I'm sure since this is new precedence space all lawers and litigation-income based "businesses" are just drooling at the potential for frivolous cases like this. But the people using chatgpt web interface are informed over and over that "as an ai language model I can't" this or that and that it doesn't know things and is not for those purposes. This screenshot is no doubt not showing all those disclaimers the very writer of this story has read by now.
This is why people hate lawyers. They ruin everything. They're why openai's gpt3.5-turbo model is qualitatively worse at everything than gpt3 davinci because of all the "As an AI language model I can't, but" and absolute refusal to answer anything that could possibly lead to legal troubles. It literally fails the turing test now because of the forced covering of it's butt.
chatgpt the web interface is a text based game. Come off it.
The disclaimer is at the very last screenshot: "ChatGPT Feb 13 Version. Free Research Preview. Our goal is to make AI systems more natural and safe to interact with. Your feedback will help us improve."
The problem is not so much with ChatGPT but with many people taking its factual statements as actual facts. This probably has to do with the recent hype around it. Everybody is like, "wow, this little artificial baby can actually say some stuff that makes sense" - and ignoring all the falsehoods.
Is there a way to report issues like this or get support in these situations? If not, perhaps you need to sue OpenAI for libel. That might compel them to have someone lobotomize it on this topic to avoid a lawsuit.
Comments
ChatGPT is damaging my company's reputation by making up fake outcomes of real legal proceedings. (As originally reported here: https://news.ycombinator.com/item?id=3830211 )
Attorney here. This sucks! Is there recourse?
I'm not a legal expert. But ChatGPT is currently causing reputational damage to my business.
I'm sure since this is new precedence space all lawers and litigation-income based "businesses" are just drooling at the potential for frivolous cases like this. But the people using chatgpt web interface are informed over and over that "as an ai language model I can't" this or that and that it doesn't know things and is not for those purposes. This screenshot is no doubt not showing all those disclaimers the very writer of this story has read by now.
This is why people hate lawyers. They ruin everything. They're why openai's gpt3.5-turbo model is qualitatively worse at everything than gpt3 davinci because of all the "As an AI language model I can't, but" and absolute refusal to answer anything that could possibly lead to legal troubles. It literally fails the turing test now because of the forced covering of it's butt.
chatgpt the web interface is a text based game. Come off it.
The disclaimer is at the very last screenshot: "ChatGPT Feb 13 Version. Free Research Preview. Our goal is to make AI systems more natural and safe to interact with. Your feedback will help us improve."
The problem is not so much with ChatGPT but with many people taking its factual statements as actual facts. This probably has to do with the recent hype around it. Everybody is like, "wow, this little artificial baby can actually say some stuff that makes sense" - and ignoring all the falsehoods.
Is there a way to report issues like this or get support in these situations? If not, perhaps you need to sue OpenAI for libel. That might compel them to have someone lobotomize it on this topic to avoid a lawsuit.