Episode Summary
Executive Summary: Kleiner Perkins GP Mamoon Hamid argues AI is a supercycle comparable to the internet, with the biggest venture opportunities in application-layer tools that supercharge scarce, high-value labor like doctors, lawyers, and developers. He says the market is crowded and overinvested in middleware, pricing is frothy, and founders win through exceptional products, technical depth, and market creation, not data-driven consensus or top-down TAM exercises.
Main Topics: AI as a supercycle and venture opportunity (Priority: 5/5): Hamid frames AI as the most exciting time to be alive and compares it to the internet boom, but amplified. He believes the infrastructure shift will create trillions in value and that venture should focus on where AI creates durable business value. Where to invest: application layer over middleware (Priority: 5/5): Kleiner is concentrating on application-layer companies that augment scarce labor such as doctors, lawyers, and developers. Hamid thinks the middle layer between foundation models and applications is somewhat overinvested and more vulnerable to commoditization. Market creation and product-led winners (Priority: 5/5): Hamid repeatedly emphasizes backing companies that create markets, like Slack and Figma, rather than merely competing in existing ones. He prizes deep technical founders, strong product taste, and usage-driven evidence of pull. Pricing, reserves, and venture discipline (Priority: 4/5): He warns against irrational early-stage pricing, says reserve management is both art and science, and argues that funds must preserve ownership math. He allows for occasional rule-breaking via a 'YOLO bucket' but not as the norm. AI economics, labor replacement, and revenue scaling (Priority: 4/5): Hamid argues AI should be valued as labor and capability, not just software seats. This supports much higher per-seat pricing and very fast revenue ramp, especially in workflows with acute labor shortages. M&A, IPOs, and liquidity constraints (Priority: 4/5): He describes M&A as slow and IPO markets as largely closed, partly due to regulatory and election uncertainty. Still, he expects a better IPO year ahead if a major company debuts and opens the window. Firm identity, founder selection, and board behavior (Priority: 4/5): Kleiner positions itself as boutique: small team, early-stage focus, and concentrated follow-on support. Hamid dislikes voting structures, values conviction, and wants boards to focus on only one or two critical issues at a time.
Key Arguments: AI venture returns will come from applications that materially improve scarce, high-value labor rather than from generic middleware or token distribution alone. The best companies are often those that create a category or market, giving them the ability to define the playing field and win it. Competitive markets are acceptable, but the strongest signal is a product that demonstrates real usage and creates adjacent expansion opportunities. Early-stage pricing should not be driven by hype; ownership and fund math matter, and overpaying systematically breaks the venture model. Founders should not maximize capital raised at any price; the right investor-partner fit and manageable future rounds matter more. Great founders cannot always overcome bad markets, especially where structural economics are poor or customer dynamics are weak. Venture value-add is real when advice is the right advice; the wrong advice can destroy value, especially on boards. Liquidity decisions should be made around perceived local maxima, balancing current market enthusiasm with strategic acquisition value. Kleiner’s advantage comes from reputation, founder choice, and a small team that can act with conviction rather than consensus. AI pricing and revenue growth can look 'sugar-high' but still be durable if the product is effectively selling labor or outcomes.
Data Points: Top U.S. jobs referenced: Doctors, lawyers, developers - Hamid identifies these as scarce, highly skilled roles where AI copilots can create the most value. Brex startup usage share: 1 in every 3 U.S. startups - Mentioned in sponsor read to illustrate Brex’s startup penetration. FDIC protection multiplier: 20x standard FDIC protection - Brex account offering via program banks. Token price decline: 200x in the last 18 months - Hamid cites rapid token cost deflation in AI. Potential token price decline next 2 years: 10x to 20x - His estimate for continued cost decreases. Global GDP: About $100 trillion - Used to frame the scale of labor and technology spend. Share of GDP from labor: 50% to 60% - Hamid’s rough estimate of labor’s share of global GDP. Share of GDP from technology: Roughly 15% - Used to argue AI can expand technology’s economic footprint. Possible technology share of GDP: 15% to 20% - Implied expansion over the next decade. Additional annual technology spend potential: $10 trillion - His estimate of new annual spend created if tech share rises. AI capex question: $200 billion capex -> $600 billion revenue - Hamid argues the core question is whether AI infrastructure spend can generate 3x revenue. LLM company headcount example: 1,600 employees - He cites OpenAI’s size to explain why it cannot build every application itself. Kleiner early-stage fund size: $800 million - Describes firm structure and strategy. Kleiner growth fund size: $1.2 billion - Growth capital used partly to support existing winners. Early-stage fund deployment: ~15 months - He says one fund deployed quickly after team formation. Typical fund company count: About 35 companies per fund - Used to explain reserve strategy. Reserve split heuristic: 60% initial / 40% follow-on - Rough rule for capital deployment across a company’s lifecycle. First-check to total lifetime investment: About $25 million - Illustrative total per early-stage company over time. Initial check size example: $15 million - Example of the first portion of a total $25 million commitment. Follow-on reserve example: $10 million - Example of capital reserved for later rounds. Figma early valuation: ~$100 million post - Hamid invested before major revenue inflection. Figma revenue state: A few hundred thousand dollars in revenue - He notes the company was early but showed strong usage. Slack early valuation: $250 million post - Hamid invested when Slack had little revenue. Slack ARR at investment: ~$500K ARR - Early Slack investment context. Slack outcome: $27 billion - Referenced as ultimate sale value. Initial engagement at Slack: ~10,000 users; about one-third daily for multiple hours - Usage patterns that supported conviction. Glean investment valuation: $35 million post - Example of paying up for a strong founder. Aleph investment valuation: ~$35 million post - Another example of fair pricing for a strong repeat founder. Tally loss: ~$30 million - Hamid’s largest recent capital loss after the company shut down. Box near-death bridges: Three bridge rounds - He recounts supporting Box during the financial crisis. Box bridge valuation: $25 million - Price at which incremental rescue capital was invested. Box public market data: ~$1 billion revenue and $4.5 billion market cap - Used to discuss public market valuation and venture returns. Yammer acquisition: $1.2 billion - Example of selling near a perceived local maximum. Example 70x exit: $3 million to about $170 million - Illustrates large multiple but modest IRR over a long hold period. Holding period example IRR: ~15% - Hamid notes long-duration winners can still underwhelm on IRR.
Pivotal Quotes: "I love products that create markets. Slack created a market. Figma created a market." — Mamoon Hamid: He explains his preference for companies that define a new category rather than compete in an existing one. "In the age of AI, we have to think about what are we doing? We're not just providing software, we're providing labor, we're providing capabilities." — Mamoon Hamid: Used to justify higher AI pricing and focus on workforce augmentation. "It's about the founders wanting to work with you." — Mamoon Hamid: He argues venture differentiation comes from reputation and founder choice, not just deal access.
Implications: For investors, the message is to back market-creating, technically deep founders in AI application layers that replace or augment scarce labor. For founders, fit, product quality, and sensible pricing beat hype; for the industry, liquidity and M&A remain constrained, so discipline matters.