AI’s survival of the fittest

AI’s Shifting Sands: Navigating the New Investment Landscape at Web Summit Vancouver 2026

At Web Summit Vancouver 2026, a panel of leading investors, including Managing Partner Andy McLoughlin of Uncork Capital, Global Head of Venture Capital Emily Fontaine from IBM, and Managing Director George Mathew of Insight Partners, discussed the evolving AI funding landscape. Mr. McLoughlin focuses on seed-stage, seeking non-obvious ideas and founders with unique building and storytelling capabilities. Ms. Fontaine, with IBM’s $500 million AI fund, targets strategic co-investments for growth in AI and Quantum. Mr. Mathew, representing Insight Partners’ multi-stage approach, prioritizes product-market fit, strong teams, and early commercialization from Series A.

To stand out, Ms. Fontaine advises a five-point strategy: strategic fit, differentiated technology, competitive landscape analysis, strong founders with a robust cap table, and sound financials. Mr. Mathew emphasizes identifying the precise moment founders have established their team, shipped a product, and show a clear path to commercialization. A significant challenge is the increasing reach of frontier models, demanding long-term durability from startups. Mr. McLoughlin highlights the shift in investor questioning from “Why can’t Google build this?” to “Why can’t frontier model providers do this?”

Success requires building truly unique solutions or leveraging proprietary data sets. Agility is paramount; Ms. Fontaine stresses founders’ willingness to pivot with market shifts. Mr. McLoughlin cited Mutiny, an Insight co-investment, whose CEO bravely transformed their successful marketing tech business into an agent-first solution, showcasing extreme adaptability. The panel agreed that “spray and pray” investing is obsolete, replaced by a focus on individual company opportunities and founder vision, rather than broad categorical bets.

Emerging categories like power, nuclear fusion, and power consumption software, previously niche, are now significant investment areas. Ms. Fontaine and Mr. Mathew also identified material science, particularly ceramics for storage and memory, as a re-emerging market. Physical AI, encompassing robotics and manufacturing, presents another exciting frontier. Ms. Fontaine sees substantial opportunity in the software layer for physical AI, focusing on governance, orchestration, and security for interacting physical agents. Mr. Mathew detailed physical AI’s future, involving physics-based and “world models” that transcend traditional transformer architecture, leveraging diffusion models, images, and videos to grasp physics and dimensionality.

Conventional wisdom is rapidly evolving. Ms. Fontaine advocates a “fit-for-purpose” multi-model strategy, combining large and small models, a trend already adopted by 85% of enterprises. Mr. McLoughlin notes a departure from strict margin fixation, with venture capital now flowing more freely, even accepting negative margins. Ms. Fontaine highlights the growing importance of early strategic partnerships, even at the seed stage, driven by the demand for early revenue, elevating the role of Corporate Venture Capital (CVCs). Mr. Mathew concluded by suggesting the potential end of the SASS cycle, with new economics driven by token generation and API usage, exemplified by Anthropic’s revenue from agentic coding.