I Built a Team of AI Employees (They Actually Work)
The speaker demonstrates an AI team built using the ICORE methodology, where specialized AI agents (Larry, Pixel, Nolan, etc.) work together using Claude's API within a structured framework. The system leverages natural language interaction, MCP servers, and business knowledge management to automate tasks while maintaining human oversight, representing a shift toward accessible AI implementation for non-technical users.
Summary
The speaker presents a functioning AI team composed of specialized agents that operate according to the ICORE project management methodology. Each agent has defined characteristics including personality, intelligence level, reliability, and motivation that update dynamically based on interactions. The team includes Larry (manager), Pixel (designer), Nolan (HR), Quinn (quick research), Bus (video editor), Pen (YouTube scriptwriter), and others, each with access to specific tools like Perplexity, Metricool, and Nano Banana Pro.
The infrastructure runs on a SQLite backend with a Telegram interface, utilizing two Claude Max plans ($200/month total) rather than expensive API calls. Importantly, the system requires manual human interaction to comply with Claude's terms of service, avoiding autonomous autopilot operation. The speaker emphasizes that the visual representation is secondary—the core is Claude functioning through natural language commands.
The ICORE framework provides the structural foundation enabling effective delegation and team communication. All agents operate within this methodology, understanding goals, projects, and team dynamics. The system includes a business knowledge management (BKM) system that stores reference images, branding guidelines, and context, enabling consistency across outputs. Tasks are created automatically in Click Up with full context awareness, allowing team members to track progress and updates.
The speaker positions this as a template for human team management rather than pure automation—the skills required (delegation, management, organization) are identical whether managing humans or AI. He argues that 2026 will be a turning point when Apple launches Gemini on iOS/macOS, making AI adoption mainstream. The speaker concludes that while AI operates from existing patterns rather than true innovation, humans remain the creative force, with AI handling friction points like administrative work and research, freeing humans to concentrate on their strengths.
Key Insights
- The speaker runs an AI team on two Claude Max plans ($200/month total) rather than individual API calls, which he describes as 'insane' and unsustainable in cost comparison, demonstrating that subscription models are more efficient for multi-agent systems.
- All AI agents' motivation levels update automatically based on interaction tone—if the speaker is consistently angry, motivation drops—creating a feedback loop that mirrors human employee dynamics.
- The speaker deliberately maintains manual human interaction rather than full automation to comply with Claude's terms of service that restrict accounts running on autopilot, yet argues this limitation 'goes a long way with the right skills.'
- The speaker claims that skills required to manage AI agents (delegation, management, organization) are identical to managing human employees, and that treating AI 'much closer to human employees than ever before' closes the gap between rigid automation and flexible natural language systems.
- The speaker predicts 2026 as a pivotal moment when Apple's launch of Gemini on iOS/macOS will shift AI adoption to mainstream non-technical users, similar to how ChatGPT succeeded because users could 'talk whatever they had on mind' without needing prompt engineering expertise.
Topics
Transcript
[0:00] you into AI. This is the moment. This is the moment. Well, let me introduce you, Larry. [laughter] Yet another AI personal assistant and so on. No, this is not what's all about. Today, we will talk about how I got to this point, how the potential is there for all the people out there leveraging AI in 2026, and it will go wild. What I'm showing here is an AI team working [0:32] according to the ICO methodology. It knows perfectly what a goal, a project, and all these things are because that's what we define in ICORE. It knows perfectly how team communication works, all these things. And when I started building this, it it was literally…
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.