E30: Sunil Dhaliwal, Founder of Amplify Partners on Lessons from 27 Years in VC

Episode Summary

Executive Summary: Sunil Talib describes how venture and technology cycles repeat in different forms: internet-era hype was a true gold rush driven by vibes, while today’s AI boom is real but still creates short-term hype and consensus investing. He argues that early-stage venture must prioritize conviction, domain expertise, trust, and founder-market fit over committees and consensus, and that AI will shift value toward both foundation models and application-layer companies. He also explains Amplify’s evolution, its relationship-driven model, and why a dedicated bio fund now makes sense.

Main Topics: Internet-era bubble vs. today’s AI cycle (Priority: 5/5): Sunil contrasts the dot-com era’s demand-saturated, vibe-driven public-market frenzy with today’s AI boom, which he says is more supply-constrained and supported by real broad-based demand across consumer, enterprise, government, and sovereign buyers. Early-stage venture requires independent conviction (Priority: 5/5): He strongly rejects investment committees at early-stage firms, arguing that seed investing is about seeing things others don’t, backing controversial deals, and empowering high-agency investors to make calls without consensus drag. Where AI value will accrue (Priority: 5/5): He says value will likely accrue both to foundation models and to application-layer companies, with advantage coming from domain knowledge, product taste, focus, and trust rather than simply building a larger model. Trust, brand, and shipping as moats (Priority: 4/5): He argues that in a world of infinite software choices and rapidly replicable products, customers will increasingly choose vendors they trust to ship frequently, stay focused, and reliably meet evolving needs. Amplify’s origin and firm-building strategy (Priority: 4/5): Sunil explains how Amplify was built around technical founders, focused domains, and a no-committee decision structure. He credits concentrated focus, a strong platform team, and relationship compounding for the firm’s growth. Bio as the next breakout area (Priority: 4/5): He makes the case that AI plus modern biology tools, cheaper sequencing, and dual-trained talent are making biotech investable in a new way, especially at the intersection of life sciences and AI. Relationships as a compounding asset (Priority: 4/5): He emphasizes that authentic long-term relationships create unexpected downstream value for founders, investors, recruiting, and deal flow, and that these effects compound over years.

Key Arguments: Internet-era technology was overcapitalized and driven by story and momentum; many companies went from inception to IPO to bankruptcy in roughly three years. Today’s AI boom differs because demand is broad and real, while supply of compute, chips, and data centers is the current bottleneck. Early-stage investing cannot be reduced to analytics; it depends on judgment, contrarian insight, and the ability to back people and markets before evidence is obvious. Investment committees slow down the very thing early-stage firms are paid to do: make asymmetric bets based on conviction. Foundation model labs can capture value, but many of the best opportunities will be in application-layer businesses that combine technical capability with deep domain expertise and taste. As software becomes easier to generate, trust, brand, reliability, and shipping cadence become stronger differentiators, especially in enterprise purchasing. The so-called SaaS apocalypse is mainly a valuation and growth-rate problem for overvalued companies, not proof that AI will wipe out all software businesses. Bio is attractive now because new biological measurement tools and AI can improve speed, cost, and success rates in drug development and discovery. Relationship-building is not transactional networking; it is a long-term compounding mechanism that feeds fundraising, recruiting, customer access, and future deal flow. Amplify’s recruiting philosophy is to give junior investors real agency so they become better decision-makers and source better deals. Data Points: Years in venture: 28 years - Sunil says he has been in venture capital for 28 years. Amplify AUM: about $2.5 billion - He describes the firm’s scale today. Amplify age: about 14 years - He notes the firm has been around roughly a decade and a half. First fund size: $49.1 million - He recalls the size of Amplify’s first fund. Fundraising time for first fund: 1 year - He says raising the first fund took a year despite his prior track record. Maximum LP check size in first fund: 10% of fund / about $4 million - He capped LP concentration to avoid dependence on any single investor. AI and bio dedicated fund size: $200 million - He says Amplify raised a dedicated bio fund alongside its main fund. Main fund size mentioned: $400 million - He references the main fund raised alongside the bio fund. Bio investments already made: about 15 - He says Amplify had made around 15 AI-in-life-sciences/biology investments over a decade before launching the dedicated fund. Fastly seed timing: seed stage - He says Amplify backed Fastly at seed through a founder introduction and domain conviction. Temporal valuation: $5 billion - He cites Temporal’s most recent round as an example of a large technical infrastructure company built from domain expertise. Nvidia growth trajectory: $500 billion to $5 trillion - He uses Nvidia to illustrate long compounding and delayed value realization. Biome sequencing cost decline: 1,000th of what it was - He says sequencing costs have fallen dramatically, enabling new biology workflows.

Pivotal Quotes: "If you have an investment committee in a very early stage firm, you're an idiot." — Sunil Talib: His blunt explanation for why early-stage venture requires speed, judgment, and individual conviction rather than committee consensus. "The truism that the controversial deals are always the best deals is a truism for a reason." — Sunil Talib: He argues that the best early-stage opportunities are usually non-consensus before the market catches up. "We are 100% demand constrained on this boom. We are not supply constrained on this boom." — Sunil Talib: He explains why the AI cycle differs from the dot-com bubble: real demand exists, but compute and infrastructure are the bottlenecks.

Implications: For founders and investors, the message is clear: build around real domain insight, move fast, earn trust, and expect AI to reshape software, infrastructure, and biology. The winners will likely be the most focused, credible, and execution-oriented teams.