---
id: "concept-levels-of-ai-fluency"
type: "concept"
source_timestamps: ["01:31:00"]
tags: ["ai-adoption", "skill-development"]
related: ["concept-active-level-3", "framework-context-sandwich", "concept-playbooking-method", "prereq-basic-prompting"]
definition: "A three-tier model of AI adoption progressing from basic Q&A to collaborative work, and finally to autonomous delegation."
speakers: ["Rachel Woods"]
---
# Levels of AI Fluency

## The three levels

### Level 1 — AI for Answers
Treating AI like a super-powered search engine. Asking ChatGPT for recipes, basic facts, definitions. The intuitive beginner level. Most casual users plateau here.

### Level 2 — AI as a Daily Work Partner
Using AI to assist with ongoing work: drafting emails, brainstorming, coding side-by-side. The human is still heavily involved, directing the AI step by step. The [[framework-context-sandwich]] is the prompting structure that elevates a user *into* Level 2.

### Level 3 — AI Working *For* You
The AI operates autonomously or semi-autonomously in the background, executing full workflows based on triggers. Minimal human intervention per execution. Achieved via the [[concept-playbooking-method]].

## Why the taxonomy matters

The summit's argument is that the gap between Level 2 and Level 3 is where actual leverage lives — and that most learners get stuck at Level 1 or 2 because they think AI mastery means *better prompting* rather than *better operating*. See [[contrarian-stop-prompting]].

## Sub-distinction inside Level 3

Level 3 itself bifurcates into passive and active forms — see [[concept-active-level-3]].

## Prerequisite

The taxonomy presumes [[prereq-basic-prompting]] — basic chatbot familiarity.

## Enrichment context

Parallels Gartner's AI maturity model (assistive → autonomous) and Forrester's adoption stages. Level 3 maps closely to *agentic AI* in recent multi-agent workflow papers.