The collision is happening in real exam rooms with zero guidance for either side: - Patients using ChatGPT for diagnosis with no framework for safety - Physicians encountering AI-informed patients with no training
Dual curriculum: - Patient pathway: sensing gap, hallucination detection, appropriate use - Physician pathway: integration scripts, documentation, liability
Technical choices: - WordPress (pragmatic, not exciting) - Focus on content over platform - All resources downloadable - Zero tracking beyond basic analytics
Built free because the problem is urgent. Monetization kills distribution speed.
Feedback wanted especially on: - Technical accuracy of AI limitations explanations - Accessibility for non-technical users - What's missing
The "velociraptor test" and "intelligent humility" modules might interest HN crowd - applying evolutionary biology and architectural constraints to AI safety.