How I review and approve my AI team's work
The creator demonstrates how an AI team generates interactive HTML review dashboards inside a local folder system, allowing him to review, approve, and copy social media campaign content without writing any code. The HTML file includes embedded videos, custom images, and copy-to-clipboard functionality for scheduled posts. This workflow is built entirely through natural language instructions to AI team members.
Summary
The creator walks through his personal workflow for managing AI-generated deliverables using a local folder system he runs his entire life and business from. He references a previous video explaining the full folder structure and focuses here specifically on the 'deliverables' folder, where AI team outputs land for his review and approval.
The centerpiece of the demonstration is an HTML file generated for a social media launch campaign called 'One Life, One Folder.' Rather than reviewing raw markdown or text files, the creator opens a single HTML file in a browser, which renders as a fully interactive campaign review dashboard. The dashboard allows him to navigate by scheduled date, preview posts, copy content to clipboard with a click, view associated images, and watch embedded video clips cut from longer source videos.
He explains that his AI team includes specialized members: 'Larry' who handles general tasks and generates the HTML review interface, 'Real' who cuts video clips from a database of existing footage, and 'Charter' who creates custom infographics and images. All of this is orchestrated through natural language conversation, with no coding on his part.
The creator clarifies that while HTML is powerful for deliverable review interfaces, he does not convert his entire knowledge base to HTML — his personal knowledge management system stays in Markdown with Obsidian wiki-links for interconnectedness. He also shows the workflow running inside VS Code terminal, demonstrating that the same HTML preview works without leaving the development environment. He mentions a free scaffold available through 'myICOR' that lets others replicate this setup, with expansion packs planned for the HTML deliverable review format.
Key Insights
- The creator runs his entire life and business from a single local folder combined with Claude, and his co-founder Paco independently uses the identical folder structure to manage his own life and four businesses.
- Rather than delivering raw markdown files for review, the creator's AI team generates a fully interactive HTML dashboard that lets him navigate by date, preview posts, copy content to clipboard, and watch embedded video clips — all from a single local file.
- The creator has a specialized AI team member called 'Real' who can cut clips from a database of existing videos, enabling natively formatted short-form video content for platforms like X and LinkedIn to be embedded directly in the review dashboard.
- The creator explicitly argues against converting his entire PKM system to HTML, stating that Markdown with Obsidian wiki-links is essential for maintaining a deeply interconnected, AI-automated knowledge graph — HTML is only appropriate for deliverable review interfaces.
- The entire HTML review interface, including embedded videos, custom images, and interactive copy buttons, was produced through natural language instructions to AI team members with zero coding knowledge from the creator.
Topics
Transcript
[0:00] guys, I'm shocked, and I'm the one who created this, okay? In a previous video, I showed you that I run my whole life and business just from within this one local folder and Claude. And today I want to show you how powerful it can get just having a local folder when it comes to working with AI. So if you just getting started with working AI, or if you already are advanced, you might be surprised what I'm about to share with you, what you can actually do without using any additional tools in addition to this local folder. [0:31] for those who didn't watch the video, I highly recommend to watch the video because it's a…
Full transcript available for MurmurCast members
Sign Up to AccessMore from ICOR with Tom | AI Productivity
I Made Opus and Fable Grade Each Other. Opus Admitted It Lost.
The creator demonstrates real-world AI applications in their productivity business using Claude Fable and Opus, showing how Fable excels at comprehensive analysis of complex systems while comparing its performance to Opus on specific tasks. The key finding is that Fable's holistic understanding justifies its higher cost for business-critical audits and optimizations.
Our AI invented the numbers. We caught it.
The hosts discuss the importance of double-checking AI outputs and validating information, emphasizing that this is not a new problem but rather a longstanding practice required with any information source. They explore how AI tools can be customized to individual workflows and contexts, arguing that understanding foundational methodology matters more than specific tools.
An AI Just Handed Me a Fake $67B Statistic
The speaker shares how an AI confidently generated a completely fabricated $67.4 billion statistic while researching AI hallucinations, then explains his systematic approach to combating AI misinformation: using dual independent search engines to cross-check answers and identify where hallucinations occur.
AI Is Quietly Taking Your Memory
AI companies are racing to own users' context and memory through integrated tools like Claude in Slack and ChatGPT's auto-memory features, creating a lock-in trap. The speaker argues users should own their context (expensive, irreplaceable) in local plain text files while renting the AI model (cheap, replaceable), and demonstrates his personal knowledge assistant system built on markdown files that work with any AI model.
My AI Team Now Has an Interface. All 12 Agents. Free.
Tom, creator of the iCola methodology, releases My PKA version 3 with a free interface (My PKA Cockpit) that provides 12 pre-built AI agents for productivity, note-taking, health tracking, document management, and task planning—all running locally in a folder structure accessible via Claude and a web interface.