Stanford's Method Turns Claude Into a PHD Level Research Team
A researcher demonstrates the STORM method from Stanford, which uses five expert perspectives (practitioner, academic, skeptic, economist, historian) to create verified research reports. The method produces 25% more organized articles than competing approaches and is packaged as a reusable Claude skill that generates HTML briefings with peer-reviewed citations.
Summary
The transcript presents a detailed walkthrough of the STORM research methodology developed at Stanford, which the speaker has adapted into a Claude skill for AI-assisted research. The core innovation is using multiple agent perspectives simultaneously to identify blind spots that single-perspective research would miss. Each agent role (practitioner, academic, skeptic, economist, historian) brings different expertise and identifies gaps the others overlook.
The speaker compares STORM to Claude's native Deep Research feature, demonstrating that while Deep Research spins up 100+ agents to produce a markdown report with limited sources, STORM uses a more structured 5-agent approach followed by 6 verification agents to produce a more reliable, organized HTML briefing. In their head-to-head comparison, another AI model (Codex) rated the STORM output superior in six categories: evidence quality, source diversity, thesis strength, actionability, risk control, and content usability.
The methodology works through four sequential prompts: first spinning up the five expert lenses, second creating a contradiction map to identify where perspectives disagree, third synthesizing everything into a single report, and fourth conducting adversarial peer review that verifies citations against primary sources. The skill incorporates an HTML template for consistency.
The speaker emphasizes this skill is freely available in their community and can be installed in Claude's .Claude folder. He demonstrates the live execution showing how subagents run in parallel, how to monitor their research in real-time, and how the final output ranks findings by reliability based on which perspectives supported or challenged each claim. He also explains the distinction between subagents (which only communicate with the main session) and agent teams (which can communicate with each other), noting that agent teams are more expensive.
The broader lesson emphasized is not just about this specific skill, but about the principle that multiple contradicting perspectives conducting research together produce more holistic, accurate results than single-perspective approaches, and that using agents to borrow subject matter expertise can help overcome knowledge gaps.
Key Insights
- Stanford's STORM method produces articles 25% more organized than the next best method through peer-reviewed testing by simulating five distinct expert perspectives that each identify holes the others miss.
- STORM requires a verification phase where sources are not just collected but actively confirmed, corrected, or demoted based on accuracy, whereas comparable methods like Deep Research produce unvetted brain dumps of statistics.
- When comparing STORM's HTML briefing output against Claude's Deep Research on the same topic, a third AI model (Codex) rated STORM superior across six categories while using 100+ fewer agents and being 100% cheaper to run.
- The skill automatically identifies its own blind spots by noting missing lenses—for example, all five original perspectives analyzed the research from ownership/ROI viewpoint, missing customer and frontline employee perspectives entirely.
- Subagents architecture (where multiple agents work for one main session but cannot communicate with each other) differs fundamentally from agent teams (where agents can debate each other to consensus) and costs significantly less while still enabling parallel research.
Topics
Transcript
[0:00] So Stanford has a research method called storm, which has actually been shown in peer-reviewed testing to produce articles 25% more organized than the next best method. So I put all of those storm principles into my own Claude skill, which I'm going to give you guys for completely free, and you end up with the result that looks like this. It is an HTML briefing that has been put together by five different perspectives of agents, and it has been verified. Meaning if I scroll down to the bottom, you can see that the different perspectives are giving analysis on each parts of the report. But at the very bottom, you can see that we have different sources…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Nate Herk | AI Automation
Every Codex Concept Explained for Non-Coders
A comprehensive tutorial explaining 18 core Codex concepts for non-technical users, organized into four parts covering basics (projects, agents, agent loops, goals), environments (on-premises vs. cloud, working trees), customization (AI models, effort levels, permissions, skills), and scaling tools (browser, sites, subagents, scheduled tasks, voice mode).
Claude Code Mods Are Game Changers. This One Saves Me Money.
Claude Code now supports mods that allow customization of the Cloud desktop app. A user demonstrates a cache management mod that tracks usage data, displays costs, monitors session limits, and provides notifications before cache expiration to help optimize spending.
5 Claude Code Mods That Everyone Needs
The video demonstrates five useful Claude Code modifications (mods) that are plugins enabling customization of the Claude desktop application interface. The creator showcases practical mods including Cache Keeper for token management, Recording Mode for privacy, Goal tracking interface, and Collision Guard for file conflict detection, explaining how anyone can create mods using plain language.
How to Actually Build & Sell Software with AI as a Non-Techie
Dave Fabrikant, a 10+ year Python programmer and AI engineer, discusses how AI coding tools have transformed software development from a specialized skill into an accessible field for non-technical founders. He explains the progression from personal tools to scalable products, emphasizing architecture, security best practices, and the shift from specification-driven to intent-based development.
I Tested Codex's NEW $500/mo Ultrafast mode
A reviewer tests Codex's new $500/month Ultrafast mode, which claims to be 8x faster than standard mode. While results vary by task complexity, Ultrafast delivers significant speed improvements but at substantially higher resource consumption, making it suitable primarily for time-sensitive projects.