---
title: "Can your AI survive one mistake?"
code: JZ-006
series: Behavior
published: 2026-11
updated: 2026-11-01
author: Joost van der Zwan
canonical: https://softwarebehavior.com/notes/jz-006
skill: studio-jz-algorithm-miss
license: CC BY 4.0
---

# Can your AI survive one mistake?

Field Note JZ-006 · Behavior · One mistake

**15–29%** more error, trusting themselves

## The guess
Better tools get forgiven: a model that is right more often keeps its users.

## What the research found
In forecasting tests with a bonus for accuracy, people who watched a statistical model make mistakes chose their own judgment instead, though the model was better: their own forecasts had 15–29% more error. A follow-up study found the way back: when people could adjust the model's forecast, even slightly, they were far more willing to use it.

## Why
The same mistake cost a model more trust than it cost a person. A small say in the result was enough to bring people back.

## What to do
- Let people adjust what the system gives them: edit a suggestion, nudge a forecast, re-rank results. Even a small degree of control helps.
- Design the moment the system is wrong before it happens: acknowledge it, make correction easy, keep the person in charge.
- Set honest expectations about accuracy up front.

## What not to do
- Make the output take-it-or-leave-it.
- Promise a model that never errs. One visible mistake costs more than a modest claim.

## Limits
- Lab forecasting tasks with a bonus at stake, not everyday products.
- The studies measured whether people chose the model, not how they used it over months.

## Example
- Before: A sales forecast shown as one fixed number; after one bad month the team stops opening it.
- After: The forecast with a small adjustment control (for example ±10%) and a note on how often it has been within range (illustrative).

## Sources
- Dietvorst, B. J., Simmons, J. P. & Massey, C. (2015). Algorithm aversion: People erroneously avoid algorithms after seeing them err. Journal of Experimental Psychology: General, 144(1), 114–126. https://doi.org/10.1037/xge0000033
- Dietvorst, B. J., Simmons, J. P. & Massey, C. (2018). Overcoming algorithm aversion: People will use imperfect algorithms if they can (even slightly) modify them. Management Science, 64(3), 1155–1170. https://doi.org/10.1287/mnsc.2016.2643

## Use with your AI
This note is also a skill: `studio-jz-algorithm-miss` (version 1.0.0). Download: https://softwarebehavior.com/skills/studio-jz-algorithm-miss.zip · All skills: https://softwarebehavior.com/field-kit

Software Behavior Field Notes by Joost van der Zwan · Studio JZ · CC BY 4.0 · softwarebehavior.com