20VC: Leading Anthropic's First Ever Round | Will Open Source...
The Twenty Minute VC (20VC)Full Title
20VC: Leading Anthropic's First Ever Round | Will Open Source Threaten Anthropic's Business | Do Margins Matter in a World of AI | Why Triple, Triple, Double, Double is Not Good Enough Today | Why Series A is Hard Today with Matt Murphy @ Menlo
Summary
Matt Murphy of Menlo Ventures discusses leading the investment in Anthropic and other AI companies, highlighting the evolving VC landscape.
The conversation covers the challenges and opportunities in AI investment, the role of open source, the importance of margins, and the difficulty of Series A funding.
Key Points
- Menlo Ventures led the initial investment in Anthropic, a significant move given the company's early valuation and Menlo's fund size, driven by conviction in the AI market and Dario Amodei's vision.
- The firm demonstrated flexibility by investing in Anthropic despite it not fitting traditional fund mandates, prioritizing market entry over strict ownership percentages in an outlier market.
- Ownership relevance has shifted; in the current AI market, being a small percentage of a massive outlier company is more valuable than a large percentage of a moderately successful one.
- Special Purpose Vehicles (SPVs) are increasingly used to allow firms to make larger, more concentrated investments than their main fund might permit, particularly in fast-moving AI opportunities.
- Taking liquidity off the table is less prioritized for winning investments, with a focus on letting outliers compound fund returns, although secondary sales can occur for LP liquidity or fund maturity reasons.
- The most nerve-wracking time for Murphy was the first large SPV for Anthropic, which involved significant fundraising and coordination, but ultimately proved exhilarating.
- The rise of open-source AI models poses a challenge but doesn't entirely displace the need for sophisticated foundation models like Anthropic's, which offer superior performance for critical workflows, leading to a hybrid model approach.
- Companies like Lovable are seeing explosive growth, demonstrating the potential for massive valuations even with initial lower margins, with a clear path to improvement through optimization and specialized models.
- Margins matter, but the focus is on the credible path to improvement, with AI infrastructure and application companies needing to optimize costs and leverage unique data or models to achieve higher gross margins.
- The venture funding landscape has shifted, making Series A rounds difficult due to high valuations and compressed timelines between seed and Series A, leading firms to adopt a barbell strategy of investing very early (seed) or in established leaders (growth).
- Smaller boutique seed funds may struggle in the current environment, as larger, more established firms (like Menlo, Founders Fund, Sequoia) are effectively competing at the seed stage, and the trend is towards larger, more collaborative rounds.
- The venture market has moved away from rigid "swim lanes," with many firms becoming more full-stack and engaging in collaborative syndication for larger rounds.
- Menlo Ventures raised a $3 billion fund to maintain its "small and mighty" firm structure, focusing on agility and high-quality partnerships rather than expanding into a large, multi-team structure.
- While San Francisco's prominence in AI is returning, there's growing interest in global talent, particularly in Europe, which has a reputation for fostering gritty entrepreneurs.
- Losing deals is often due to being late to build relationships, with personal connections and trust being critical in venture capital, making shotgun approaches less effective.
- The conventional "triple, triple, double, double" growth trajectory is no longer sufficient to excite investors, who now expect significantly more explosive growth from outlier companies.
- The trend of companies developing their own chips (e.g., Google, Amazon, Meta) is driven by the need for cost optimization and performance improvements at scale, although it's a difficult business to enter.
- The routing business, exemplified by OpenRouter, is seen as a significant opportunity due to its ability to optimize model selection for applications and its strong developer traction.
- The infrastructure and tooling stack above foundation models (e.g., observability, routing, AI frameworks) is an area of growing excitement and investment as AI ecosystems mature.
- Future excitement is centered on medical breakthroughs driven by AI, transformative changes in the healthcare system, and the overall potential of AI to reshape society over the next decade.
Conclusion
The venture capital landscape is rapidly evolving, with a shift towards investing in outlier companies, a greater acceptance of SPVs, and a focus on flexible strategies to capture market opportunities.
AI is driving unprecedented growth and valuation potential, requiring VCs to adapt their investment theses and embrace a more dynamic approach to identifying and supporting category leaders.
The future of venture capital involves deeper sector expertise, strategic partnerships, and a focus on transformative technologies that promise significant societal impact.
Discussion Topics
- How should VCs balance the pursuit of massive outliers with maintaining traditional investment theses in a rapidly evolving AI market?
- What are the key indicators that distinguish truly defensible application companies from those whose value could be replicated by foundation models?
- As AI continues to advance, what are the most significant ethical considerations and regulatory challenges investors and founders need to address?
Key Terms
- Foundation Models
- Large-scale AI models trained on vast amounts of data that can be adapted for various downstream tasks.
- OpenAI
- An artificial intelligence research laboratory that developed models like ChatGPT and GPT-3.
- ChatGPT
- A conversational AI model developed by OpenAI, known for its ability to generate human-like text.
- Compute Multipliers
- The efficiency or cost-effectiveness of computational resources used in AI model training and inference.
- SPV (Special Purpose Vehicle)
- A subsidiary created for a specific limited purpose, often used in venture capital to pool investor capital for a particular deal.
- LP (Limited Partner)
- An investor who contributes capital to a fund but has no role in its management.
- ARR (Annual Recurring Revenue)
- The predictable revenue a company expects to receive from its customers over a year.
- POC (Proof of Concept)
- A small project or test to demonstrate the feasibility of an idea or product.
- NeoLabs
- A term likely referring to new or emerging AI labs or companies focused on foundational AI research and development.
- Full Stack
- A company that operates across all layers of a technology stack, from hardware to software applications.
- GPUs (Graphics Processing Units)
- Specialized processors that are highly efficient for parallel processing tasks, crucial for AI computations.
- TPUs (Tensor Processing Units)
- Custom ASICs developed by Google specifically for machine learning workloads.
- Inference
- The process of using a trained machine learning model to make predictions on new, unseen data.
- COGS (Cost of Goods Sold)
- The direct costs attributable to the production or purchase of the goods sold by a company.
- SaaS (Software as a Service)
- A software licensing and delivery model where software is licensed on a subscription basis and is centrally hosted.
- YC (Y Combinator)
- A startup accelerator program that provides seed funding and mentorship to early-stage companies.
- Carry
- The share of profits that a general partner receives from a venture capital fund's successful investments.
- Benchmarking
- The process of comparing the performance of a company or investment against industry standards or competitors.
Timeline
Matt Murphy details the introduction to Anthropic through Anjan, highlighting the initial conviction and the challenges of investing in an early-stage company with a high valuation request relative to Menlo's fund size.
Murphy explains the internal debate regarding Anthropic's valuation and the firm's decision to invest despite the unconventional entry point, framing it as a necessary pivot towards AI.
Murphy discusses the size of the initial check into Anthropic and the potential return scenarios that convinced the partnership.
Murphy describes the partnership's discussion around investment thesis and the justification for entering the foundational AI market despite valuation concerns.
Murphy elaborates on how ownership percentages are less relevant than being in outlier companies in the current venture market.
Murphy addresses whether valuation still matters, stating that while there are limits, his preference is to be in the most amazing companies.
Murphy discusses the normalization of dilution in the AI sector due to increased funding rounds and the need for capital to play offense.
Murphy details the decision-making process for the larger, second investment round in Anthropic, emphasizing building relationships and providing value beyond capital.
Murphy explains the factors that validated the second investment, including Amazon and Google's involvement and Anthropic's revenue growth.
Murphy discusses the increasing use of SPVs and how they are managed within Menlo's fund strategy to play offense and accommodate larger investments.
Murphy addresses the timing of taking money off the table, emphasizing the focus on letting winners compound returns.
Murphy reflects on the most nerve-wracking period as an Anthropic shareholder, relating it to the firm's first SPV.
Murphy touches upon the dynamics of VC firms lending their brand and the transition of power when companies legitimize VCs, referencing the controversy around SPVs.
Murphy discusses the investment in Lovable, detailing its rapid growth and the underwriting thesis based on market potential and founder vision.
Murphy addresses the importance of margins in AI, acknowledging that many great companies currently have lower margins but a path to improvement.
Murphy explains how Lovable's margin structure will improve with open-source utilization, clarifying that it doesn't contradict the Anthropic investment due to market size and differing user bases.
Murphy explains why he believes foundation models like Anthropic will remain essential despite the rise of open-source options, citing their superior performance.
Murphy discusses the downward cost curve for AI tokens and the combination of model families and open-source solutions that will emerge.
Murphy discusses the trend of companies developing their own chips as an optimization strategy for cost and performance, noting the difficulty of the chip business.
Murphy contrasts full-stack approaches with specialized solutions, highlighting the value of routing businesses like OpenRouter due to developer trust and intelligence.
Murphy expresses high hopes for OpenRouter's potential, citing its rapid trajectory and profitability.
Murphy discusses the investment in Lagora, highlighting its strong execution and the founder's vision beyond just legal applications.
Murphy explains how application companies can defend their market share by building distinct workflows and value on top of foundation models, distinguishing Lagora's complex B2B approach.
Murphy agrees that Lagora will likely expand beyond legal to other professional services like compliance and tax to justify future valuations.
Murphy states that Series A is currently the hardest stage to invest in due to high valuations and uncertain competitive landscapes, leading to a barbell investment strategy.
Murphy explains Menlo's shift towards a barbell strategy: investing in outliers above $10M ARR or very early at the seed stage, acknowledging the compression of the seed-to-Series A window.
Murphy argues that small, boutique seed funds may underperform in the current vintage due to strong competition from larger, established firms with effective seed products.
Murphy discusses the dissolution of traditional "swim lanes" in venture capital, with firms becoming more full-stack and collaborative rounds becoming more common.
Murphy explains that Menlo raised a $3 billion fund to maintain its agility and "small and mighty" firm culture, prioritizing alignment and collaboration over sheer size.
Murphy believes investors are best off focusing on a specific "swim lane" (stage or sector) to build deep expertise, although fluidity exists.
Murphy reflects on the difficulty of investing in sectors like semiconductors without deep domain expertise, contrasting it with more opportunistic bets on exceptional founders.
Murphy notes the increasing interest in global AI talent, including European companies, and the inherent grit of European entrepreneurs.
Murphy identifies losing deals primarily due to a lack of proactive relationship building and being late to the party.
Murphy highlights the shift from prioritizing high ownership percentages to ensuring participation in outlier companies, even with smaller initial stakes.
Murphy believes LPs understand the changing investment landscape where smaller initial stakes in massive outliers can yield significant returns.
Murphy explains that the "triple, triple, double, double" growth model is no longer sufficient in venture, with companies achieving much faster growth and higher valuations.
Murphy reflects on past investment misses, now seeing them as learning experiences that reinforced the need for continuous pursuit of large opportunities, particularly in AI.
Murphy selects a seed fund (Sousa), a Series A fund (Benchmark), and a growth fund (larger firms like Lightspeed) if he could only invest in one of each.
Murphy believes the growth market has dramatically changed, with boutique growth funds now needing to operate at a much larger scale.
Murphy identifies the NeoLabs sector as currently overheated due to the sheer number of companies and large rounds raised, predicting consolidation.
Murphy expresses excitement about the developer stack and tooling above foundation models, an area that was previously underinvested.
Murphy's greatest excitement for the next 10 years lies in AI-driven medical breakthroughs and the transformation of the healthcare system, as well as Menlo's ability to capitalize on the AI wave.
Murphy believes the current era of AI represents the biggest technological shift in decades, with unprecedented potential for societal transformation.
Murphy discusses the importance of a strong relationship with financial partners like JP Morgan for navigating startup growth.
Murphy highlights Corgi Insurance as a solution for the slow and confusing insurance process faced by tech companies.
Murphy mentions Flex as a platform that simplifies financial management for business owners.
Murphy emphasizes the importance of sustained momentum and a challenger mentality for firms like Menlo to continue to thrive after significant wins.
Murphy suggests that being a "richer investor" allows for a focus on upside optimization rather than downside mitigation.
Murphy talks about a past miss on a company called Winclad, which taught him that one loss doesn't define a career.
Murphy expresses excitement about companies in drug discovery and the transformation of the healthcare system via AI.
Murphy feels privileged to be investing in AI during this transformative period.
Episode Details
- Podcast
- The Twenty Minute VC (20VC)
- Episode
- 20VC: Leading Anthropic's First Ever Round | Will Open Source Threaten Anthropic's Business | Do Margins Matter in a World of AI | Why Triple, Triple, Double, Double is Not Good Enough Today | Why Series A is Hard Today with Matt Murphy @ Menlo
- Official Link
- https://www.thetwentyminutevc.com/
- Published
- July 27, 2026