Menlo Partner: The AI Bet That Paid Off 100x | Amy Wu Martin
Menlo Partner: The AI Bet That Paid Off 100x | Amy Wu Martin
YouTube52 min 38 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights
  • Treat AI venture exposure cautiously: funding rounds are moving quickly, valuations are heated, and rushed diligence raises downside risk; prioritize companies with durable growth and defensible advantages.
  • Watch AI infrastructure and inference—including CoreWeave, Nebius, Fireworks AI, and Modal—for public-market opportunities, but demand valuation and profitability evidence before investing.
  • Monitor AI-native application companies for potential IPOs in roughly 12–24 months; rapid growth stories such as Higgsfield are promising, but the discussion provides no valuation or profitability data to justify buying specific names.
  • Track Meta, Amazon, Walmart, and Shopify for evidence that AI shopping features improve conversion and engagement without weakening customer relationships; no stock-specific targets were provided.
Detailed Analysis

AI Venture Market

  • Menlo said more than 90% of its investments have been in AI, reflecting a deliberate, highly concentrated strategy.
  • Amy Wu Martin described the current market as extremely heated: some large rounds are completed in about a week, with limited time for diligence, and founders may take substantial secondary proceeds.
  • She compared the market’s behavior with 2021, while cautioning that no one can predict when an adjustment might occur. If growth slows, companies may find it harder to raise large, frequent rounds.
  • Menlo’s strategy spans early-stage companies and later-stage businesses that appear to be emerging winners, including in inference and applications.

Takeaways

  • AI is attracting significant capital, but the discussion highlights the risk of paying high prices amid rushed diligence and optimistic growth assumptions.
  • For investors evaluating AI companies, the discussion suggests looking beyond funding momentum to evidence of durable growth, competitive advantage, and the ability to withstand a slower funding market.
  • The guest expects some AI-native application companies to begin reaching public markets in roughly 12–24 months, but noted that their growth and margin profiles may differ from traditional SaaS companies.

Anthropic

  • Menlo described Anthropic as a major investment in a leading frontier AI model company. Menlo’s first investment was in its Series C, a round led by Spark at a $5 billion valuation in 2024.
  • The firm said it had an existing relationship with the company and had observed its team over time. A key point in the decision was Anthropic’s focus on AI safety.
  • Menlo ultimately made an investment of more than $500 million. Wu Martin said the firm viewed the bet as so consequential that being wrong could threaten the firm.
  • She said the investment performed far better than the firm had imagined, but did not specify a return multiple.

Takeaways

  • The discussion emphasizes the potential value—and substantial risk—of concentrated investments in leading AI-model companies.
  • Menlo’s account points to prior familiarity with the founders and conviction in the market as important parts of its decision; it does not establish that the same investment terms or outcomes are available to other investors.
  • No future price target or specific recommendation for Anthropic was given.

AI Infrastructure and Inference: CoreWeave, Nebius, Fireworks AI, and Modal

  • Wu Martin said the first AI-related companies to reach public markets have largely been in neocloud infrastructure, naming CoreWeave and Nebius and mentioning SpaceX in the context of companies needing access to capital.
  • Menlo has invested in Fireworks AI and Modal, which she cited as companies in the inference space.
  • She expects another wave of inference and neocloud companies to pursue public listings, though she did not identify specific listing dates or price targets.

Takeaways

  • The transcript points to AI computing infrastructure and inference as areas where demand for capital may create public-market opportunities.
  • Investors should distinguish infrastructure providers from AI application companies: the transcript offers no company-level valuation analysis or comparison of their business risks.
  • The guest’s broader warning about heated markets applies here: high demand for financing does not by itself demonstrate that a company’s valuation is sustainable.

AI Applications: Higgsfield, Lovable, Legora, and Suno

  • Higgsfield was presented as a company whose founder, Alex, impressed Menlo with his ability to hire strong teams and anticipate market shifts. The company initially pursued video APIs, then shifted toward its application layer and the tools built around models.
  • Wu Martin said Higgsfield went from launching its product to a $1 billion revenue run rate in 18 months. She used its rapid growth as an example of why some companies are raising new rounds quickly.
  • Lovable, Legora (rendered as “Lagora” in the transcript), and Suno were named among AI companies Menlo had backed in growth rounds after they appeared to be emerging winners in their markets.

Takeaways

  • Higgsfield’s story illustrates the potential importance of founder judgment and willingness to change strategy as the market evolves.
  • Rapid revenue growth can help a company raise capital and compete for talent, but the transcript also warns that very fast fundraising can coincide with limited diligence and heated valuations.
  • The discussion did not provide valuation, profitability, or other financial details for these companies, so it offers no basis for comparing their investment merits.

Consumer AI Agents and Commerce: Manus, Instinct, Muse, and Alta

  • Wu Martin described personal AI agents as an emerging consumer opportunity, driven by improvements in both execution capability and user experience.
  • Manus, Instinct, and Muse were discussed as examples of consumer-facing agent products. She said horizontal platforms may have an advantage over a collection of separate apps for individual tasks.
  • Alta was described as an AI styling and shopping recommendation app. The guest said about $1 billion of items are uploaded each month from users’ closets for recommendations.
  • Consumer AI adoption appears to be growing, based on the survey Wu Martin discussed: AI spending was said to have roughly tripled, about half of surveyed AI users were paying for a product, and about half said they had made money using AI.
  • She also highlighted adoption barriers: some users remain reluctant to use AI, and tasks involving medical appointments, payments, personal data, and purchases require greater reliability and trust.

Takeaways

  • Consumer AI agents and AI-assisted shopping are potential growth areas, but the transcript makes clear that trust, privacy, security, and reliable execution are essential hurdles.
  • For commerce-focused businesses, a useful area to monitor is whether users complete purchases through agents—not simply whether they download or try them.
  • The adoption statistics cited were from a survey discussed on the podcast; they should not be treated as a guarantee of market-wide adoption or company revenue.

Meta, Amazon, Walmart, and Shopify

  • Meta was described as well positioned to develop consumer AI products because of its large user base, personalization data, and advertising business. Wu Martin characterized its Muse launch as a significant new consumer interface.
  • Amazon had not integrated with Muse at the time of the discussion. Wu Martin suggested that Amazon has incentives to retain control of the customer transaction and shopping data, and speculated that it could eventually offer its own deeply integrated agent.
  • Walmart and Shopify were mentioned as platforms moving toward agent-based commerce. Wu Martin said Walmart had reported lower conversion on an initial AI checkout experience than on its website, while the gap was narrowing.
  • She noted that agent-driven shopping still faces friction and that adoption may vary across consumers and retail categories.

Takeaways

  • The discussion suggests that consumer AI could affect how retailers attract customers, collect transaction data, and convert purchases.
  • For these companies, a key signal to watch is whether AI features improve shopping conversion and customer engagement without weakening control of the customer relationship.
  • The transcript did not give stock-specific forecasts, valuations, or recommendations for Meta, Amazon, Walmart, or Shopify.

AI Gaming and Video: Roblox, ByteDance, and MiniMax

  • Wu Martin said recent model improvements could make it more feasible to build complex games, beyond visually impressive demos or simpler games. She pointed to possible opportunities for new platforms and for Roblox to adopt AI tools.
  • In video, she cited ByteDance’s Seedance and MiniMax H3 as examples of model advances. She said lower cost and latency could enable more real-time, personalized video experiences.
  • She also described AI-generated short-form entertainment and AI-assisted live commerce as emerging areas.

Takeaways

  • AI tools for game creation and video production could enable new forms of entertainment and lower the cost of content creation.
  • The investment opportunity remains speculative in the transcript: it gives examples of technical progress but no company-specific revenue forecasts or valuations.
  • Investors should watch for evidence that new tools produce engaging content and sustainable business models, rather than relying on demonstrations alone.
Ask about this postAnswers are grounded in this post's content.
Video Description
Amy Wu Martin, partner at Menlo Ventures, co-leads consumer investing at the 50-year-old firm behind Anthropic, Uber and Roku, which recently raised $3 billion to go all in on AI. Before Menlo, she was a partner at Lightspeed Venture Partners and CFO of Discovery’s Global Digital and Sports unit, and she studied at Harvard University. Subscribe to stay ahead of technology trends. ------------------------------------ Timestamps (0:00) Meet Amy Wu Martin (2:22) Menlo at 50 (4:06) Inside Menlo Ventures (6:50) Supporting struggling companies (8:30) Why personality matters in VC (10:13) Who holds the power: founders or investors? (10:45) Menlo’s all-in AI bet (13:19) How Menlo underwrites deals (15:58) Where did due diligence go? (18:25) When does the market crack? (19:54) Raising every 6 months (21:19) Backing Higgsfield (22:49) Backing Anthropic (25:57) What AI IPOs will change (28:29) Is consumer back? (33:49) How many AI subscriptions per household? (37:35) What makes personal agents special (42:10) Big ideas in agentic commerce (44:34) Why Amazon isn’t integrating with Muse (45:36) Was Muse essential for Meta? (46:51) The next big breakthroughs (51:52) What Amy would build ------------------------------------ Where to connect with us Follow Ollie on X: https://x.com/ollieforsyth Follow Amy on X: https://x.com/amywumartin Visit Menlo Ventures: https://menlovc.com ------------------------------------ Our Partners VCX: The public ticker for private tech. Head over to http://getvcx.com to learn more. Prelint: Keeps every decision reviewed, recorded, and traceable. Head over to https://prelint.com to learn more.
About NEW ECONOMIES
NEW ECONOMIES

NEW ECONOMIES

By @neweconomiespod

Uncovering the origin stories of today’s most iconic founders and companies.