Video summary

Authors, You Need to Learn This AI Skill

Main summary

Key takeaways

Educational

Main ideas, concepts, and lessons

  • Authors should learn AI literacy: stay current with AI developments or risk falling behind; start early to gain “first mover advantage.”
  • Two key AI-driven approaches:
    • Vibe coding: using AI to write code (software, tools, even game code), enabling non-programmers to prototype and build.
    • AI agents / AI harnesses: systems where AI can make decisions and take actions beyond fixed chatbot steps, based on provided guidance.
  • Software will become personalized: people who can assemble/customize AI-powered systems will gain major advantages—both professionally and creatively.
  • Writing is expanding beyond traditional silos: writers can become “story directors,” coordinating multiple media forms (books, scripts, games, etc.).
  • Monetization opportunity: learning to build and customize AI systems for others (e.g., businesses, freelance platforms) is a major near-term way to earn money.
  • Personalization over “one tool for everything”: rather than choosing a single AI product, combine multiple tools by connecting them to a shared structure (folders/data layout).
  • A practical product-building workflow: build an AI-driven “operating system” using iterative steps—structure → design → development (MVP) → testing/refinement.

Methodology / instructional steps (detailed)

1) Build AI literacy (starting point)

  • Commit to learning AI developments relevant to creative work and tooling.
  • Treat AI as a skill set to master, not just a novelty.

2) Use vibe coding and agents as core capabilities

  • Learn to use AI to generate code even if you’re not a developer.
  • Use AI agents for tasks that require ongoing action and decision-making rather than scripted automation.

3) Choose tools—but don’t lock yourself into one

  • Tools mentioned as options to combine:
    • Open Claude / Claude Code (for coding/vibe coding)
    • Hermes (for continuous/running tasks)
    • ChatGPT desktop app (daily interactive work, including voice and browser/computer control)
    • Grok (another bot option)
  • Key principle: things change quickly, so test and swap tools over time.
  • Avoid dependence on one platform by wiring tools into the same system structure.

4) Create shared folder structure to unify multiple AI tools

  • Set up a common folder structure so different AI tools/agents can operate over the same repository/data.
  • Implement careful usage (e.g., avoid running multiple tools simultaneously if they conflict), while still allowing tool switching.

5) Define what each tool does in your system

  • Claude Code: software development / “vibe coding.”
  • Hermes:
    • Runs continuously for automation and scheduled/recurring tasks.
    • Uses always-on network storage (personal server/NAS repository).
    • Examples: daily summaries, end-of-day notes, respond to a text message when needed.
  • ChatGPT desktop app:
    • Daily interface for tasks.
    • Strengths highlighted: image generation, computer/browser control, and especially voice mode as an “orchestrator.”
    • Voice mode lets you continue work while away from the keyboard by managing threads/processes.
  • Obsidian:
    • Local, markdown-based knowledge base storage.
    • Chosen because AI handles markdown well, and files remain human-readable.
    • Knowledge base folders are accessible to AI for viewing/updating; also manually browsable/editable by the user.

6) Establish an “Enterprise” root system with multiple sub-systems

  • Create a root called the “USS Enterprise system” (Star Trek themed).

  • Under it, create multiple dedicated folders (“OS” modules), each behaving like separate software connected through a “Bridge” control panel:

    • Personal OS (also “Captain’s cabin”)

      • Knowledge base folder, projects folder, additional needed folders.
      • Instruction-like markdown files (examples: claw.md, agents.md, design.md) to guide AI identity/purpose.

    • Team OS (“team”)

      • For team coordination: task management, support, communication tools.

    • Fiction OS

      • Fiction writing workflows: reviewing, editing, and related processes.

    • School OS

      • Dedicated to the Story Hacker group: gather data, identify key questions; AI helps but won’t “write answers.”

    • YouTube OS

      • Video research and production support: hooks, idea generation, and potential visual accompaniment.

    • Bridge

      • The personal control panel that brings the system together.

  • Hosting/continuity:

    • This structure is hosted on always-on server storage (Ugreen NAS).
    • ChatGPT desktop app and Claude Code are already connected to it.

7) Follow an MVP-driven build/design cycle

  • Step 3: Design

    • Decide desired design early so AI can create files that carry through the system.
    • Suggests design first for minimum viable product flow (noting delays can happen, but recommending you do it earlier).

  • Step 4: Development

    • Combine with design because design and implementation iterate.
    • Use input/output thinking:
      • Ensure each feature correctly handles expected inputs and produces correct outputs.
    • Example: a diary app workflow
      • User writes diary entries manually.
      • User later takes a photo and sends it to an agent.
      • Agent transcribes text and adds it to the knowledge base.

  • Step 5: Testing and improvement

    • Continuous refinement—never fully “finished.”
    • Build, test, refine, and then add enhancements.

8) Prototype a custom dashboard (demonstration of “custom software”)

  • A Star Trek-inspired dashboard prototype is described (with dummy data).
  • Display components include:
    • Three main daily goals
    • Meetings list from Google Calendar
    • Backup tasks to replace/augment goals
    • Greeting (e.g., “Good afternoon, Jason”) and motivational quote (“I want to become the best.”)
    • Study group data
    • YouTube data
    • Amazon KDP account data
  • Core lesson: there is no off-the-shelf tool that merges all these needs; therefore, build personalized software to reduce business friction.
  • Encouragement: authors should consider custom systems to spend more time on what only they can do.

Speakers / sources featured

  • Speaker / narrator: Jason (referred to repeatedly; e.g., “Good afternoon, Jason,” “Okay, Jason”).
  • Named organizations/tools/systems:
    • OpenAI / ChatGPT (specifically “ChatGPT desktop app” and “voice mode”)
    • Claude Code (Claude Code)
    • Open Claude (mentioned as an option)
    • Hermes (AI agent/harness)
    • OpenK3w / Open Claw (mentioned as a famous agent environment; subtitle appears as “Open Claw”)
    • Grok (mentioned as a bot option)
    • Obsidian (note-taking/markdown knowledge base)
    • Ugreen NAS (always-on storage/server)
    • Google Calendar
    • Amazon KDP
    • Upwork
  • Creators recommended:
    • Alex Finn
    • Nate Gurk
  • Community mention:
    • “Story Hacker” / “School membership” (Story Hacker group; referenced as the speaker’s School membership space).

Original video