Harrison Chase (Co-founder & CEO of LangChain) joins Jay to discuss the evolution from LangChain's Twitter origins to becoming the infrastructure backbone for thousands of production agents. Harrison talks about how LangChain started on Twitter and quickly grew into a multi-product ecosystem including; LangChainLangSmithLangGraph, and LangGraph Platform. The conversation revolves around deep agents, permission models, the issue with memory, and what the future holds for the coding space.Harrison explains why competing directly with model providers on their specialized domains (like Anthropic's Claude Code) is nearly impossible, but argues the real opportunity lies in UX innovation and bringing these capabilities into existing workflows.Tune into the full episode to learn why memory isn't the bottleneck yet and how the bitter lesson applies to agent architecture!HIGHLIGHTS: 0:00 Intro 1:24 LangChain's evolution from Twitter prototype to production platform 3:23 Model capabilities progression from 2023 to today 4:05 Deep agents - planning, subagents, and file systems for long-term tasks 6:37 Why string replacement beats line-by-line editing for Claude 8:14 The impossible challenge of competing with Claude Code directly 11:28 UX differentiation and workflow integration as winning strategies 13:55 Unix commands and composability for non-coding agents 16:14 Sandboxing approaches - individual VMs vs shared environments 20:03 Agent runtime primitives - streaming, human-in-loop, time travel 22:55 Why CLI tools might beat MCP for agent interactions 25:13 Why base performance matters more than persistence 28:10 External vs model-weight memory systems for auditability 30:12 Product admiration - from cooking to Cursor's UX masteryConnect with Harrison - https://www.linkedin.com/in/harrison-chase-961287118/Connect with Jay - https://www.linkedin.com/in/jayhack/ or https://x.com/mathemagic1an Visit https://langchain.com/ for agent development tools

Ver todos los episodios