What happens now if someone uses traditional search and comes across an outdated wiki or Confluence page with outdated/wrong/bad instructions, or interprets something incorrectly?
I’m responding to your point that it would be “unacceptable” for a regular government website to provide wrong information. The case I linked to says that the government isn’t responsible for mistakes in what the government tells you about the law. If someone in the social security office says the law is X, but it’s actually Y, then the citizen is still required to follow Y.
The case isn’t about whether it’s “illegal” for the government website to be wrong. It says you as the citizen are held responsible for the actual law, even if you followed incorrect government advice. So the current rule already acknowledges that advice provided by the government may be mistaken and says that’s no excuse.
So the question is is the chat bot better than people trying to understand complicated government websites by themselves?
It has always been the person's problem to solve, even after this website.
No government employee is your lawyer. If the government gives you incomplete or wrong information, you are still liable for not doing the correct thing.
Having a website that does a best guess at what steps a person must take after changing a name or getting married is better than nothing, but the status quo isn't "nothing".
It needs to be kept up-to-the-minute updated and it can't skip any nuance / details. Does anyone here really trust the lackeys who spectacularly failed at DOGE to do boring and highly detailed work?
My team experimented with this a couple of years ago with a complex government regulatory body of knowledge.
Google and Microsoft have APIs that will ground with and reference the specific policy documents, which helps. Even then, it was more useful as a job aid for people already knowledgeable of the data than for a layman.
With all LLMs, the quality of the question correlates with the quality of the answer.
