A Data-Driven Course on AI Proficiency

From Prompts
to Pipelines

The AI Skills Nobody Teaches

Based on 8,043 AI conversations and 278,532 messages across 39 months on 4 platforms. A practitioner's guide to the skills between "asking questions" and "writing code."

Explore the Course
0
Conversations
0
Messages
39
Months
4
Platforms

Ten thousand courses teach you
how to prompt. Zero teach you
what comes next.

Saturated
Level 1 - 4
10,000+ courses, tutorials, bootcamps
Nothing
Level 5 - 8
The skills that matter most
CS Programs
Level 9 - 10
PhD-level AI/ML engineering
Browser AI is a smart advisor.
Terminal AI is a workforce you command.

8,043 Conversations.
278,532 Messages.
39 Months. 4 Platforms.

Every conversation logged. Every platform tracked. This is what the journey from Level 1 to Level 8 actually looks like.

0
ChatGPT
Jan 2023 - Oct 2025
0
Claude Code
Nov 2025 - Mar 2026
0
Claude.ai
2024 - 2026
0
Grok
2024 - 2025

Monthly Conversation Volume

Jan 2023 - Mar 2026 across all platforms

ChatGPT
Claude.ai
Grok
Claude Code
Rating Trajectory
39 months of measured growth
1
Jan 2023
8
Mar 2026
Platform Share Shift
From single-platform to orchestration

7 Phases. 39 Months.
One Practitioner's Path.

Not a curriculum. A lived experience, measured and documented in real time.

Phase 1
January 2023

First Contact

The first ChatGPT conversation. Simple questions, simple answers. The novelty phase that nearly everyone experiences but few move beyond.

48 conversations
Phase 2
March 2023

First Explosion

Something clicked. 374 conversations in a single month. Every workflow got tested. Every use case explored. The compulsion phase.

374 conversations/month
Phase 3
2023 - 2024

The Working Professional

AI became infrastructure, not novelty. Sustained 150-200 conversations per month across instruction design, writing, and analysis. The tool became invisible.

150-200 conv/month sustained
Phase 4
Jul - Sep 2025

Second Explosion

GPT-5 era. New capabilities demanded new approaches. Multi-platform usage surged. Claude, Grok, and specialized tools entered the stack.

300+ conversations/month
Phase 5
October 2025

The Terminal Leap

The moment everything changed. Moved from browser-based chat to terminal-based AI. Claude Code replaced ChatGPT as the primary tool. AI stopped advising and started building.

The inflection point
Phase 6
Nov 2025 - Feb 2026

Orchestration

MCP servers, multi-agent systems, automated pipelines. AI became a managed workforce, not a conversational partner. 1,190 Claude Code sessions in 5 months.

1,190 sessions in 5 months
Phase 7
March 2026

Mastery

Personal frameworks codified. Teaching methodology developed. The ability to not just use AI, but to architect systems of AI agents working in concert. Level 8.

Level 8 / 10

6 Modules.
From Level 3 to Level 8.

A practitioner-built curriculum that bridges the gap between "I use ChatGPT" and "I orchestrate AI systems."

👋
Module 01

First Contact

Level 1 - 3

AI fundamentals, platform selection, and the art of the first conversation. Learn why most people plateau here and what makes the difference between casual users and power users.

🌐
Module 02

Multi-Platform Exploration

Level 3 - 4

Prompt compression, custom instructions, system prompts. The techniques that separate effective communicators from everyone else. Why one platform is never enough.

💻
Module 03

The Terminal Leap

Level 4 - 5

Installing Claude Code. Your first terminal session. The moment AI stops being a chatbot and becomes an engineering partner. This is where everything changes.

Module 04

Integration

Level 5 - 6

APIs, automation scripts, CLI tools, and workflow integration. AI touches every system you work with. Build the connective tissue between AI and your existing stack.

🎯
Module 05

Orchestration

Level 6 - 7

MCP servers, multi-agent architectures, memory systems, and managed AI workforces. Move from using one AI to commanding many. This is the workforce paradigm.

🧠
Module 06

Methodology

Level 7 - 8

Developing personal frameworks, codifying institutional knowledge, and teaching AI proficiency to others. The transition from practitioner to architect.

Where do you fall?

A 10-point scale based on observed behavior patterns across 8,043 conversations. Most people never leave Level 3.

10
Pioneer
<100
9
Architect
~1,000
8
Orchestrator
YOU COULD BE HERE ←
~5,000
7
Integrator
~25,000
6
Builder
~100,000
5
Terminal User
~500,000
4
Power User
~5M
3
Regular User
~50M
2
Experimenter
~200M
1
First Contact
~500M+

The 5-Part
LinkedIn Series

The full analysis, published as a series. Data, methodology, frameworks, and the case for a new kind of AI education.

Part 1

I Talked to AI 8,043 Times. Here's What I Learned.

Read article →
Part 2

From Prompts to Pipelines: Why Nobody Teaches Levels 5-8

Read article →
Part 3

The Terminal Leap: When AI Stops Advising and Starts Building

Read article →
Part 4

From Asking to Commanding: The Language of AI Mastery

Read article →
Part 5

The Course We Built (And Why We Built It)

Read article →
CM

Dr. Charles Martin

Ed.D. Educational Technology

8,043
Conversations
Level 8
AI Rating
39 mo
Journey

Dr. Charles Martin holds an Ed.D. in Educational Technology from the University of Florida, an MA in Middle Grades Education from Georgia State University, and certifications in school leadership from USC. His career spans K-12 teaching in Abu Dhabi, instructional design at Age of Learning and Microsoft, intelligence work at DNI, curriculum development for the US Army, and clinical education at UCLA Health.

Today he serves as an Instructional Systems Designer at Eccalon (federal contracting) and VP at ModelIt through Discovery Collective, pioneering AI-powered workflows in education and defense training.

  • Ed.D. Educational Technology, University of Florida
  • ISD at Eccalon (DoD federal contractor)
  • VP at ModelIt / Discovery Collective
  • 10+ years K-12 teaching (including Abu Dhabi)
  • Microsoft, DNI, US Army, UCLA Health alumni
  • Founder, Alexandria's Design LLC

8,043 conversations. Level 8/10. And counting.