LangChain founder advises investors to target inference and vertical agents
Harrison Chase argues that inference providers and domain-specific agents represent the most promising AI investment opportunities. He warns against funding crowded customer support AI and robotics sectors, which he views as overfunded relative to their capabilities.
Harrison Chase, founder of LangChain, has outlined a strategic framework for investors navigating the current AI landscape. Chase identifies inference and coding agents as the two markets he most underestimated, describing the demand for inference as massive and unlikely to decrease soon. He advises investors to focus on inference providers, lower-level infrastructure primitives such as Modal, Temporal, and Superbase, and vertical agents that possess deep domain expertise. Chase argues that durable startup success in this era requires strong technical talent, rapid execution speed, and a brand advantage built from being first, citing LangChain, Cursor, and Instinct as examples of companies that leveraged early viral success to create a durable moat despite the emergence of 100 competing frameworks.
Conversely, Chase warns against investing in customer support AI, noting that 20 funded providers have raised over $100 million in the last 18 months. He also views robotics as significantly behind its hype. Chase emphasizes the importance of strong venture capital partnerships, describing Benchmark as a 'third co-founder' focused on core company building and Sequoia as a key resource for recruiting and marketing. His analysis suggests that while broad AI applications are crowded, the underlying infrastructure and specialized vertical applications offer more sustainable growth potential.