A Maneira MAIS INTELIGENTE De Organizar Arquivos Com IA

How the “Brain” system uses AI to organize a Drive automatically

The creator describes an AI-driven file organization workflow that “organizes itself” in a self-governing way. The system is built around a fixed folder structure that does more than categorize files—it triggers automation. Each folder behaves differently, and the folder connections form a “system that governs itself.”

Core folder structure (the “Paria” structure)

Six main folders form the foundation:

1. Projects

   • Holds actionable work-in-progress items and project context for AI-assisted execution.
   • Project types:
     • One-time, e.g., developing an application
     • Recurring processes, e.g., ad production
   • Workflow:
     • The AI generates/edits assets (images, carousel/video elements).
     • Then completed outputs are brought back and organized into the most relevant project folders.
   • Example: from “story” inputs, the AI produces items and returns them into the project folders needed for that week’s tasks.

2. Areas

   • For ongoing responsibilities without a completion deadline (maintenance/pace-based).
   • Access patterns:
     • e.g., monthly finance checks
     • renewing certificates when needed
   • Examples:
     • documents get updated when a task like certificate renewal occurs.
   • Treated as lower-frequency than Projects and Memory.

3. Resources

   • A reuse library that prevents starting from scratch.
   • Contains standardized materials needed to execute work (past images/photos, client testimonials, proof, data from spreadsheets).
   • Design goal: standardized reuse so outputs can be regenerated into many future results.
   • Examples:
     • Frameworks: a single design created in ~1 hour (e.g., Photoshop) reused across mentoring sessions, videos, and proposals.
     • B-Roll / Birollose (cinematic language): reusable supporting clips to avoid reshooting scenes; repeated points can improve quality without wasting time.

4. Inbox

   • A temporary holding area for intermediate files.
   • Example: a screen recording from Screen Studio stays here until the final deliverable is produced.
   • Routine: the AI clears the inbox daily by deleting finalized temp items, sometimes leaving items in Trash for later potential reuse as resources.

5. Archived

   • Keeps past items while keeping active folders clean.
   • Example counts: ~63 archived projects vs. ~15–20 active projects.
   • Benefits:
     • Retrieval of past information when needed.
     • Reusing past assets (e.g., anniversary materials from earlier years).

(Also mentioned: additional “governing” files—around 14—used to control behavior, though fewer core ones might be sufficient.)

“Memory” concept and how it’s simulated

The speaker argues that AI “memory” shouldn’t be limited to a user-settings note or a single “memory file.” Instead, memory is broken into human-style types (working, episodic, long-term), and AI can simulate them.

• Example: a “Cronos” folder as chronological memory
  • The AI automatically creates a daily note and saves context such as:
    • decisions made
    • tasks completed
    • priority changes
  • Saving happens via an implied rule, so the user doesn’t need to manually store everything.

Governing files: “single sources of truth” and self-improving behavior

To let the system run independently, it uses a set of governing files that teach AI patterns, rules, and feedback loops.

• These include design and video editing standards:
  • design.m: rules/preferences for creating designs, carousels, covers, etc.
  • edit.md: video editing rules including motion effects and selecting key points.

The speaker frames this as an information architecture principle:

All types of info should have one canonical location (“single location to be saved and to be found”).

This supposedly makes the system faster because AI doesn’t search the entire system.

How the AI is guided to use the right folders/files

Folder semantics direct usage:

• Projects → projects page
• Memory → memory folder
• Areas → areas folder

There are also identity/values files (e.g., “mi.mmd”):

• AI uses persona/essence/role/values to filter and refine responses.
• Example filtering: if the creator values family over work, the AI avoids generic “work harder” advice and instead tailors outputs to what matters.

Combining inputs (“perception”) to generate real outputs

The speaker emphasizes a “chat skills” approach:

• Use one chat (across cloud/GPT while still connected to Google Drive).
• The user guides the AI to the correct location at the moment it’s needed.

They also stress perception as a skill: teaching AI to combine the correct files for a task.

• Example shown in a “processes” file (before writing a script):
  • read avatar
  • read brand
  • read testimonials
  • read old content
  • read what the creator is studying
  • read mentoring notes
• Feedback loop:
  • after script generation, the creator provides feedback
  • the AI records that feedback to improve future outputs (learning “in real time”)

Product/education call-to-action (review/guide/tutorial elements)

The video promotes a “Brain workshop” (Sep 2–3) focused on installing the system.

• Mentions a “Brain guide” download outlining the “three initial steps” to set up the Brain in a favorite AI.
• Workshop claims:
  • Day 1: AI already knows everything about the user/company/work after installation
  • Day 2: automation becomes active so work can proceed without asking
  • Includes live installation with provided files and live support

The content is not positioned as a traditional consumer “review,” but rather as a tutorial/implementation guide for an AI “file system + instruction set.”

Main speakers / sources

• Speaker: the video’s author/creator (unnamed in subtitles), discussing “Alma Brand” and the “Brain workshop” / “Brain guide.”
• Referenced author/source: Thiago Forte (mentioned in comparison to folder/second-structure ideas).