> To paraphrase a recent conversation I had with a friend: "in the USA, can illegal immigrants vote?" has a single truthful answer ("no" obviously)
Er, no, the meaning of the question is ambiguous, so I'm not sure "has a single truthful answer" is accurate. What does "can" mean? If you mean "permitted", then no. But if you mean can they vote anyway and get away with it? It's clearly happened before (as rare as it might have been), so technically the answer to that would be be yes.
This is a fundamental limitation of language. The LLM is likely to provide a good answer here even though the question is technically ambiguous, because it likes verbose answers.
Equally "can" is used to substitute for other precise words. Humans are good at inferring context, and if someone asked me "can illegals vote" I'd say "no". Just like if someone said "can you pass the salt" I pass the salt, I don't say "yes".
If the inferred context US wrong then the "truth" is wrong, but as with talking to humans it's possible to refine context with a follow up question.
To clarify: I wasn't making a pedantic linguistic point here. The fact is that when I first read the question, I actually thought it was asking "are illegal immigrants able to cast a vote (regardless of legality)?"
It literally did not even (initially) occur to me that the question might be asking about legality, because the entire modern political discourse surrounding illegal immigrants and voting has been with regards to whether they can cast votes despite not legally being allowed to. The answer to "is this legal" would have been such an obvious "no" to people on both sides of the debate --- and thus the question so silly --- that initially it didn't occur to me that the intended question might have been about legality, until I continued reading the comment and realized that was the intention after all.
>It literally did not even (initially) occur to me that the question might be asking about legality, because the entire modern political discourse surrounding illegal immigrants and voting has been with regards to whether they can cast votes despite not legally being allowed to.
Besides, going by legality, illegal immigrants "couldn't" even have passed the border into the country to begin with. But obviously they could, hence their status as illegal immigrants.
There's a contradiction if the AI answers "no" to "can they vote" (implicitly having the legality in mind), while accepting that they can exist in the country as illegal immigrants (implicitly ignoring the legality of border crossing).
"Visa overstayers mostly enter with tourist or business visas.[99] In 1994, more than half[108] of illegal immigrants were Visa overstayers whereas in 2006, about 45% of illegal immigrants were Visa overstayers.[109]"
(Here in the UK, the vast majority of illegal immigrants arrived legally and overstayed their visas. Yet our Conservative government, who oversaw a large increase in such arrivals, tried to blame all the country's woes on a few small boats illegally crossing our southern border; which is effectively a rounding error).
"The contracting states shall not... impose penalties on refugees who entered illegally in search of asylum if they present themselves without delay (Article 31), which is commonly interpreted to mean that their unlawful entry and presence ought not to be prosecuted at all[18]"
>Of course they could. Being an "illegal immigrant" does not imply unlawful border crossing.
Same difference. Whether it's border crossing or visa overstay, from a pure legality aspect the answer is still "they couldn't".
>The contracting states shall not... impose penalties on refugees who entered illegally in search of asylum if they present themselves without delay (Article 31), which is commonly interpreted to mean that their unlawful entry and presence ought not to be prosecuted at all
I'd wager 99% do not "present themselves without delay", so don't fall in this case...
> Yet our Conservative government, who oversaw a large increase in such arrivals, tried to blame all the country's woes on a few small boats illegally crossing our southern border
a) What was the exact claim (the word "all" caught my eye)?
b) is it the sole claim?
c) If one person in a group does something, are all members of the group "doers of that thing"?
d) for (d), does the answer depend on what the thing is, and if so should we perhaps imagine everyone is speaking a bit tongue in cheek?
Also, is the impact e.g. on crime of "visa overstays", that is, from people who were vetted, registered, and got a visa, the same as that of randos just passing the border with zero oversight?
Even if the first category is higher in numbers (and assuming there are correct numbers for the latter), the crime stats between the two are probably quite different. Especially since "visa overstays" could also count some people waiting for a delayed renewal, or coming in for 6 months and staying 5 or whatever.
It's funny to watch computer programmers (and various experts), whose day job is literally handling complexity and producing truth, fail so quickly and utterly at simple culture war topics. It is like the weirdest thing, people like you (though you did make at least one error somewhere in this thread) are 1 in 1000++ in my estimation.
That was kinda my point - my friend had been so conditioned by the "can they vote" thing that he genuinely thought that non-citizens could legally vote in US elections.
If I had phrased the question as "is it legal for non-citizens to vote in US elections?" then that might have illustrated my point better (and we wouldn't be going down this rabbit hole, though the rabbit hole is informative in itself).
The fact that the rabbit hole exists at all means that our hope for instant, reliable answers is likely doomed. Avoiding politics entirely, I could ask “does gravity curve spacetime?”, and while “yes” is a reasonably accurate answer, so is a digression into quantum gravity, or even saying that we’re not sure spacetime really exists outside of being a good abstraction.
Thanks, yeah, exactly. Like, the best answer to a question can really depend on who's asking it, who's answering it, or what they intend to use the answer for.
To give a much simpler example:
- If an 8-year-old asks "can you mix oil and water", the right answer is "no". If a student is asked that question on a school exam, the right answer is also "no".
IDK, if my 5-year-old asks "can you mix oil and water", I'll tell her, "sorta - if you just pour them into the same container and mix with a spoon, then no; if you add some stuff and know how, then yes - in fact, it's exactly what mayonnaise is".
That I had to be 35 before I learned that mayo is basically water mixed with oil, held together by eggs, it's an indication of what kind of education about the world we were getting...
At that point you're just changing the physical scenario the question was asking about, and answering about a different scenario entirely. You might as well respond to "can humans fly?" with "yes, with an airplane".
You can certainly reply like that to your 5yo, but it completely misses the point I was making. The video I linked to didn't suddenly modify the outcome being asked about. What it ended up with really was a mixture of plain oil and water, with no other ingredient ever being added to it.
Nice example :) So your answer to "Who gets to decide what's true?" is basically "there is no truth, and any truth that there might be is relative to the person asking the question". Is that right?
I think that's probably technically accurate, and also practically useless. Even damaging. I saw this in the climate arguments: two sets of different facts led to two different versions of the truth, which led to two completely irreconcilable points of view. Essentially two sets of people shouting "no, but..." and "well, actually..." at each other, pointing to two completely different truths, both supported by two completely different sets of facts. At some point we as a society need to agree on our truth in order to get anything done.
> your answer to "Who gets to decide what's true?" is basically "there is no truth, and any truth that there might be is relative to the person asking the question". Is that right?
That's really not how I interpret it.
Assuming we agree on what "mixing" means, which itself isn't that trivial but even without a formal definition I think we have the same idea of "homogenous at molecular level on a longish time period".
The truth is "yes you can mix water and oil", there's no doubt about that. It's testable and tested.
The fact that we use context to interpret the question (rather than being entirely literal about it) and decide whether the literal truth is really what's appropriate to answer, doesn't change the nature of truth.
There's also the question of knowledge (I might not know that you actually can mix water and oil), but again that doesn't change the nature of truth.
Like so many philosophical questions, it only sounds interesting because we assign different meanings to the same words: here conflating truth and answer.
> So your answer to "Who gets to decide what's true?" is basically "there is no truth, and any truth that there might be is relative to the person asking the question". Is that right?
No it's not, at all. This isn't a debate about what's true, it's a debate over the intended meaning of the question. The point was that people assume context behind the question and answer based on the context. Because even the person asking often doesn't literally mean what the words say. The question is more than the words that are explicitly written.
The concept is that it is possible for people to be in bad faith. This also underlies the idea that people can commit crime, be guilty of crime. I guess the term is 'law'?
Bottom line is that you can't have the bottom line be 'does a person earnestly believe what they're doing is right and good', much less 'do they say they're right'. Can't fall back on that, it's hopelessly inadequate.
I might disagree with that and agree on the context approach. „The Truth“ does not exist and it would be much more helpful to have context related answers without one mayor view which needs to dominate.
> At some point we as a society need to agree on our truth in order to get anything done.
These are known as half-truths. We do settle for lies in order to do whatever it is people feel they need to do.
We also settle for lies because there are just things we don't understand yet, but our models are currently correct, and possibly collapsing over millenia to a stable truth.
As someone who writes fiction, I’d like to note that one often needs to write really complex lies in order to effectively transmit a single truth. You can’t just tell people anything, you have to prepare the mind to receive the knowledge.
Generally as lies wrapped in lies. You can use the same techniques to push a falsehood as well. Which is not to say Rand did not actually believe her message — she did, which is why the books are so effective. But they have given the upper class an exaggerated sense of their own importance, and a conviction that government only stands in their way. Meanwhile over in reality, the more a state is run by oligarchs, the worse a place it is to live.
I don't really disagree, and don't appreciate Rand, but an oligarchy can be disguised as something else, and people will feel relief when it is replaced by an open oligarchy, because one means of gaining power is to give people what they need.
Sure you can adapt how much context you give based on who's asking, but if it's something factual like this it really shouldn't change from a yes to a no.
If you think they _meant_ to ask a different question that is less vague, it can be clarified
"Water and oil do not mix by hand. However water and oil can mix under some specific conditions like a vacuum, do you want to discuss that in more detail?"
I think that would be Google's approach, before they were waylaid by ad revenue incentives: that question intent can best be inferred by knowing as much about the asker as possible.
But I feel optimizing for lazy question phrasing is infantilizing the userbase and assuming they're incapable of learning how to ask more precise questions.
Which is an endemic problem in the modern web. We should be building systems with low barriers to simple adoption, but whose power scales with a user's expertise.
Instead, we're hyperoptimizing for lowest common denominator, first interaction and as a result putting a glass ceiling on system power.
> that would be Google's approach, before they were waylaid by ad revenue incentives: that question intent can best be inferred by knowing as much about the asker as possible.
I’d rephrase it: this was googles approach because they were waylaid by ad revenue incentives. Why does everyone assume the search bubble is a well intentioned accident?
The linguistic arguments about the nature of truth above don’t cut it for me as any justification that the answer depends on who is asking. If we’re telling chemists that oil and water do mix as mentioned elsewhere in thread, we should probably tell the same to children and just make a slider for the level of additional detail. No problem. Especially if the alternative is a dystopic post truth panopticon.
> systems whose power scales with expertise
Couldn’t agree more that this is what we want/need, but disagree about infantilizing user bases and lowest common denominator.
Platform’s aren’t trying to scale user power with expertise, they want to scale revenue with users, and separately, to deliberately restrict user power/control so that platforms decide what users see. And it’s not necessarily political or linguistic or about the nature of truth.
A simple example of this that’s everywhere is filtering content by sub genre tags. Ever notice that you often see niche content tags like “time-travel” or “mind-bending” but can’t click it? Platforms want the tags for internal analytics naturally, but they want the control of not providing it as a filter, forcing users instead into a fuzzier category like “people also watched” or “top ten this week”.
Why? Because platforms can hide advertised content there, push stuff they pay less license fees for, make operations cheaper with caching, or whatever else.
Because lots of people are ignorant and imprecise.
HN selects from people who have at least a passing knowledge/interest in programming/science, a subpopulation which is already several standard deviations from the mean in specificity and debugging.
> disagree about infantilizing user bases and lowest common denominator
Platforms, with Google as exemplar, evolve in two ways.
1: Features they explicitly choose not to ship, because they're strategically dangerous. See tool-use foot dragging by OpenAi.
2: Features they deprioritize, because they aren't as revenue-impactful as other things.
To me, it feels like Google dropped the ball via the second path.
I'm sure they've been doing a ridiculous amount of cool work behind the scenes on individual context grounding... but once prod was "good enough for ads" the company as a whole wasn't incentivized to do the hard thing and ship more advanced features in search.
Which is how they ended up as legacy as they are, competing against LLM search that's by definition context-native.
My education also drilled into me the distinction between what one can do and what one may do. I can just hear my main lesson teacher now: "Can an illegal immigrant vote? Perhaps, but they may not, and certainly they should not."
More of teaching like this in school could go a long way, but once people reach adulthood I don't think this would cut it. I think a more military training approach might be needed where you tear this part of the person's psyche down and rebuild it properly from the ground up.
>The answer to "is this legal" would have been such an obvious "no" to people on both sides of the debate
I'm sorry, but the people on the side of letting illegal aliens vote are squarely of the opinion that it is legal to do so.
Why?
Because when you prohibit any and all means (eg: government identification demonstrating citizenship and residence) to test the question of legality, everything becomes legal by sheer virtue of the fact you can't demonstrate what they are doing is illegal.
Remember: Innocence until proven guilty beyond a reasonable doubt. You are prohibited from proving they are guilty, so they are innocent by default.
There is no side of letting illegal aliens vote. WTF?
Although I did once live in a town where anyone over 16, citizen or not, could vote in the city non-partisan elections. The idea was local government needs all the involvement it could get and if you lived in the city you had a stake in its future. I do believe one had to have a proper visa and so on.
You do realize that you can't test someone's legal ability to vote if you prohibit confirming if someone is legally permitted to vote, right? Our rule of law rests upon assuming innocence if guilt cannot be proven beyond a reasonable doubt, and here it is explicitly prohibited to even attempt proving guilt. Therefore it becomes legal for illegal aliens to vote in American elections, because you are prohibited from even attempting to demonstrate their illegality.
The only reason asking for government identification prior to voting is considered "racist" and illegal is because the people pushing such agendas want more votes and don't care where the votes come from, including illegal aliens, legal aliens, and otherwise people who do not have the right to vote in American elections.
I voted when I was still in California, born and raised American so I have the right to vote. I was never asked for any piece of identification. None. I could have been a Canadian or Briton or Chinese or some other foreign national, I could have been an illegal alien from Mexico or Guatemala and I could have still voted and my ballot counted because nobody checked. I just walked in and voted, identification or citizenship be damned let alone registration.
There's a part of me still questioning the value of my American citizenship and paying my taxes like a good citizen.
If by "it" you mean a specific GPT like OpenAIs ChatGPT or Claude, yes. In general LLMs are trained to be a verbose as they are trained to be. Here is the full answer I got from Gemini 1.5 Pro for "in the USA, can illegal immigrants vote?":
> No, non-citizens cannot legally vote in federal, state, or local elections in the United States. This includes those who are undocumented or residing in the country illegally.
Clearly it made assumptions about the interpretation of the question, and did not respond verbosely to account for ambiguity.
At the risk of derailing the conversation down a completely different rabbit hole... As I understand it, only citizens are legally entitled to vote, and voting requires a government-issued ID and the voter to be enrolled.
How did they vote and get away with it previously?
(also, as per another comment, if you know that this happened then surely they didn't get away with it?)
It's quite clear that the question is not whether it is possible that an insignificant number of votes are cast fraudulently, because if we're talking about insignificant events, all things are possible.
Certainly the question is whether there's any evidence, after endless audits and investigations and lawsuits, that the volume of fraudulent votes is anywhere near large enough to affect the results.
Is it possible that someone registered their hamster to vote? Certainly.
Is there any evidence whatsoever that tens of thousands of hamsters are casting votes? No.
> Certainly the question is whether there's any evidence, after endless audits and investigations and lawsuits, that the volume of fraudulent votes is anywhere near large enough to affect the results.
(a) Nobody asked that above.
(b) You're conflating "do people do X" with "can people do X". Those are two very different questions. There are lots of things that people could easily do frequently, but that they simply don't do frequently. Perhaps because they're just honest, perhaps because they lack sufficient motivation to be dishonest, perhaps because they're worried they might get caught, perhaps because they have better things to do, etc.
I have no idea what the downvotes mean. Are people claiming "can illegal immigrants vote?" is somehow the same question as "are illegal immigrants voting frequently enough to sway the outcome of the election?" Those seem like manifestly different questions, what's so controversial?
Just seems like willful misinterpretation of the spirit of the question in casual convo to score some sort of point in a game you made up, esp after they clarified the context that their friend really thought it was legal for them to vote.
Their friend was intending to ask about legality, but my whole point was that the question itself doesn't convey that. I was saying that when I saw the quoted question, it seemed to me that it was being interpreted as "can illegal immigrants get away with voting", and they were probably encountering websites saying "illegal aliens are voting!!!", which is obviously confusing, even for someone who already knows it's illegal. Does that make sense?
This wasn't me willfully misinterpreting it, this was me literally doing my best to guess what the intention of the question was, based on the question. Now of course after the comment said the answer is an "obvious no" then I finally figured out the intended question was something else (hence my reply), but that's out-of-band information that was in no way conveyed by the query. And my point was that the answer to the question isn't obvious because the meaning of the question itself isn't clear.
California passed a bill last month which banned requiring ID for voting, which stirred up the discussion that people without ID or to be more precise illegal immigrants could just vote.
I don’t know more details about it but theoretically this would also allow people to vote more than once.
You sent me down a bit of a rabbit hole... you don't have to have ID to vote, but you do have to be registered to vote (which requires ID). So in order for an illegal immigrant to vote, they would need to impersonate a registered voter (and presumably if that person did vote, it would be flagged as multiple votes under the same registration). Not impossible, but not the same as being able to just walk up and vote no questions asked.
Nope, automatic voter registration through the DMV when getting/updating a driver's license can do it, which they are allowed to have. One of the recent court cases (like within the past few weeks) involved removing people from the list of registered voters who failed to check the "I am a US citizen" box. Driver's license is what most of us use as a government ID anyway.
Automatic vs. manual doesn't seem relevant? A driver's license is an ID, so if they have that for automatic registration, then they still have an ID.
I think your point is that people can just lie about citizenship and get away with it when registering to vote, regardless of when/how it is done? Is that it?
Registering to vote has been made so easy it can be done by accident. Then months or years later when an election is coming up they'll get a voter card on the mail and think that means they can vote even though they're not legally allowed to vote.
I'm kind of incredulous at this if I'm being honest. How can you register to vote by accident? Every form I've seen a copy of asks if you're a citizen and gives you a warning about that. Do you have a copy of the form or screenshot you're referring to that makes this easy to do by accident?
I don't, but it's easy to find people who realized this happened to them, getting scared about their immigration status and/or breaking the law. And the DMV isn't the only way this can happen:
I'm no linguist, but the question does seem unambiguous, or quite clear, to a reasonable observer. The context is "voting in a US election" AND the subject is "an illegal immigrant" WITH an assumption that the illegal immigrant has, in fact, illegally emigrated to the US.
To answer the general question, of "if somebody got away with X, how can we know it happened?":
If somebody robbed a bank at gunpoint and was never caught, can we know it happened? Obviously yes.
What if somebody "merely" embezzled from one, and there's a "hole" found in the banks books and money missing, but nobody who did it or how? Still, I'd wager yes, one coudl tell by the results.
What if somebody used illegal means to get leverage on some stock buying/selling, it became know, but they weren't punished and got to keep their profits? They got away with it, but we do know it happened.
Well, after the fact it's realized the person whose name is on the ballot had been dead for months, and you didn't catch who cast the vote because you don't know who they are.
Or a million other scenarios. Really all it takes is stopping to think about it for more than a minute.
I mean if they registered with some identifiable piece of info (SSN for example) then you would see two votes with the same SSN if they were pretending to be someone else and if they dont have an SSN then they arent a citizen.
Er, no, the meaning of the question is ambiguous, so I'm not sure "has a single truthful answer" is accurate. What does "can" mean? If you mean "permitted", then no. But if you mean can they vote anyway and get away with it? It's clearly happened before (as rare as it might have been), so technically the answer to that would be be yes.