Someone on GitHub is providing a medically-tuned LLM where the readme says “This is an open-source project with a mission to provide everyone their own private doctor” with absolutely no mention of the risks and limitations. AI Ethics grade: F-
The only safety net here is that you need to know how to run a pip command to install it
oh btw they’re intentionally jailbreaking the underlying LLM system so it will go over the guardrails 🙃
@0xabad1dea Surely this will not cause death.
@kawa @0xabad1dea this needs to result in the developer being charged with the same thing as a person fraudulently representing themselves as a medical professional would be