Uploaded February 2026 | Updated September 2026, 2 weeks ago
A special double pod on the 1 year anniversary of Claude Code: we chat with one of its most vocal fans, who thinks it will write 25-50% of all code on GitHub, plus get a breakdown on the memory crunch
00:00 AI Slop vs Expertise
01:00 Reconnecting and Origin Stories
03:40 Falling for Semiconductors
06:44 Moores Law Thesis and Nvidia
12:31 Claude Code Awakening
33:15 Agent Swarms Reality Check
37:44 Claude Bot Security Limits
41:13 Claude Code Workflow Setup
46:53 Hygiene and Junior Analysts
59:41 GDPval and Economic Impact
01:06:51 Railroad CapEx Parallel
01:08:07 Funding Bubbles and Demand
01:09:36 AI as Junior Analysts
01:11:32 Death of the IDE
01:23:43 Microsoft and Oracle Stakes
01:36:46 TPU Window Opens
01:38:11 Supply Chain Reality Check
01:41:53 HBM Memory Squeeze
01:48:14 Context Rationing Era
01:57:43 Writing And Trail Lessons
A special double pod on the 1 year anniversary of Claude Code: we chat with one of its most vocal fans, who thinks it will write 25-50% of all code on GitHub, plus get a breakdown on the memory crunch
00:00 AI Slop vs Expertise
01:00 Reconnecting and Origin Stories
03:40 Falling for Semiconductors
06:44 Moores Law Thesis and Nvidia
12:31 Claude Code Awakening
33:15 Agent Swarms Reality Check
37:44 Claude Bot Security Limits
41:13 Claude Code Workflow Setup
46:53 Hygiene and Junior Analysts
59:41 GDPval and Economic Impact
01:06:51 Railroad CapEx Parallel
01:08:07 Funding Bubbles and Demand
01:09:36 AI as Junior Analysts
01:11:32 Death of the IDE
01:23:43 Microsoft and Oracle Stakes
01:36:46 TPU Window Opens
01:38:11 Supply Chain Reality Check
01:41:53 HBM Memory Squeeze
01:48:14 Context Rationing Era
01:57:43 Writing And Trail Lessons







![⚡️Context Graphs: according to the authors — Jaya Gupta, Ashu Garg, Foundation Capital
In this Lightning pod, swyx hosts Jaya Gupta and Ashu Garg from Foundation Capital to discuss the emergence of context graphs. They define this new framework as the institutional memory of the why behind business decisions, captured through decision traces—the sequence of steps and human reasoning that models often miss. The conversation explores how these graphs will become the defensible moat for the next generation of applied AI companies and systems of agents.
Section Timestamps
[00:03] – Introductions and the early vibe of AI hackathons post-ChatGPT.
[02:00] – The origin story of the Context Graph thesis at Foundation Capital.
[04:59] – Defining the Context Graph and the Decision Trace.
[07:32] – Who is building this today? Examples like Player Zero and Glean.
[09:37] – Technical implementation: Is there an ideal data structure?
[12:09] – Explaining Systems of Agents vs. standard chatbots.
[15:26] – The importance of the Right Path (operational) vs. the Read Path (analytical).
[18:46] – Why these will be new platforms rather than features in Slack or GitHub.
[21:48] – Addressing pushbacks: Can you truly capture the Why or just the How?
[24:04] – Privacy, data governance, and Metadata 3.0.
[26:43] – Context Graphs vs. Data Mesh: Why universal graphs are unlikely.
[31:18] – 2026 Predictions: The Context Graph Stack and production scale.
show notes
https://foundationcapital.com/context-graphs-ais-trillion-dollar-opportunity/
https://x.com/JayaGup10/status/2003525933534179480
https://simple.ai/p/what-are-context-graphs ⚡️Context Graphs: according to the authors — Jaya Gupta, Ashu Garg, Foundation Capital](https://i.ytimg.com/vi/zP8P7hJXwE0/mqdefault.jpg)
