Inside Silicon Valley’s VC Playbook | WTF is Venture Capital? - 2025 Edition | Ep. 24

Three VCs break down the 2025 playbook: AI turned data-rich, high-CapEx sectors investable while demographics and digital fatigue open new verticals.

Core Thesis

• Rising rates plus ChatGPT created instant demand for efficiency, making manufacturing, agriculture and robotics attractive (2:05-3:06).
• Second-mover advantage now favors execution over first-mover distribution in AI (2:31-2:52).

Investment Hotspots

• Longevity & senior living: community-driven models like The Villages deliver measurable lifespan gains (1:30-2:10).
• Beauty & male grooming: Ozempic and Turkey hair-transplant boom show attention + scarcity economics (2:10-2:31).
• Data centers and prediction markets (Kalshi, Polymarket) ride compute growth and regulatory tailwinds.

Societal Headwinds

• Birth rates keep falling as phones satisfy emotional needs; digital detox and new rituals are rising (0:50-1:20).

00:00:00Major Topic: Episode Overview and Guest Introductions

00:00:00 Subtopic: Framing the Conversation

• This episode features a group of venture capitalists (VCs) and operators (Deedy, Nikunj, Niko) gathered with host Nikhil Kamath to discuss the VC playbook, sector tailwinds, and long-horizon bets in tech and AI. The central theme is identifying where tailwinds exist for the next decade and how to navigate a rapidly changing landscape.
• Definitions inline: VC = venture capital investor who funds early-stage companies (seed, Series A, etc.). Seed/Series A are early rounds; later rounds include Series B, C, etc.

"How you guys see different sectors evolving... Where can me, as either an investor, or somebody looking to start my career as a young man or woman have enough tailwinds that it makes sense to spend the next decade in?" (Nikhil Kamath, 00:00:00)

00:10:03Major Topic: Personal Journeys into VC

00:10:03 Subtopic: Debarghya Das (Deedy) — Background and current focus

• Deedy hails from Calcutta, India; his real name is Debarghya Das, and the nickname ‘Deedy’ stuck in high school due to pronunciation challenges. The moniker became his professional name.
• Education: Cornell graduate with CS (Computer Science) and a master’s degree.
• Early career: Meta (Facebook) in New York for a year; then Google in New York with international exposure (Israel, Bangalore).
• Founding experience: On the founding team of Glean, a GC-backed company; progress from engineering to managing and then leading the AI product line.
• What Glean did: Enterprise search turned into an enterprise assistant/knowledge-finding tool for large companies. Examples include onboarding, holiday policy queries, and org-structure/document search.
• Glean scale: Glean reached about a $7B valuation; Deedy spent four years there.
• Current role: Joined Menlo Ventures; Menlo is a long-established fund (48 years) investing out of a $1.5B fund with about $7B under management across the wider platform.
• Focus areas at Menlo: Seed and Series A deals primarily in AI, SaaS infrastructure, with broader exposure to biotech and consumer practices; broader fund activity up to Series B/C.
• Anthropic: Deedy highlights Anthropic as a landmark investment and notes it began as a foundational-model company following departures from OpenAI; Anthropic focused on enterprise-facing, text-based models with a strong emphasis on coding and reasoning.
• Anthropic summary: Enterprise-facing, text-to-text models; primary business centered on enterprise customers with a consumer app as a secondary line.

> "Glean started as an enterprise search business... evolved into an enterprise assistant and knowledge finding tool for bigger companies" (Deedy, 00:00:00)

01:20:03 Subtopic: Nikunj — Background and current focus

• Nikunj’s path: Moved to the U.S. 17 years ago; studied at Urbana-Champaign (Electrical Engineering and Computer Science); 13 years in the Bay Area.
• Career arc: Operator for 11 years; first PM roles, four startups (early-stage founder experiences); involved with Hall (enterprise comms) and Spykes (Shyp-like; last-mile delivery) before OpenDoor.
• Opendoor context: Real estate tech platform enabling cash offers for homes via ML-driven valuations; the model reduces transaction friction and 6% agent fees—highlighted as a dramatic shift in home transactions in the U.S.
• IPO experience: Firing 35% of the company at Opendoor during 2020 crisis; IPO six months later; a crucible for learning in high-velocity, high-uncertainty scenarios.
• Meter: Wired into networking hardware enabling customers to provision spaces with no upfront hardware costs; pay-per-square-foot model with maintenance handled by Meter.
• Transition to investing: Explored private equity/VC with a tilt to seed and Series A only; joined FPV Ventures (Founder's Point of View) for a concentrated, founder-first approach; fund size around $525M; focus across biotech, quantum computing, AI chip design, etc.
• FPV ethos: Generalist fund with a founder-first lens; emphasis on early-stage bets with a preference for deep technical founders who build for themselves and then scale.

"We mostly care about the founder and the team that we meet" (Nikunj, 01:25:00)

02:40:00 Subtopic: Niko — Background and current focus

• Niko’s origin: From Greece; split time between Greece and the UK; settled in Silicon Valley since 2009 after Stanford; joined General Catalyst and focused on first-time founders with bold ideas.
• Investment stance: Focus on young founders (Gen Z) who are AI-curious and building for themselves; shift from consumer to enterprise and AI-centric opportunities; emphasis on trajectory over sector specialization.
• Geographic stance: Spends time between Silicon Valley and New York; values the founder’s potential more than sector alignment.
• Collaboration and networks: Mentions Neeraj of Venture Highway; emphasizes cross-border collaboration and India connections.

"I spend a lot of time with technical young founders... building products for themselves" (Niko, 02:40:00)

00:40:00Major Topic: The Crossroads — Sector Outlook and Tailwinds

00:40:00 Subtopic: The Central Question — Which sectors to pursue or avoid over the next decade?

• Deedy argues that venture investing is biased toward sectors with scalable, favorable returns, which often excludes extremely high-Capital Expenditure (CapEx) sectors because of longer ROI horizons. However, he acknowledges exceptions (e.g., real estate tech like Opendoor) and notes that the lens should be broad—every sector has potential.
• On a broader scale, Deedy suggests staying open-minded about sectors; manufacturing, which may appear non-sexy, can yield profound business models given labor dynamics, tariffs, and manufacturing localization challenges; AI can unlock value where data is abundant.
• Nikunj pushes back on sector bias: venture returns can be achieved in any sector if the founder is exceptionally capable; argues AI-driven data-driven sectors have become investable due to efficiency gains; critiques cycles of high-capex and long sales with nuance around financing strategies.
• Niko emphasizes a preference for highly technical, early-stage founders rather than chasing faddish markets; notes that the pace of change and AI-driven transformation create tailwinds across multiple industries; underscores the importance of founder quality and execution capability.
• The collective view: No hard sector ban; the rate of change and the ability to solve real, scalable problems with robust unit economics matter more than the sector label; AI is a macro tailwind that extends to traditionally “uninvestable” sectors when approached with the right data-driven thesis.

"AI has actually made most industries that were uninvestable suddenly investable" (Nikunj, 20:05:00)

01:00:00Major Topic: AI Trends and the 10-Year View

01:00:00 Subtopic: Three Nuanced AI Trends Highlighted

• Reinforcement Learning (RL) trajectories: RL is highlighted as a key curve for enabling models to learn via rewards in simulated environments; the idea is to move beyond static data and into interactive learning contexts where the model improves through feedback loops.
• Data constraints and China: The discussion emphasizes that China’s AI capabilities are highly competitive, often leveraging less resource-intensive processes; the point is that frontier models from China can rival or complement Western approaches, underscoring a more layered global AI landscape.
• Evals and the real-world utility of models: The concept of evals (evaluations) is introduced—assessing model performance in real-world tasks with corner cases and expert oversight; the value of human experts as a validation or correction layer to refine model outputs.
• Embodied intelligence and data sourcing: The group explores data strategies for training embodied agents (robots, autonomous systems) and discusses data collection (synthetic vs. real-world) and labeling, including offshoring the labeling work to cost-advantaged regions.
• The data-labeling economy: The crowd-labelling economy (e.g., Philippines, India) is discussed as a backbone for training data, with scaling implications for AI-driven businesses across healthcare, robotics, and defense tech.
• The near-term horizon for AI agents: The panel notes that agents capable of longer, multi-step reasoning (ten hours of sustained operation) are becoming plausible; the capabilities uplift expands AI’s utility from mere task automation to long-form research and planning.
• The broader spectrum: They discuss the boundless potential of AI-enabled automation across labor-intensive industries, with particular attention to labor replacement and augmentation.

"But the point being is, like, we've already reached here and everybody’s kind of just like, ‘Whoop-di-la’, walking away next" (Niko, 01:30:00)

01:40:00 Subtopic: AI-Ecosystem Hot Spots and Market Signals

• Cursor and AI code generation: Cursor is named as a notable player in AI code generation and automation; the ecosystem is rapidly expanding with a variety of AI-enabled tools (receptionists, coding assistants, etc.).
• AI scribes in healthcare: The market for AI-assisted medical documentation and scribes is highlighted as a rapid-growth area with double-digit ARR growth (started recently and expanding quickly).
• Enterprise AI adoption: Enterprise AI software is highlighted as a fertile field because enterprise buyers require cost efficiency and productivity improvements, driven by macro tailwinds like rising costs and the need for efficiency.
• RL as a service and other services: The conversation touches on RL as a service and other tooling that lowers the barrier to entry for developers and enterprises to leverage advanced AI capabilities.
• The ongoing importance of real-world data vs. synthetic data: The panel debates the role of synthetic data in RL environments and the tradeoffs between synthetic and real-world data for robust model performance.

"Cursor’s doing AI code gen... AI receptionist, voice applications... crowded with hot markets" (Niko, 02:40:00)

02:20:00Major Topic: Sector-Specific Trends and 10-Year Forecasts

02:20:00 Subtopic: Identified Trend Set — Which areas will grow fastest in the next decade?

• Education and outcome-focused education: Indian context specifically discussed as a driver of premium education and better outcomes; education is rated highly as a near-term growth area because it directly correlates with better life outcomes for families investing in children.
• Secondary markets for private companies: The expectation that private market secondary markets (trading existing private stake in companies) will become more prevalent as private company valuations and liquidity evolve; potential for more mature access to private equities.
• Wellness, longevity, and senior living: Aging demographics, higher life expectancy, and the demand for wellness will push longevity solutions and senior living services; high engagement in elder care communities with community benefits improving quality of life.
• Content and brand-led marketing: The need for brands to own attention in an attention-economy makes content creation and storytelling critical; video remains a dominant channel, with a focus on authentic storytelling and retention rather than only acquisition.
• Electric vehicle (EV) adoption: Adoption of EVs is expected to rise, with regional variances; infrastructure (charging networks) and cost considerations (battery costs) influence adoption. The panel expects majority of new cars to be EVs in some regions within 10–15 years; India and other large markets discussed with caveats around infrastructure and policy.
• Climate tech and energy: Energy is a core pillar; climate tech encompasses a broad spectrum from renewables to nuclear/fusion and energy storage; the potential for climate tech investments remains strong but with long return horizons. Data centers and energy arbitrage are touched as infrastructure plays with geopolitical and regulatory dynamics.
• Live events and offline experiences: In a world with increasing digital immersion, offline experiences may command premium pricing and safety; scarcity and brand experiences can drive demand and monetization.
• Beauty and luxury as durable growth: Beauty and luxury brands are highlighted as durable, high-margin sectors with strong consumer willingness to spend; often bolstered by branding and storytelling.
• Aging and senior care as a demographic tailwind: Senior living models with community benefits and high-quality care show promise, particularly in markets with aging populations and rising demand for eldercare solutions.
• Data and AI-enabled healthcare/simulation: A broad set of AI-enabled tools for healthcare, defense tech, and robotics are anticipated to unlock new business models tied to data, evaluation, and training.

"Beauty and luxury... live events... climate tech energy... education... aging/wellness... EV adoption" (summary of 02:20:00 topics)

03:40:00Major Topic: The Practicalities of Investment — David vs. Goliath, Product vs. Distribution

03:40:00 Subtopic: David vs. Goliath — Competition dynamics and investment philosophy

• David vs. Goliath: The panel discusses whether smaller, nimble founders can disrupt large incumbents; the consensus is nuanced—while big incumbents have advantages, smaller players can disrupt if they have a strong product, compelling distribution, or unique strategy. The group rates the likelihood of David success across various contexts.
• Distribution as an advantage: One strong thread is the idea that distribution (marketing reach and channel access) can be a powerful equalizer; content can be used to craft a narrative and build brand reliability.
• Product quality and retention: The counterpoint emphasizes that product quality and retention are essential; growth through attention must be anchored by a robust product that retains users, otherwise the growth is unsustainable.
• The role of capital intensity: High-CapEx sectors demand different capital strategies; efficient capital deployment and creative financing are critical, including leveraging external financing to enable long-run CapEx projects.
• The tension between innovation and regulation: Speculation markets (e.g., Kalshi, Polymarket) are recognized as part of the broader financial ecosystem, with regulatory risk as a key factor; the panel seems to entertain regulatory nuances as a gating factor for innovative financial products.

"David versus Goliath will be strong forever... the last mile problem will always be there" (Niko, 03:40:00)

04:40:00Major Topic: Indian Market and Global Dynamics

04:40:00 Subtopic: India as a Growth Engine for AI and Startups

• India’s potential: The panel discusses how India sits at an advantageous demographic position (young workforce) but notes structural challenges, including brain drain, capital access, and the need for more GPU-oriented AI talent development.
• Observations about local vs global advantages: While India offers a large talent pool and cost benefits, the panel argues that AI competition is increasingly global; local advantages may rely on data, language-specific models, and sector-specific applications rather than merely localized AI models.
• Recommendations for India: Build around strengths in manufacturing, healthcare, and traditional tech-enabled sectors; avoid purely “me-too” AI playbooks; focus on niche applications where India can outperform globally (e.g., data-rich domains in healthcare, language-specific AI tools). Retain talent through supportive policy, investment, and infrastructure.
• The importance of culture and risk tolerance: The panel debates the cultural propensity for risk-taking inside India; suggests nurturing a climate that supports ambitious founders and reduces risk aversion to capture global opportunities.

"Unlike other things, other kinds of technology in the past, you could create a local economy with that technology... AI has no obvious local advantage" (Nikunj, 04:40:00)

05:50:00Major Topic: Personal and Societal Impacts — Data, Privacy, and the Future of Work

05:50:00 Subtopic: Data, Privacy, and the Public Sphere

• Data and privacy in a world of pervasive sensing: The conversation acknowledges that the world is increasingly data-rich, with cameras, sensors, and ubiquitous devices collecting information that creates a real-time, public-sphere-like environment.
• The social contract of living in a data-rich world: There is concern about privacy, surveillance, and data traceability; people are living more of their lives in public in a context where devices and services capture more information than ever before.
• Youth and technology adoption: The group discusses how young users interact with AI and mobile devices, including concerns about social media, digital identity, and self-presentation in a data-driven world.
• Speculations about future tech and policy: They touch on the need for thoughtful policy around AI governance, data rights, and safety to avoid negative societal outcomes while still enabling innovation.

"Everything you ever say online, in theory, somebody can track... digital identity is being published without your choice" (Nikhil, 05:50:00)

06:50:00Major Topic: Real Estate, Urbanization, and Infrastructure

06:50:00 Subtopic: Real Estate Economics, Urban Trends, and Data Center Markets

• Real estate dynamics and CapEx: The discussion on real estate highlights how asset prices, construction productivity, and supply constraints drive long-run real estate value; data centers emerge as a critical infrastructure asset with high entry costs and long-term returns.
• Urbanization and density: The group generally supports ongoing urbanization and density, arguing that cities remain central to economic activity; but acknowledges the interplay of climate, cost of living, and migration patterns.
• Data centers as an investment theme: Data centers are considered a compelling but nuanced opportunity; debate centers on operator quality, energy sources, energy arbitrage, and the geopolitical aspects of energy and compute resources.
• Global energy dynamics: The panel discusses energy security, geopolitics around chips and energy, and the importance of local data sovereignty policies (e.g., India’s data localization for Indian data).

"Data centers as an investment... Nine for 2035; depends on operator quality and energy considerations" (Niko, 06:50:00)

07:40:00Major Topic: Personal Reflections on Future Society and Possible Scenarios

07:40:00 Subtopic: The Abundance Narrative and Human Purpose

• Abundance vs. scarcity: The conversation wrestles with whether a future enabled by AI and productivity gains will lead to abundance and shorter workdays, or whether inequality will persist due to governance and distributional issues.
• The role of religion, meaning, and culture: Several participants discuss the search for meaning, social cohesion, and how AI and technology could influence religion, community life, and social norms.
• Long-term human adaptation: There’s speculation about future human-AI co-evolution, potential AI-driven societal structures, and how humans might reorient toward art, philosophy, and more meaningful work as automation absorbs routine tasks.
• The potential of AI to augment or replace labor: The conversation covers potential shifts in labor markets, the creation of new job categories, and how people will adapt to AI-driven productivity.
• Governance and policy: The risks associated with rapid AI advancement call for policy foresight, ethical frameworks, and governance structures to ensure broad societal benefits while avoiding negative externalities.

"The genie is unlocked... once you get used to high efficiency, why would you stop? Gen Z wants to build and create; technology unlocks a new era" (Nikunj, 01:30:00)

08:30:00Major Topic: Risk, Regulation, and the Balance Between Innovation and Safeguards

08:30:00 Subtopic: Regulatory and Ethical Considerations in AI and Finance

• Regulation in speculative markets: The discussion around Kalshi and Polymarket points to regulatory risk in new financial instruments tied to prediction markets; the panel sees potential but emphasizes regulatory risk as a critical factor.
• AI safety and governance: There’s emphasis on the need for governance frameworks to keep pace with AI capabilities; policy makers face a hard problem in balancing innovation with safety and ethics.
• Privacy and data sovereignty: Re-emphasis on data localization (e.g., India) as a tool for data governance and economic strategy; this is linked to technology development and data-driven business models.
• The balance of open-source vs. proprietary AI: The panel hints at the trade-offs between broader access to AI capabilities and controlled, enterprise-grade solutions; concerns about safety and strategic initiatives shape the ecosystem.

"The hardest thing we face right now is even if you're a PhD in AI, you can't keep up with the pace" (Niko, 01:10:00)

09:40:00Major Topic: Synthesis — Key Takeaways and Forecasts

09:40:00 Subtopic: Recap of Insights and Practical Takeaways

• Core sectors with tailwinds: Education (outcome-focused), wellness/ longevity, senior living, content/brand storytelling, EV adoption, climate-energy tech, and AI-enabled enterprise software. These are highlighted as sectors with growth potential shaped by demographic shifts, technology, and consumer behavior.
• The role of founders: Across all discussions, founder quality, execution, and the ability to attract talent and partners are repeatedly called out as the strongest predictor of success—more so than sector guesswork.
• The enterprise AI opportunity: AI is a wide-ranging engine for efficiency; coders, clinicians, engineers, and other professionals will be augmented by AI, enabling new product categories and business models.
• India’s strategic opportunity: India presents a large, young talent pool with the potential to become a global innovation hub if supported by policy, investment, and capital formation; a call to focus on sectors where India can create global impact rather than chasing purely local opportunities.
• The importance of content and distribution: In an attention-driven economy, the ability to tell compelling stories and distribute them effectively is a powerful competitive edge; however, product quality and retention remain essential to sustainable growth.

"Content is going to be the biggest differentiator; attention is all you need" (Nikhil, 03:46:00)

10:56:0010:56:00 Subtopic: Final Reflections and Closing Thoughts

• The cohort ends with reflections on personal lives, ethical considerations, and the reality that the venture ecosystem thrives on intelligent risk-taking, strong mentorship, and the ability to adapt quickly to a shifting landscape.
• The host closes with gratitude and a nod to the audience, acknowledging that the discussion has been a long, deep dive into the VC playbook, with many actionable takeaways for founders and investors alike.

Resources Mentioned

• Anthropic: Foundational AI company focusing on enterprise AI and code-centric models.
• Cursor: AI tool for code generation and automation; discussed as a notable example of hot AI tooling.
• Glean: Enterprise search and knowledge tooling; Deedy’s prior company.
• OpenAI: Foundational AI company referenced in context (authoritative source of modern LLMs).
• Waymo, Tesla, and other autonomous/robotics discussions: References to autonomous systems and infrastructure needs.
• Alpha Evolve (DeepMind): Reference to a DeepMind project/paper demonstrating novel algorithmic breakthroughs.
• Three-Body Problem (Liu Cixin): Sci-fi reference to future society and the role of AI, science, and policy.
• Ozempic: Diabetes/weight-management drug cited as a case study in consumer health and lifestyle trends.
• FPV Ventures, Founder's Point of View: The small, seed-to-Series-A focus fund where Nikunj is located.
• Khosla Ventures and General Catalyst: VC firms mentioned in the context of career paths and collaboration networks.
• Neeraj and Venture Highway: India-focused venture ecosystem connections referenced for collaboration.
• Hints at India-based policies: Local data localization, talent development, and cross-border collaboration as pivotal themes.

Notable Verbatim Quotes (verbatim with timestamp and speaker)

• 00:00:00 — Nikhil Kamath: "How you guys see different sectors evolving... Where can me, as either an investor, or somebody looking to start my career as a young man or woman have enough tailwinds that it makes sense to spend the next decade in?"
• 00:10:03 — Deedy: "Glean started as an enterprise search business... evolved into generally an enterprise assistant and knowledge finding tool for bigger companies to figure out answers to questions..."
• 01:25:00 — Nikunj: "We mostly care about the founder and the team that we meet."
• 01:30:00 — Niko: "AI has actually made most industries that were uninvestable suddenly investable."
• 01:40:00 — Niko: "Cursor is doing AI code gen, of course. AI receptionist, voice applications..."
• 02:21:03 — Deedy: "Superpowers. Patent spotter, relentless editor, agency amplifier." (ChatGPT exercise)
• 03:40:00 — Niko: "David versus Goliath... the last mile problem will always be there."
• 04:40:00 — Nikunj: "Unlike other things, AI competition is global; data labeling, languages in India, and manufacturing vs. services dynamics will shape opportunities."
• 05:50:00 — Nikhil: "Everything you ever say online, in theory, somebody can track... digital identity is being published without your choice."
• 06:50:00 — Niko: "Data centers as an investment... Nine for 2035; depends on operator quality and energy considerations."
• 07:40:00 — Deedy: "The product and retention matters; content can catalyze growth, but you need a truly good product to sustain it."
• 08:30:00 — Debrief on AI governance: "The hardest thing we face right now is even if you're a PhD in AI, you can't keep up with the pace."

Topics

• Venture Capital Strategy
• AI Trends and Markets
• Sector Tailwinds and 10-Year Forecasts
• Founders and Team Dynamics
• India as a Global AI Hub
• Education, Wellness, and Aging Markets
• Content, Brand, and Marketing in the Age of AI
• Climate Tech and Energy Infrastructure
• Data Centers and Infrastructure
• Regulatory and Societal Impacts of AI
• Live Events, Digital Detox, and Social Media
• David vs. Goliath in Startup Competition
• Urbanization and Real Estate Dynamics
• Privacy, Digital Identity, and Safety in an AI World

Resources Mentioned

• Anthropic (AI company)
• Cursor (AI code-generation and automation tool)
• Glean (Enterprise search/knowledge tool)
• Opendoor (Real estate fintech)
• Meter (Networking hardware/pay-per-use model for spaces)
• FPV Ventures (Founder's Point of View)
• Khosla Ventures (VC)
• General Catalyst (VC)
• Venture Highway (India-focused investment community)
• Alpha Evolve (DeepMind) – reinforcement-learning/novel problem-solving paper
• Three-Body Problem (book by Liu Cixin) – sci-fi influence on future tech discourse
• Ozempic (weight management drug)
• Waymo/Tesla (autonomous mobility references)
• Kalshi/Polymarket (prediction markets)
• Tariq references to Lipstick and Beauty/Luxury sectors
• Risers in the Indian tech ecosystem (Freshworks example; manufacturing and non-AI sectors)
• DC and energy policy topics (data sovereignty and local data storage)