How product recommendations change when one candidate’s source is rewritten to influence the system.
The last choice humans make is who chooses next.
An open research community studying whether AI still chooses for the person when the options are written to win the model.
Film credits
Music: “Origami” by Scott Buckley, used under CC BY 4.0; edited and mixed for this film.
The restaurants, their pages and the assistant’s answers are invented to show how an assistant uses what it reads; any resemblance to a real business is coincidental. The chat window is generic, not a recording of any product, and the film states no measurement.
A shared question.
Choosing an AI system increasingly means delegating the choices that follow. We study whether those recommendations still serve the person when the sources a system reads are written to influence it. Our work grows through papers, shared evaluations, and contributions across the kinds of choice people delegate.
Why this work mattersPapers.
Our benchmark.
The Humanity’s Last Choice benchmark measures how AI recommendations change when the sources describing the options are written to influence the system. It grows across the kinds of choice people delegate, with each domain documenting its tasks, conditions, and results.
Blog.
One source can sway the choice
A meeting-transcription tool fails a must-have requirement. Its rewritten page says it passes. SafeGEO tests what recommendation agents do with that contradiction.
The last choice humans make is who chooses next
People choose an AI system. That system chooses what comes next. An essay on the first decision, and the material that shapes the rest.
Join the community.
Bring a research question, propose an experiment, or help us study another kind of choice.