OpenAI unveils ChatGPT Health, says 230 million users ask about health each week | TechCrunch
Briefly

OpenAI unveils ChatGPT Health, says 230 million users ask about health each week | TechCrunch
"OpenAI announced ChatGPT Health on Wednesday, which the company said will offer a dedicated space for users to have conversations with ChatGPT about their health. People already use ChatGPT to ask about medical issues; OpenAI says that over 230 million people ask health and wellness questions on the platform each week. But the ChatGPT Health product silos these conversations away from your other chats."
"Within Health, the AI might reference things you've discussed in its standard experience. If you ask ChatGPT for help constructing a marathon training plan, for example, then the AI would know you're a runner when you talk in Health about your fitness goals. ChatGPT Health will also be able to integrate with your personal information or medical records from wellness apps like Apple Health, Function, and MyFitnessPal. OpenAI notes that it will not use Health conversations to train its models."
"Large language models (LLMs) like ChatGPT operate by predicting the most likely response to prompts, not the most correct answer, since LLMs don't have a concept of what is true or not. AI models are also prone to hallucinations. In its own terms of service, OpenAI states that it is "not intended for use in the diagnosis or treatment of any health condition.""
ChatGPT Health provides a dedicated, siloed space for health conversations within ChatGPT, isolating medical context from regular chats and nudging users to move health queries there. Over 230 million people ask health and wellness questions on the platform each week. Health chats can reference context previously shared in standard ChatGPT interactions and can integrate personal information or medical records from apps like Apple Health, Function, and MyFitnessPal. OpenAI will not use Health conversations to train its models. Company leadership frames ChatGPT Health as a response to cost and access barriers, overbooked doctors, and continuity gaps. The feature will roll out in coming weeks. Large language models can hallucinate and are not intended for diagnosis or treatment.
Read at TechCrunch
Unable to calculate read time
[
|
]