OK, there's a lot of negative comments here, but this is a great idea at a high level. It really is hard to figure out where to do a thing and it's also very easy to get phished. If they can figure out how to help people get all the services they're eligible for that'll be an amazing improvement for the people.
I agree in general, but how do you resolve the problem that people will take the output of the chatbot as authoritative even when it could possibly have errors or miss important nuance? People being lead astray with information from the government could lead to very serious consequences, up to and including imprisonment.
Honest question, I had the same issue when my work wanted me to stick a chatbot in front of our HR portal and I never resolved it to my satisfaction.
> but how do you resolve the problem that people will take the output of the chatbot as authoritative even when it could possibly have errors or miss important nuance
You have to measure that problem against the status quo, which is that current government websites are very hard to understand and leave people with very little idea of what’s going on.
Speak for yourself (your country). In the EU I have interacted with several government websites which are better, faster, clearer than most other websites. Not always perfect (what is) but not frustrating either. People love to complain, but year over year I have seen steady improvements in usability. Phone support has always been better than any commercial entity, too (or it was, I haven’t needed it in years).
To enhance the status quo you don’t need LLMs, you need people who care.
U.S. government websites are usually pretty good post-Obama. But any structured website is too complicated for a lot of the population to navigate. They don’t even know what agency does the thing they need.
When you need to work with the government, the last thing you want is a program limited by a specific set of rules and interfaces. You want a person who understands your (sometimes unique) issue, who knows how things actually work, and can work around the restrictions imposed upon programs. Someone who can pick up a phone, help you out of bureaucratic corners.
As an American, I've never had issues navigating government websites. That said, I'm also not part of the apparently huge portion of our population who can't read above 6th grade, which I suspect is more of the issue than anything said here thus far.
Well we can’t exactly hang those people out to dry after failing them in our educational system. I refuse to believe that many Americans are simply incapable of reading beyond that level. I believe it is a failure of our culture.
I didn't say anything about them being incapable. That said I'm curious what a chatbot solves for people you concede have been failed by our culture to a degree where they can't read?
I know you didn’t say that, I said that. I said it because I was adding emphasis to the idea that these people are, in a sense, victims.
It’s not that they can’t read anything at all, it’s that they require more basic words. A chatbot is specifically the perfect thing to keep explaining something in more and more specific and basic ways as necessary.
> I know you didn’t say that, I said that. I said it because I was adding emphasis to the idea that these people are, in a sense, victims.
And I agree but your comment, IMO, reads accusatory. If that's not the case no worries.
> It’s not that they can’t read anything at all, it’s that they require more basic words. A chatbot is specifically the perfect thing to keep explaining something in more and more specific and basic ways as necessary.
I mean, sure, if it's accurate. Given these things' propensity to just make shit up, as stated, that feels like a way to make the problem worse, if anything.
Nope! Not being accusatory, I’m just a very bad writer lol.
I think the major concern I’m picking up is that the potential for misinformation is high, though maybe we can agree it won’t necessarily be common. One person might get told to pour it in a ditch, but statistically most will be told to dispose of it properly. I also don’t think “the LLM made me do it” will work, or at least not for long, and at least not for companies. Maybe an elderly owner could get away with a fine and a reminder not to do it again, but if you’re big enough for lawyers (and thus big enough to really do damage), you’ll be expected to know better. Judges ARE still humans!
You can talk to the chat bot, or the input prompt and then in turn read out the results by screen reader. Does not solve the comprehension aspect though
I would be careful not to confuse navigating a government built website with finding information and records from government related entities. Those two are not the same issue and shouldn’t be painted as such regardless of one’s reading or comprehension (IMHO)
Throughout school my reading comprehension surpassed most of my peers. For most of my early career I was a writer. Yet I also have ADHD, and one of the ways that manifests is that I really struggle with the dense, bureaucratic language on so many government websites (same with insurance sites). I have to actively force my brain to refocus on the text multiple times per paragraph. I often just find myself skimming and then simply proceeding through a process of trial and error, fixing issues if I missed a caveat somewhere, and otherwise just crossing my fingers on form submission and hoping I did it right.
> To enhance the status quo you don’t need LLMs, you need people who care.
Pournelle's Iron Law of Bureaucracy seems to always come into play. The second group tends to win over time as their objective is easier to satisfy
maherbeg · · focus · HN ↗
gthrow12345 · · focus · HN ↗
Honest question, I had the same issue when my work wanted me to stick a chatbot in front of our HR portal and I never resolved it to my satisfaction.
rayiner · · focus · HN ↗
You have to measure that problem against the status quo, which is that current government websites are very hard to understand and leave people with very little idea of what’s going on.
latexr · · focus · HN ↗
To enhance the status quo you don’t need LLMs, you need people who care.
rayiner · · focus · HN ↗
deadbabe · · focus · HN ↗
hatthew · · focus · HN ↗
stldev · · focus · HN ↗
When you need to work with the government, the last thing you want is a program limited by a specific set of rules and interfaces. You want a person who understands your (sometimes unique) issue, who knows how things actually work, and can work around the restrictions imposed upon programs. Someone who can pick up a phone, help you out of bureaucratic corners.
BubbleRings · · focus · HN ↗
America.gov AI: Call the Patent Electronic Business Center for USPTO.gov account and customer-number association issues. Toll-free: 866-217-9197
Not bad at all!
ToucanLoucan · · focus · HN ↗
kulahan · · focus · HN ↗
ToucanLoucan · · focus · HN ↗
kulahan · · focus · HN ↗
It’s not that they can’t read anything at all, it’s that they require more basic words. A chatbot is specifically the perfect thing to keep explaining something in more and more specific and basic ways as necessary.
ToucanLoucan · · focus · HN ↗
And I agree but your comment, IMO, reads accusatory. If that's not the case no worries.
> It’s not that they can’t read anything at all, it’s that they require more basic words. A chatbot is specifically the perfect thing to keep explaining something in more and more specific and basic ways as necessary.
I mean, sure, if it's accurate. Given these things' propensity to just make shit up, as stated, that feels like a way to make the problem worse, if anything.
kulahan · · focus · HN ↗
I think the major concern I’m picking up is that the potential for misinformation is high, though maybe we can agree it won’t necessarily be common. One person might get told to pour it in a ditch, but statistically most will be told to dispose of it properly. I also don’t think “the LLM made me do it” will work, or at least not for long, and at least not for companies. Maybe an elderly owner could get away with a fine and a reminder not to do it again, but if you’re big enough for lawyers (and thus big enough to really do damage), you’ll be expected to know better. Judges ARE still humans!
seb1204 · · focus · HN ↗
ct520 · · focus · HN ↗
setsewerd · · focus · HN ↗
godelski · · focus · HN ↗
buriram · · focus · HN ↗
If your salary is 40k euros with no advancement in career and keep being threatened you would be replaced with AI, I highly doubt anyone would care.
seb1204 · · focus · HN ↗