Anthropic investor remains bullish despite $8 billion in annual losses
A major venture capital firm defends Anthropic's high valuation and massive operating costs, arguing that the market is shifting toward infrastructure optimization.
The short version
- Anthropic reportedly racked up $8 billion in operating losses last year, according to leaked financials obtained by Reuters.
- Menlo Ventures, an early investor, could see its $1 billion stake grow to $25 billion at a $2 trillion valuation.
- Matt Murphy of Menlo Ventures argues that the market is moving from experimentation to optimization and cost-cutting.
- The firm currently allocates 60% to 70% of its AI investments toward the infrastructure layer, including compute and routing.
- While some institutional investors fear competition from Chinese models, others believe American labs have a major cost advantage.
Why it matters
This highlights a growing divide in the AI industry between those who fear 'froth' in valuations and those who believe massive infrastructure scale creates an insurmountable moat. For the reader, it signals that the next era of AI success may depend more on efficient routing and cost management than on simply building the largest models.
Defending the massive losses
Menlo Ventures, a prominent venture capital firm, is standing firm on its investment in Anthropic despite recent reports of heavy spending. Leaked financials obtained by Reuters revealed that the AI company recorded an eye-watering $8 billion in operating losses over the last year. These figures have sparked a debate among investors regarding whether the current valuations for leading AI labs are sustainable.
Matt Murphy, a partner at Menlo Ventures, dismissed these concerns by urging the public to wait for the company's official S-1 filing before making a final judgment. He remains extremely bullish on the company's trajectory and expressed a desire to invest even more capital if he could. The firm has participated in every funding round for Anthropic since 2023, positioning itself as a primary stakeholder in the company's future.
The stakes for Menlo Ventures are significant if the company proceeds toward an initial public offering. If the listing succeeds at a $2 trillion valuation, the firm's estimated $1 billion investment could transform into approximately $25 billion. This potential return underscores the high-risk, high-reward nature of the current AI venture landscape where early movers are betting on total market dominance.
The shift to optimization
Murphy argues that the AI industry is moving past the initial phase of simple experimentation and into a period of intense optimization. He believes that companies are now increasingly focused on how to cut costs and move work to the most efficient systems available. This involves refining post-training models and developing smarter routing to manage how tasks are distributed across different AI systems.
This shift in focus is reflected in how Menlo Ventures is currently allocating its capital. The firm has directed roughly 60% to 70% of its AI investments toward the infrastructure layer. This includes the foundational technologies required to run models, such as compute power, inference, and the systems that route data to the correct processing units.
By focusing on the plumbing of the AI world, the firm aims to capture value from the underlying systems that make large-scale models functional. Murphy suggested that while some startups in robotics or neo-labs are interesting, many are raising hundreds of millions of dollars without showing real traction. He views the infrastructure layer as the more stable and necessary component for long-term growth.
Competition and market moat
A major point of contention in the industry is whether Chinese open-weight models will eventually commoditize the work of major American labs. Murphy pushed back against this idea, suggesting that the sheer scale of American infrastructure provides a distinct advantage. He believes that the massive investment in chips and data centers creates a barrier that smaller or foreign competitors may struggle to overcome.
This perspective contrasts with some institutional voices who have expressed growing skepticism about the dominance of American AI. For example, the head of Washington University’s endowment recently warned that competition from Chinese models could put AI labs in major trouble. This investor noted that they were already skeptical of the original OpenAI deals and have become more concerned over time regarding the sustainability of these valuations.
However, other prominent figures in the space disagree with the fear of commoditization. Early OpenAI investor Vinod Khosla stated that American labs like OpenAI and Anthropic possess a huge cost advantage due to their access to high-end chips and massive data centers. This debate highlights the uncertainty surrounding whether the current leaders can maintain their lead as the technology becomes more accessible globally.
The 'froth' in the ecosystem
While the leaders of the AI space are being defended by some, there is a growing sense of caution regarding the broader startup ecosystem. Murphy pointed out that there is significant 'froth' in the market, specifically noting that some companies are raising billion-dollar rounds before demonstrating any meaningful progress. This suggests a disconnect between the massive amounts of capital flowing into the sector and the actual utility being delivered by newer startups.
The sentiment among the broader investment community is often less enthusiastic than that of the primary backers. During a recent town hall, many investors expressed that they would not be willing to buy Anthropic stock at a $3 trillion valuation. This tepid response indicates that while the technology is impressive, there is a limit to how much the market is willing to pay for future promises.
The contrast between the
Sources and links
- Postlater chimed x.com
- PostMy tweet x.com
- ReportingBusiness Insider: the article we sourced
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.
Listen to the AI Wire headlines
Listen to the AI WireListen to our daily AI shows every morning: on your commute, in the shower, wherever you are
Listen to the daily showsFirst reported by Business Insider · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
