In short

1. The Core Relationship

The graph shows a clear negative correlation (inverse relationship) between human Competence (x-axis) and AI reliance/usage (y-axis).

• Low Competence, High AI: On the far left, where a person’s skill or knowledge level approaches zero, their dependency on AI spikes dramatically toward the vertical asymptote. If you don’t know how to do something at all, AI becomes your absolute crutch.

• High Competence, Low AI: On the far right, as human expertise grows, the curve flattens out, approaching a horizontal asymptote. The master needs the AI far less than the novice does.

2. The Two Truths

Depending on how you look at it, this diagram carries two distinct truths:

• The Practical Truth (Efficiency): Experts don’t need to prompt an AI for things they can do in seconds with their own muscle memory or deep knowledge. They only use it for mundane, repetitive tasks or broad brainstorming. Novices, on the other hand, use it to bridge a massive skill gap.

• The Critical Truth (The Dunning-Kruger AI Effect): There is a darker psychological layer here. People with the lowest competence often use AI blindly because they don’t know enough to realise when the AI is hallucinating or giving poor advice. Meanwhile, true experts are often skeptical, using AI sparingly because they understand its limitations and biases.

3. The Irony

The tragedy of this graph is that AI is often most dangerous to the people who use it the most. When Competence is near zero, the user lacks the critical thinking required to verify the AI’s output. Ideally, we (or is it you?) want a curve where high competence pairs with strategic AI leverage to achieve superhuman productivity, rather than AI just acting as a substitute for basic (human) skills.

Conclusion

It’s probably a great doodle that perfectly captures the “shortcut culture” vs. “expert mastery” tension of our times.

Did I draw this to highlight how people are skipping the learning process?