The Receipts
Data extracted from 366K messages. Patterns that emerged.
Weekly Velocity
"i want to build it in 1 week"
"Great and I want to build this product in a week"
Peak: July 2025 — 12,991 messages in a single week. 151 weeks tracked total. The velocity correlates with shipping: highest output during WOTC and Brain MCP builds.
Signature Phrases
Recurring expressions that define communication style. Not what was said — how it was said.
| Phrase | Count | Insight |
|---|---|---|
| "give me the" | 789x | Direct request, expects output not process |
| "and give me" | 429x | Chaining requests, transactional style |
| "no no no" | 333x | Strong disagreement, immediate redirect |
| "please give me" | 262x | Politeness + directness combined |
| "what is the" | 251x | Direct questions expecting direct answers |
| "step by step" | 244x | Sequential, monotropic processing |
| "now give me" | 215x | Urgency, immediate delivery expected |
| "make sure to" | 206x | Precision focus, need for certainty |
| "all of the" | 180x | Comprehensive, exhaustive coverage |
| "want to be" | 180x | Aspirations, future-oriented thinking |
432 total phrases tracked across 8 categories: command, correction, process, emphasis, intention, question, evaluation.
The Core Rhythm
This pattern is consistent across 2023, 2024, 2025, 2026. The THINKING didn't change. The TOOLS adapted to match it.
Teacher + Conductor
"Mixture between kindergarten teacher and orchestra conductor. Early ChatGPT was like a garden hose. As context windows grew, it became a fire hose. Being a firm conductor has become critical."
The Evolution
As AI output grew from trickle to flood, firm conductor became critical. This is what SHELET is — the framework for conducting.
Correction Patterns
When and how AI gets redirected. The "no no no" pattern quantified.
Monthly Breakdown
| Month | Corrections | Rate |
|---|---|---|
| 2026-01 | 4 | 9.1% |
| 2025-12 | 81 | 0.6% |
| 2025-11 | 34 | 1.5% |
| 2025-10 | 72 | 1.9% |
| 2025-09 | 238 | 2.5% |
| 2025-08 | 44 | 1.0% |
September 2025 peak: 238 corrections during SHELET development.
Deep Dive Stats
Metrics that show depth, not just volume.
What This Shows
Intellectual Evolution
How language changed over time. As fluency increased, verbosity decreased.
"what if...", "can you help..."
"give me...", "show me..."
Minimal words, assumed context
Shorthand, shared vocabulary