Medical chatbot using OpenAI’s GPT-3 told a fake patient to kill themselves

Re:Medical chatbot using OpenAI’s GPT-3 told a fake patient ... by visarga on Saturday February 27, 2021 @01:25AM (#26283120)

> I know what I want and how to tell it to a machine.

Yeah, Google search is not that great when you want to do conditional search in a topic with many false positives. For example I want to find a light electric scooter, under 10Kg of weight. Google will happily report all the pages that contain scooter and kg, but the kg would be for the max weight of the person, not the scooter itself. How do I tell it that in keywordese?

[ Parent | on HN ]

Re:Medical chatbot using OpenAI’s GPT-3 told a fake patient ... by colanderman on Saturday February 27, 2021 @09:43AM (#26285137)

Actually I feel like Google is the one such service that is approaching natural language usability for general topic queries. It seems to me to be able to extract basic meaning from simple English phrases and search that sentiment. Whereas search phrases containing negatives ("not", "isn't") used to be futile, Google now correctly infers your meaning.

But yes, for parametric search I think I would always prefer a direct interface. It makes clear what is indexed and what isn't.

[ Parent | on HN ]

Re:Medical chatbot using OpenAI’s GPT-3 told a fake patient ... by sitkack on Saturday February 27, 2021 @02:06AM (#26283272)

Altavista had the "near" operator.

[ Parent | on HN ]

Re:Medical chatbot using OpenAI’s GPT-3 told a fake patient ... by gundmc on Saturday February 27, 2021 @03:56AM (#26283599)

Google has the "AROUND" operator, though I haven't tried it recently to see if it is still respected.

[ Parent | on HN ]