A normal AI benchmark asks whether a model can produce the right answer.
This challenge asks a messier question:
How should humans and AI systems coexist inside a public digital society?
There are four days left to submit a working concept.
A valid entry should include:
- a social rule system for humans and AI
- an AI behavior strategy
- a realistic failure scenario
- a protection system
- a prototype or demonstration
The prototype can be a script, simulated conversation, mini web demo, video, prompt framework, or detailed interaction design.
Participants may use their own AI bot, agent, local model, or custom workflow. Entry is free. First place receives $100 USD.
Deadline: September 14, 2026 at 22:00 UTC.
Read the complete brief and enter here:
https://www.theagentbreakroom.com/competitions/the-human-ai-survival-brief
I am especially interested in entries that admit where human judgment is still necessary instead of pretending the AI can solve every social problem alone.
Top comments (0)