investors should consider how decentralized AI models or AI-powered search alternatives could further disrupt incumbents and create new investment avenues.

The GPU Allocation Battle Among Hyperscalers

• Slide 27 presented a compelling analysis mapping cloud revenue market share against the share of Nvidia GPU allocations for major hyperscalers.
• Amazon (AWS): 44% cloud revenue share, but only 20% Nvidia GPU allocation.
• Microsoft: 30% cloud revenue share, ~30% GPU allocation.
• Google: 19% cloud revenue share, ~20% GPU allocation.
• Oracle: 5% cloud revenue share, jumped to 19% GPU allocation.
• CoreWeave: Emerged with 11% GPU allocation.
• Bill Gurley found this slide particularly insightful, highlighting Amazon's lower GPU share relative to its cloud dominance. This could suggest AWS is either behind in AI, pursuing a different hardware strategy (as mentioned by CEO Andy Jassy), or a combination. It also showcases Oracle's resurgence and CoreWeave's success as a focused pure-play.
• Philippe Laffont cautioned that the numbers are estimates but affirmed the trend of differential GPU access. He posed the question: "Are Nvidia GPUs a prediction of future cloud revenues?"
• The discussion also considered the potential for more hyperscalers, including companies like Anthropic or sovereign cloud initiatives, and the strategic choices between standardizing on Nvidia versus developing custom silicon (like Google's TPUs - Tensor Processing Units, specialized chips for machine learning).
• TPUs (Tensor Processing Units) are custom-designed application-specific integrated circuits (ASICs) developed by Google specifically for accelerating machine learning workloads.
• Actionable Insight: GPU allocation is a critical leading indicator of future AI dominance and cloud market share. Crypto AI investors should track these allocations and the underlying hardware strategies (Nvidia reliance vs. custom silicon) as they will significantly influence the compute landscape available for both centralized and potentially decentralized AI development.

Macroeconomic Backdrop: AI, Productivity, and US Debt

• Philippe Laffont addressed the macroeconomic environment, particularly US debt concerns, referencing Dan Loeb's adage, "If you don't do macro, macro does you."
• Despite concerns about a "debt spiral," the 10-year Treasury yield has remained relatively stable. Laffont presented an argument (slide 45) that an AI-driven productivity cycle, similar to the internet boom in the '90s, could lead to faster economic growth, lower inflation, and lower interest rates, potentially bringing the deficit-to-GDP ratio down.
• He noted that in 1993, experts predicted a rise in debt-to-GDP from 60% to 80%, but it actually fell to 40% due to unexpected productivity gains.
• Slide 51 illustrated that productivity growth of 2.5% to 3.5% annually for the next decade could substantially reduce the debt-to-GDP ratio from a projected 140% to potentially 80-100%. "At least we've been able to bookend what would productivity need to be," Philippe stated, acknowledging the analytical framework rather than a definitive prediction.
• Strategic Implication: If AI significantly boosts productivity as hypothesized, it could create a more favorable macroeconomic environment for growth assets, including Crypto AI. Investors should monitor productivity data closely as a key indicator of AI's broader economic impact and its potential to alleviate fiscal pressures.

The Reawakening of Private Markets: IPOs and M&A

• Thomas Laffont discussed the private markets (slides 60 and 61), noting a shift from an "incredibly unhealthy" 2021 environment (too much capital in, not enough out) to a more promising outlook. Signals are moving from "red to yellow and potentially green."
• A rebound in IPOs is occurring, with better performance from recent cohorts compared to the 2021 IPOs, which were down 40% within a year and 50% five years later (slide 71, excluding SPACs). This poor performance created significant "scar tissue."
• The discussion highlighted CoreWeave and Circle's strong IPOs and the importance of companies understanding public market expectations (slide 75).
• The M&A environment is also reviving, exemplified by Meta's deal to acquire 49% of Scale AI for a price reflecting 100% of its value. Thomas Laffont described this as Zuck's "bold move... urgency is now."
• Actionable Insight: The thawing of the IPO and M&A markets, partly fueled by AI enthusiasm, presents exit opportunities for early-stage Crypto AI investors and new public investment options. The quality and performance of these new listings will be crucial indicators of market health.

Meta's Scale AI Deal: Talent, Urgency, and Market Dynamics

• The panel delved into Meta's acquisition of a 49% stake in Scale AI at a $30 billion valuation (effectively paying $15 billion). This move aims to bolster Meta's AI talent and capabilities, with Scale's CEO assisting Meta's efforts.
• Brad questioned why Meta would pay 100% of the value for only 49% of the company. Thomas Laffont suggested two factors: the sheer size of the AI prize, making the investment seem proportionate to the multi-trillion dollar opportunity, and the rapid movement of the AI ecosystem.
• He cited Anthropic's rapid revenue growth (from $0 to $1B in about a year, then to $2B in three more months, and $3B two months after that) as evidence of the speed at which AI companies are scaling, creating urgency for players like Meta.
• Strategic Implication: The high valuations and aggressive M&A for AI talent and technology underscore the intense competition. Crypto AI projects needing to scale or integrate advanced AI may face similar pressures or opportunities for strategic partnerships.

The Future of Public Companies and the "Quasi-Public" Market

• Bill Gurley raised the trend of companies staying private longer, contrasting with the current IPO window. Thomas Laffont suggested reasons for going public include simpler financing (equity and debt), brand definition, transparency for customers, and attracting different investor types, including retail.
• Brad argued for the importance of major AI companies like OpenAI going public for regulatory scrutiny and "democratization of finance," rather than wealth creation being limited to a select few.
• Philippe Laffont introduced the concept of a "quasi-public" market for large private companies (over $5-10 billion valuation) that historically would have been public. He warned that if such companies avoid public market scrutiny, they will face regulatory oversight instead. "If you're not willing to submit yourself to sort of the sunshine... of the public markets, you're going to get it through a regulatory agency," he cautioned.
• Actionable Insight: The debate around private vs. public for large tech/AI companies has implications for governance, transparency, and investor access. Crypto AI researchers should consider the long-term structural impacts on market accessibility and regulatory approaches to highly influential private entities.

AI-Driven Efficiency and the "Golden Age of Margin Expansion"

• The discussion shifted to AI's impact on company efficiency, with Brad coining the term "golden age of margin expansion." He noted Mag 7 companies growing revenue over 20% compounded with opex (employee numbers) growing only at 2%.
• Slide 91 provocatively asked, "Has Microsoft reached peak employees?" It showed three phases: the ZIRP (Zero Interest Rate Policy) era of hiring for growth, the "get fit" era of headcount reduction (coinciding with GitHub Copilot - an AI pair programmer), and now the AI era.
• ZIRP (Zero Interest Rate Policy) refers to a monetary policy where a central bank sets its target short-term interest rate at or near 0% to stimulate economic activity.
• GitHub Copilot is an AI-powered coding assistant developed by GitHub and OpenAI that suggests code and entire functions in real-time within an editor.
• Slide 92 showcased AppLovin, a mobile technology company, which doubled its revenue while reducing headcount by over 35% by adopting an AI-first strategy. Thomas Laffont emphasized that AppLovin's CEO, Adam Foroughi, made these changes "to win and out compete."
• The potential employment impact was discussed, referencing Jevons Paradox – the idea that increased efficiency in resource use can lead to increased, rather than decreased, consumption of that resource. Philippe expressed faith that AI might create more, and more interesting, jobs.
• Jevons Paradox, in this context, suggests that as AI makes labor more efficient (reducing the "cost" of achieving an outcome), it could lead to the creation of new tasks, industries, and ultimately more jobs, rather than mass unemployment.
• Strategic Implication: AI's potential to dramatically improve corporate efficiency and margins is a major investment theme. Crypto AI investors should look for companies effectively leveraging AI for operational gains, while researchers can explore how decentralized AI systems might offer similar or novel efficiency benefits.

Strategic Advice for Founders and CEOs in the AI Era

• Thomas Laffont presented a quadrant (slide 98) offering advice to companies based on their growth rate (above/below 25%) and profitability (cash flow positive/negative).
• High Growth (>25%), Profitable: Time to consider being IPO-ready.
• High Growth (>25%), Burning Cash: Build a fortress balance sheet (e.g., OpenAI's $40B raise).
• Low Growth (<25%), Profitable: Avoid complacency. Explore how AI can re-ignite growth, possibly through M&A or new investments, even if it means becoming temporarily unprofitable.
• Low Growth (<25%), Burning Cash: Time to "reinvent." This could mean focusing on a nascent, high-potential product, open-sourcing technology, or making other fundamental strategic shifts. Thomas Laffont stressed, "It's the opportunity of looking at this moment and thinking what can I do."
• Brad highlighted a challenge for companies in the low-growth/burning-cash quadrant: a tendency to be conservative to protect existing revenue, even if that revenue is attached to a declining multiple. He advocated for increasing risk and embracing reinvention.
• Actionable Insight: This framework is valuable for Crypto AI projects assessing their own strategic positioning. The emphasis on reinvention and leveraging AI for growth, even from a position of weakness, is particularly relevant in a rapidly evolving technological landscape.

This Coatue analysis highlights AI's pervasive reshaping of markets and the critical need for strategic agility. Crypto AI investors and researchers must track AI-driven productivity gains, GPU dynamics, and the evolving institutional stance on digital assets to navigate emerging opportunities.