public good survey to help people suffering from mental health issues
Been thinking about this a lot lately.
More and more people struggling with depression, loneliness, anxiety, suicidal thoughts etc are talking to AI instead of actual humans. And now we’re already seeing the first AI-related suicide lawsuits.
This is scary, so I started an open page where anyone can test or flag how different AI models respond to mental health prompts. https://humanjudge.com/auth?track=empowerment&project=7a97ef08-3d89-42c5-a4db-f6b43b4700ca
Not to attack AI companies. Just because vulnerable people are using these systems every day, and we need more accountability around what they’re being told when they’re at their lowest.
If an AI gives dangerous advice, people should know.
Would genuinely love help/input from people here who care about this stuff!