Oct 19, 2026, 10:00 AM ETInterview, Fireside Chat

Who Wins in the Next Decade of VC | Margin and Price Matter Less in a World of AI | Sameer Dholakia

Bessemer Venture Partners recently closed a $5.75 billion fund, marking a significant increase in capital deployment capacity.
The fund structure includes a $1.75 billion seed and early-stage vehicle and a $4 billion growth-stage vehicle.
Samir (General Partner at Bessemer) describes the current AI market as a "trillion-dollar super cycle," asserting that intelligence will eventually exceed oil as the world's largest market.
Bessemer's investment thesis relies on the belief that capital is a primary moat in the AI sector due to the massive compute and infrastructure costs required.
Samir predicts a structural shift where companies will remain private longer, with a potential unlock of IPO activity expected in late 2026 or 2028 driven by public market demand for pure-play AI exposure.

Anthropic: Bessemer led a Series D round at a $15 billion pre-money valuation; this was internally voted the lowest-voted deal in firm history at the time.
Valuation Adjustment: Samir admitted to underestimating the scale of the AI market, noting they initially capped their imagination at a $100 billion exit for Anthropic before OpenAI's trajectory shifted that view.
Agora: Bessemer invested $1.8 billion in the legal AI company Agora, which had approximately $20 million ARR at the time of investment.
Fireworks: Bessemer invested in Fireworks (an AI inference platform) alongside Anthropic, betting on both closed and open-source model ecosystems.
FieldGuide: An early-stage investment in AI for auditors, which Samir describes as a "generational business" with extraordinary post-launch inflection.
Waymo: Bessemer invested nearly $200 million in Waymo, viewing the company's potential to expand beyond ride-hailing into personal autonomy, trucking, and logistics as a potential trillion-dollar opportunity.

Capital Intensity: The primary differentiator in the current AI wave is the sheer capital required for compute; Samir argues that only firms with massive war chests can compete, creating a self-fulfilling prophecy of market consolidation.
Closed vs. Open Models: Samir rejects the binary "closed versus open" debate, arguing both will generate massive value; he notes that while 99% of token traffic may flow through open models, the majority of dollars could still flow through closed models.
Sovereign AI: Bessemer is investing in sovereign AI models (e.g., Sarvam in India, Mistral) to address geopolitical concerns and specific language/data needs in regulated markets.
Enterprise Preference: Despite the rise of open source, Samir believes enterprises will increasingly seek to own their intelligence rather than rent it, driving demand for private inference infrastructure.
Open Source Monetization: He argues that open source businesses can succeed as providers of talent and support services (similar to Red Hat), noting that most enterprises prefer a single provider for open source solutions rather than managing them internally.

Check Sizes: Seed checks in the AI era are significantly larger than historical norms, with Samir noting that small seed funds are unlikely to succeed in finding winners in this capital-intensive sector.
Valuation Multiples: Bessemer targets 10x to 15x Next-12-Months (NTM) revenue multiples for hyper-growth companies, arguing that standard historical metrics (like 100x current ARR) are irrelevant in this context.
Portfolio Concentration: The firm is comfortable allocating 5% to 10% of a single fund to one company if conviction is high, citing a 20% allocation to Anthropic in a previous growth fund.
Liquidity Strategy: Samir advocates for taking profits along the journey (secondary sales) to provide DPI for Limited Partners, viewing this as a rational move rather than a lack of faith in the company.
T2D3 vs. Hypergrowth: While Bessemer still invests in 2-3x growth companies, the firm is prioritizing companies growing at 10x or faster due to the scarcity of capital and the compounding nature of the current cycle.

"Teammates vs. Colleagues": Samir emphasizes a culture where employees view each other as teammates who "beat colleagues and co-workers every day and twice on Sunday," built through shared non-work experiences and a focus on collective over individual success.
Values: Bessemer's culture is defined by "Happy, Hungry, Humble, Honest," with specific definitions for each term to ensure alignment (e.g., "Happy" means optimism and a "glass half-full" mindset).
Leadership Roles: Samir identifies Byron Deeter as Bessemer's best "closer" and a proven "picker" with extensive IPO experience, and names David Cowan as a top "picker" for identifying unconventional wins like Rocket Lab.
Investment Committee Dynamics: The firm prioritizes intellectual honesty in decision-making, where direct disagreement is encouraged over polite ambiguity to ensure rigorous capital allocation.

Robotics & Embodied AI: Samir predicts the robotics sector is in a phase analogous to generative AI in 2021, with massive potential in autonomy for construction, mining, agriculture, and personal transportation.
Therapeutics: He views the application of AI to drug discovery and "undruggable" targets as the highest-impact potential area, anticipating a decade of discoveries that would have previously taken 50 years.
Market Consolidation: He warns that the venture capital industry is professionalizing and consolidating; only giant global platforms or extreme specialists will survive, as mid-tier firms face a "hard world to make money."
IPO Window: Samir expects the public markets to open for AI companies in late 2026/2028 as demand for pure-play AI exposure from public investors exceeds supply.
Risk Assessment: He believes there is no bear case for frontier labs solving unsolvable problems, stating that human demand for intelligence is virtually infinite and not yet saturated.