01 · YOUR DISCOVERIES
Follow your curiosity.
Use the task manager however you like. When something seems wrong, file a quick note in your own words.
TEN MINUTES. TWO PERSPECTIVES.
See what you
and AI
find together.
Give it ten minutes. Explore a buggy app, report what you notice, and discover where your findings overlap with AI—and where your perspective adds something different.
Find my overlapOne free session per email and browser. Your comparison to take away.
DIFFERENT EYES. A FULLER PICTURE.
What do we both notice?
What does each perspective add?
02 · YOUR COMPARISON
See your overlap.
Get a visual comparison of shared findings, observations not matched by the AI collection, and the types of issues you explored.
03 · OUR UNDERSTANDING
Help testing move forward.
Your session contributes to a growing study of how people and AI can complement each other when testing software.
ONE SMALL SESSION. A BIGGER QUESTION.
How can humans and machines
do better work together?
Every participant brings a different way of looking at software. By comparing those observations with AI findings, we can explore what gets noticed, what gets missed, and where combining perspectives adds value.
We plan to share aggregate findings so the testing community can learn from the results. Your ten minutes helps build that understanding.
READY TO EXPLORE?
Ten minutes.
Your perspective matters.
- Start a session. The timer begins when you press the button.
- Find and file. Explore the app and save short reports as you go.
- Get your comparison. See the overlap, explore differences, and download your session.
AI matching is a starting point for analysis. An unmatched observation is a candidate for review, not automatically a confirmed new bug. The reference collection combines multiple ten-minute AI runs.
THE APP UNDER TESTIntentional bugs included
Time is up.
Your reports are saved locally. Your comparison is below.
TEN MINUTES, DIFFERENT PERSPECTIVES
Your testing, in context.
Your free attempt has been used. Your results remain available here.
How this comparison works
We compare your reports with the deduplicated findings from two organizer-observed, equal-effort 10-minute AI coding-agent sessions on the same Buggy build. The collection is broader than a single machine run. Matching is AI-assisted and can be wrong; it does not reproduce your findings. Preferences, questions, vague reports, and duplicate reports are kept separate. Unmatched candidates are not automatically new confirmed bugs.
The timer is issued by the server. Reports are cached in your browser, including during connectivity loss, and synced during and after the session. Full-session completion requires ten minutes of observed browser time; leaving early or a long interruption is marked partial. This is a self-reported workshop exercise, not an identity-verified examination. One free session is allowed per email address and per browser. Partial sessions also use that attempt. Refreshing resumes the same session.
The $25/hour human rate is a conservative assumption requested by the organizer. Ten minutes costs $4.17 under that assumption. Baseline AI run costs were not recorded. We do not substitute the small cost of matching your reports for the cost of the AI discovery runs.