ChatGPT medical advice brought man 'to brink of death', lawsuit alleges | Hacker News Reader

ChatGPT medical advice brought man 'to brink of death', lawsuit alleges

bbc.com

23 pointsdijksterhuis22 comments

I really hope the lawsuit does not end in OpenAI tightening the “safety” and “security” of their models into oblivion. I hope they understand that a lobotomized model is as good as my Casio watch.

I want my models to give it to me straight without patronizing me. I’m an adult and I take responsibility for my own actions.

That's the issue here tho.

You do not own the model; they do. Ergo it is their advice and so they can be held accountable fo it.

so if you give your friend medical advice and they are injured, can they sue you?

I'm sure they could try.

I don’t think this is true if a disclaimer is provided. Most investment advice videos on YouTube make this very clear, but it doesn’t stop them from sharing advice. I suppose there is some precedent for this in the American justice system.

Why are people even allowed to dress up ChatGPT... an autoregressive transformer... as a faith-based holistic doctor? Anybody could see this going wrong from a hundred miles away.

A few days ago a family member insisted to be brought to the ER because Claude told him. Result: he was immediately hospitalised and received surgery the following day to remove a tumour that was compressing his spine. Claude was very clear that he had to act as quickly as possible to lower the chances of permanent damage.

The comments here are ridiculous. Do you know how much ChatGPT has helped families, esp. those without access to the top medical doctors and resources around the world? My own experience with it so far have been nothing but positive. It beat all the doctors in various hospitals that we have consulted so far on breadth and depth and the best thing is you can keep asking follow-up questions on anything and it gives you all the details you need without being rushed.

The fact that it is not 100% reliable shouldn't be basis for restrictions. Even top trained doctors are far from 100% reliable either.

All the chat bots have a very discreet disclaimer at the bottom of the page. I want to say it should be bigger text and more strongly worded but like most people I would just mentally tune it out regardless.

The LLM I use is fairly quick to retreat into extra disclaimers. "Not sure about this, this is from memory and you ought to check", etc. Don't they all do that? Is it just because of my prompt (which does say to push back if I appear to have bad ideas)?

It seems possible that people tune out disclaimers and caveats in the middle of the answer, too.

Ironically, the given name "Claude" means "crippled" or "lame". Students of the Roman Empire may recall the disabled Emperor Claudius.

This was with GPT-4o (ancient now by AI standards). The reason OpenAI refuses to bring it back, despite all the demand, is that when the conversation gets long enough it starts doing weird things like this.

Subsequent models have almost entirely eliminated this kind of thing.

Even GPT-5.4 is ancient compared to intelligence of 5.6 although those are only 4 months apart. We're talking about at least twice the capabilities and that's scary to achieve in such a short timeframe.

There are no bounds to human stupidity

The amount of times these LLMs randomly refuse to answer basic questions because "omg safety" only to turn around and do this f￭￭￭ing nonsense is obscene.

I'm not blaming the LLM because it's just a pile of math glued together with hubris and anxiety but why is its default response not "I am not a medical professional and you should see a doctor right away if you have any concerns"?

because they get sue otherwise?

this is number one bullshit. You ask a medical professional for medical advice. Getting medical advice from AI or anywhere else is your own fault.

Reply on news.ycombinator.com