Claude just killed ALL Note-Taking Apps. Here is proof.
A video demonstrates how Claude AI and a simple folder structure replace traditional note-taking and PKM tools by creating a personal knowledge assistant with an AI team. The speaker shows building a local database, organizing files, and creating custom interfaces entirely through natural conversation with Claude, eliminating the need for apps like Notion, Obsidian, and Heptabase.
Summary
The speaker, who has extensively used various note-taking and PKM tools over years (Rome Research, Obsidian, Notion, Mem, Tana, Heptabase), argues that the era of traditional PKM tools is ending. He claims these tools lock users into developer-defined structures and scaffolds, requiring dependency on plugins and external developers for customization. The fundamental shift involves replacing PKM (Personal Knowledge Management) with PKA (Personal Knowledge Assistance) using Claude AI.
The core setup is remarkably simple: a basic folder on a computer plus Claude. The speaker creates a folder structure with three main directories (owner's inbox, team inbox, team folder) and uses Claude Code in the terminal to set up an AI team. Through natural language conversation, he establishes Larry as an orchestrator AI agent, who then hires team members like Pax (researcher) and Nolan (HR director). Each agent has a persona, identity, and specialized role.
Critically, the speaker emphasizes that the data remains completely local and portable—not locked into any proprietary format. The system creates a SQLite database that stores all information, which can then be visualized through simple HTML interfaces built by hired AI agents. When the speaker demonstrates uploading 120+ scanned PDF files, Larry automatically organizes them, performs OCR, and indexes them in the database without any manual coding required.
The speaker shows the process of asking the AI team to hire a front-end developer (Sable) to build a custom interface for viewing the database. Initially, Sable overcomplicates it by creating a full application, but the speaker corrects course by asking for a simple HTML visualization instead. The resulting interface displays journal entries, contacts, files, and projects in a clean dashboard.
Key differentiators from traditional tools include: full local control of data, ability to swap AI providers without losing work, no plugin dependency, natural language interaction instead of learning tool-specific workflows, and the ability to build completely custom interfaces and workflows. The speaker argues this approach provides genuine ownership and flexibility that tools like Obsidian falsely promise, since Obsidian files remain bound to Obsidian's formatting and interpretation.
Key Insights
- The speaker argues that traditional PKM tools like Obsidian, Notion, and Heptabase lock users into developer-defined structures and force dependency on external plugin developers for customization, limiting true personalization.
- The speaker believes the real competitive advantage for individuals and businesses comes from employees who learn to ask AI the right questions and build their own AI workflows, not from applying AI as a generic layer on top of existing processes.
- The speaker demonstrates that Claude can perform complex organizational tasks like OCR, file renaming, and database indexing on 120+ documents automatically through natural language instruction without any manual coding.
- The speaker claims that Obsidian's promise of local file control is misleading because Obsidian files remain formatted in a way only Obsidian can properly visualize, making them effectively useless if switching tools.
- The speaker argues that the portability advantage of this AI-based system is that the data format remains tool-agnostic—users can swap Claude for Gemini or ChatGPT while keeping all their organized knowledge intact.
Topics
Transcript
[0:00] I launched this channel in 2017 with the first videos about note taking apps on the iPad. I went through all the things. I used, Rome Research. I used Obsidian, upgraded to Notion, moved on to Mem, double down on Tana and ended up in Heptabase and now all this is gone. We just entered a new era and I will show you in this video the proof why note taking apps. Absent PKM tools, as you might know it are no longer relevant in 2026. [0:31] In my previous videos I talked about it. In this video. I will show you exactly how this works and how I use it as a professional on a daily basis to…
Full transcript available for MurmurCast members
Sign Up to AccessMore from ICOR with Tom | AI Productivity
Claude sucks. And here is why.
This video explores why Claude and other AI models produce inconsistent results, examining how model selection (Haiku vs Sonnet vs Opus vs Claude 3.5 Fable), effort levels, prompt clarity, and system context dramatically affect output quality. The speaker demonstrates these differences through a Venn diagram creation task and advocates for using organized folder systems with documented SOPs and code-based solutions to ensure consistent AI performance.
Claude can now prompt its own SESSIONS. I built the boss.
Claude now has the ability to interact with multiple sessions, allowing a streamlined orchestration of tasks through an agent called Larry. This integration can simplify task management and enhance productivity by enabling a single main session to oversee various sub-sessions or tasks.
I switched off Claude's memory. My folder does it better.
Tom explains why he disabled Claude's native memory feature and instead built a local folder-based system (myPKM) to maintain persistent context with AI agents. He shares an audit revealing systemic issues like contradictory documentation, forgotten rules, and parallel session conflicts, then outlines how to systematically fix these problems using the ICOR productivity methodology.
I do not download invoices anymore. My folder and Claude do it.
The speaker demonstrates how he uses AI (Claude) integrated with his local folder system to automate invoice collection for his accountant, downloading 36 invoices in 5 minutes with minimal manual intervention. He explains his AI team methodology, work streams, and how persistent local knowledge bases enable AI to handle recurring tasks independently without constant re-explanation.
Claude finished the work. Then I lost it in the chat.
A creator demonstrates a comprehensive local-folder-based system for managing AI-generated deliverables asynchronously, featuring an interactive browser interface that tracks decisions, changes, and conversations without relying on external databases or web applications. The system enables collaborative editing, visual annotations, and automatic context preservation across multiple parallel AI projects.