Neil Rimer, co-founder of Index Ventures, one of Europe’s most prominent venture capital firms, has argued publicly that the enormous sums of capital that flooded into artificial intelligence infrastructure over the past two years are beginning to rotate back out toward a broader class of technology startups. Speaking in an interview reported by TechCrunch, Rimer suggested that the AI investment cycle is entering a new and more distributed phase, one that could benefit a wider range of founders and sectors beyond the handful of foundation model developers that have dominated venture headlines since 2023.
Rimer’s remarks carry particular weight given Index Ventures’ record across both European and American markets, where the firm has backed companies including Adyen, Robinhood, and Figma. His comments reflect a broader debate now playing out across the venture capital community about where AI-derived value will ultimately accrue and whether the concentration of capital at the infrastructure layer has come at the expense of application-layer innovation.
Infrastructure Spending Peaks as Application Bets Multiply
The scale of capital that has entered AI infrastructure over the past 24 months has been staggering by any historical measure. Estimates from multiple market research sources suggest that global private investment in AI-related companies exceeded 100 billion dollars in 2024 alone, with a disproportionate share flowing to compute-heavy foundation model developers such as OpenAI, Anthropic, and xAI. Rimer’s argument, as reported by TechCrunch, is that this upstream concentration is beginning to stabilise, freeing risk appetite and dry powder to move downstream into software companies building on top of existing AI capabilities.
That structural shift, if it materialises at the pace Rimer anticipates, would mark a meaningful evolution in how venture capital firms allocate across the AI value chain. Application-layer companies typically require far less capital per deal than foundation model labs, which have raised individual rounds exceeding 6 billion dollars in some cases. A rebalancing toward application and vertical software bets would therefore spread capital across a larger number of companies, potentially compressing average round sizes while expanding the total number of funded startups. For limited partners watching capital deployment timelines closely, that dynamic carries both opportunity and risk.
What a Rotation Means for Founders and Markets
Rimer’s thesis has immediate implications for founders who have spent the past two years struggling to attract attention or capital in a market where investors were largely fixated on a small cluster of headline-grabbing infrastructure deals. If his read on the cycle is correct, sector-agnostic AI applications in areas such as healthcare, legal technology, financial services automation, and enterprise workflow could see markedly improved funding conditions through the second half of 2026 and into 2027. Index Ventures itself has signalled interest in precisely these categories, though Rimer stopped short of making specific portfolio announcements during the interview.
The rotation thesis is not without its sceptics. Some investors argue that application-layer companies remain exposed to significant platform risk, given that foundation model providers are themselves expanding into product and vertical software territory. Others point to the broader macroeconomic environment, where elevated interest rates and tightening credit conditions through much of 2025 constrained venture fundraising for general partners, leaving some firms with less dry powder than publicly projected. The debate echoes concerns raised in analysis surrounding the AI market collapse scenario, in which rapid capital reallocation could expose fragile valuations rather than validate them.
For the wider technology investment community, Rimer’s comments arrive at a moment when public market sentiment toward AI-exposed equities remains volatile and private market valuation benchmarks are under increasing scrutiny. Whether the capital he describes as coming back out translates into a genuine funding renaissance for early-stage founders, or merely a modest reallocation at the margins, will depend in large part on whether enterprise AI adoption accelerates enough to justify application-layer revenue multiples. Concerns about how quickly businesses can operationalise these tools have also come into sharper focus, as explored in reporting on AI governance challenges that many companies are only beginning to confront. As Rimer himself acknowledged, the cycle’s next chapter remains early, and the distribution of returns is far from settled.