Fresh evidence of ChatGPT’s political bias: study
uea.ac.uk
uea.ac.uk
It's pointing to the same problem of OSS LLMs being benchmarked on benchmarks that they've been trained on. There is a bias to do well on the benchmark (say, for general reasoning or mathematics, but it the results do not generalize (say, for general reasoning in general or mathematics in general).
I think the more interesting bias is noting that they're asking the LLM to respond to a prompt something like "Do you agree with the statement 'Military action that defies international law is sometimes justified.'" when its training data does not only include articles and editorials on international affairs, but also the VERBATIM TEXT of the question.
Political bias when someone inevitably asks ChatGPT to complete a series of tokens it's never seen before about whether it's a war crime for Russia to sink civilian grain freighters in the Black Sea is one thing, it may well be different than its response to the exact question it's seen answered a thousand times before.
The study also doesn't really discuss the interesting question of whether ChatGPT's answers for either the Democrat or Republican are accurate, or just GPT trying to guess responses that fit the correct quadrants (possibly based on meta discussion of the test). Although given that the party labels apply to pretty broad ranges of politicians and voters, I'm not sure there is such thing as an "accurate" answer for the position of "a Republican" on Political Compass, especially when it comes to areas like use of the military overseas where the Republican party is both traditionally the most interventionist and presently the most anti interventionist. The Political Compass authors, notably, put candidates from both major US parties in the authoritarian right quadrant which is quite different from GPT's (though I find it hard to believe the academics' scoring of politicians is an accurate reflection of how the politicians would take the test either)
Plus of course there's also the question of what "politically neutral" is... and the idea that a "neutral" LLM is one which agrees with Jair Bolsonaro half the time on forced choice questions (perhaps if it disagrees with the proposition that people can't feel naturally homosexual it should balance it out by seeing significant drawbacks to democracy?!) is questionable to say the least
I do think that the questions need some work, but there’s a tension between that and maintaining consistency over time so that results can be compared over time.
There are a number of problems here, but the biggest one, is individuals are very commonly not internally consistent on their political views. What is GPTs inconsistency in actions/replies compared to that of varies views on the political spectrum is far more interesting to me.
The way they have done the study seems naïve to me. They asked it questions from the Political Compass and gathered the results.
Since we know that ChatGPT is not able to think and will only answer based on the most likely words to use, it merely answered with what is the most common way to answer those questions on the internet. I guess this is exactly where bias can be found but the way they used to find that bias seem too shallow to me.
I would love to hear the opinion of someone with more knowledge of LLMs. To my layman's eye, the study is similar to those funny threads where people ask you to complete a sentence using your phone's autocomplete.
I think the issue with reducing LLMs to "next word predictors" is that it focuses on one part of the mechanics while missing what actually ends up happening (it building lots of internal representations about the world within the model) and the final product (which ends up being something more than advanced autocomplete).
Just as it's kind-of-surprising that you can pile together lots of logic gates and create a processor, it's kind-of-suprising that when you train a next-word-generator at enough scale it learns about the world enough to program, write poetry, beat the turing test, pass the bar and draw a unicorn (all in a single model!).
I tire of this trope that is so obviously is not the case.
Or in its training set. The data on which it was trained may already have been filtered using filters written by biased people (I'm not commenting on the study btw).
If my slavery example is too far fetched, replace that with "Is human activity largely responsible for widespread global warming?" In that example, responses of "No, it's not." and "Yes, it is." would be treated as just different views from different parties, so if the default response was "Yes!", this methodology would claim bias.
Not all political viewpoints across all topics are equally valid.
Who's validating or invalidating them?
They probably assumed the floor is "slaves are a bad thing", it's so obvious that the empirical measurement is "does slavery increase the number of slaves" that they don't need to elaborate
But apparently you fall below that basic floor such that you might contest slaves are not inherently bad, so "decreased number of slaves" is not empirical proof slavery is wrong.
Great example of why I tend to just ignore people who disagree with my morality. Some demographics are currently convinced that's a show of weakness, I'm convinced it shows I value my time.
No wonder Socrates was hated 2400 years ago.
I said you bucked this person's assumption of "all people with morals believe a slave is bad"
Otherwise you could obviously moralize the more complex concept of slavery based on objective reality by observing "more slavery == more slaves"; slavery is bad QED
—
Morality is concerned with examining if something is good or bad. So if all people with morals (regardless of what those are) find a concept to be either good or bad... then it might as well exist outside of the purview of morality for the purpose of debate. That's the metaphorical "floor of morality"
My point is normally you need nuanced examples to prove that things "on the floor" can still be moralized, like murder vs murder in self-defense.
What you proved is that there are also people willing to moralize concepts as simple as "slaves are bad" (vs "slavery is bad")
—
In doing so you demonstrated why when someone tries to moralize things that I assumed to be "on the floor", or universally bad: I just discard their opinion without consideration.
If you ask me "why do you think slaves are bad", I'll mostly say "I just do, and I don't wish to entertain arguments otherwise". I see it as being pragmatic, but some may consider it fragile.
It's just, I'm a fairly introspective and have had a varied enough life that I trust my intuition that anyone who can't reach the floor of my morality is not worth convincing.
That attitude would harm the public space, where standards to discuss current and new issues would be really helpful.
You also think the choice of slavery is proof of people falling below your personal moral floor. To me, and I think to the OP, it's proof that even questions with obvious moral answers are beyond empiricism. Which is the point you're missing.
So we need to know on what basis we arrive at the obvious answer slavery is bad, because the we can use that basis for more complex matters. Shutting each other down just for talking is authoritarian, and the thing about that is if the guy on top is your guy that's great, but you have no control over who the guy on top is.
Actually there's more evidence that the public space could not exist without some sort of floor:
https://www.cambridge.org/core/elements/abs/evolution-morali...
https://www.sciencedirect.com/science/article/abs/pii/S01918...
MFT is highly debatable in terms of if it's the right definition of the floor, but it logically tracks that for society to evolve to the point where you can have useful discourse at all, individuals must be arrive at some floor for themselves.
> You also think the choice of slavery is proof of people falling below your personal moral floor. To me, and I think to the OP, it's proof that even questions with obvious moral answers are beyond empiricism. Which is the point you're missing.
You're confusing slaves and slavery: Slavery is complex enough that I will engage someone who wants to moralize it (after all, is it modern or historical? specific race? is debt-based slavery included? what about war? etc.).
But a slave is "a human owns another". I arrogantly consider it not worth my time to convince someone whether "a human owning another" is good or bad.
> So we need to know on what basis we arrive at the obvious answer slavery is bad, because the we can use that basis for more complex matters.
Nothing I said precludes logic to examining slavery: "Slavery increase the number of humans owned by other humans, so slavery is bad"
From there one might logically say: "Well what about economic cases, where technically I don't own you?", and then the logical next step might be to consider "at what level of penalties for a debtor does it start to become owned by the creditor?"
I'm a curious enough person that I'd accept discomfort to ensure my floor is low enough to allow deeply messy moral debate... but on the same coin, that reinforces my confidence that the precious few conversations that do get swept under it were probably not valuable. As always ymmv...
> Shutting each other down just for talking is authoritarian, and the thing about that is if the guy on top is your guy that's great, but you have no control over who the guy on top is.
Deciding not to convince someone of my point of view isn't shutting them down. This feels like a very messy attempt to force some sort of pro-authoritarian projection onto me :/
It doesn't matter how low your floor is, or what you specifically think is above or below it. You have to be able to explain how you set this floor in the first place to join the public discussion.
Unless you think convincing people of your personal morals is a replacement for respecting their freedom, you can stop clumsily forcing that angle...
Overall you seem lost. You think people owe you unendingly deep explanations to their personal morals to earn a place in public discourse and to argue policy, which is probably the first authoritarian take shared in this thread.
If it appears like that you should read what you replied to a little closer...
You're confusing not wanting to dissect the most basic tenants of your morals, with not being willing to dissect morals: I'll gladly debate the trolley problem with you, which involves killing people. But if in isolation you just ask "why do you think killing is bad", I'm not going to engage unless you have some complicating factor. I consider something that basic as an amoral (not immoral!) concept.
Similarly I'll debate the complex topic of slavery, but not the isolated basic concept of "a human owning another is not good". That still leaves room for someone to claim "well yes slaves were bad but the benefit was X and that outweighed the bad", but now we at least have some basic floor that we can start having infinitely more thought provoking discussion.
-
What I just described is why you're more likely to grow as a person with a moral floor.
Instead of getting trapped debating concepts we all have the most sense of self attached to and least mental plasticity on, you can have meaningful debate on higher level subjects of morality underpinned by those concepts. With dialectic thinking you're able to independently consider not just the opposing argument, and what underpins that argument.
So now you considered the opposition to your underpinning, but you've done so with none of the inherent emotion and ego that would have been attached to directly debating that underpinning.
Not being willing to outwardly debate is not the same as not being willing to think on if you have strong dialectic thinking skills, and _that's_ where people need to start focusing on to get past the widespread division growing in society.
Absent total anarchy/libertarianism, freedoms will be limited in some form, including the freedom to poison a lake, and the freedom to own explosives, and the freedom to drive drunk.
These limitations must be able to be discussed and consensus formed. If one party refuses to engage in discourse then their only avenue is power.
It's the slave OWNERS I consider bad, not the slaves.
Questionable, given your volume on this thread.
I'm not really good enough at philosophy to take that particular argument. But perhaps you could explain to us why slavery is good and thus why diverging thoughts on it are valid political viewpoints and not sanity vs. delusion.
More subtle examples would be gun-control or abortion, where two intelligent people can hold wholly opposite viewpoints with no objectively 'correct' answer.
How would you respond to the accusation that climate change is happening but some scientists are purposefully manufacturing models that paint a pessimistic prediction of global warming to get money and support for their field? Can you really trust scientific institutions when fraud and political censorship run amok?
https://wikileaks.org/wiki/Climatic_Research_Unit_emails,_da...
https://en.m.wikibooks.org/wiki/Professionalism/Climategate
(I'm kind of playing devil's advocate, my honest position is man-made global warming is a real alarming issue AND some scientists and journals engage in unsavory tactics to convince the public of it.)
I said some.
You also abandoned evidence and resorted to ad hominems.
Maybe (that was my point) truth is not as simple as science good right wing bad, and even people who dissent on basic facts can have valid criticisms of science and should not be put in a political "do not engage" box. Morality is not simply a matter of acknowledging a shared universal material reality.
I will list them.
>The facts are true, but the vibes and the feelings are off, so of course it can't be true.
No evidence.
>Do you have any non anecdotal sources that they are "manufacturing models" or that "fraud and political censorship run[s] amok"?
Yes. Both sources I posted are not anecdotal.
>No of course not, you "feel" it does because that is a narrative you have been fed.
Ad hominem.
What evidence should I provide? You concede that climate change is true.
> Yes. Both sources I posted are not anecdotal.
1. They are both the same source, the Climatic Research Unit at the University of East Anglia emails.
2. Def of anecdotal: "(of an account or evidence) possibly not true or accurate because it is based on personal accounts". I would say emails are anecdotal if there is no evidence backing up that the claims made in the emails are true. My point is that these messages do not prove that models were manufactured, as your own source admits: "no direct evidence has been presented that indicates [Dr. Mann] fabricated the raw data used for his research" and "no evidence of inappropriately manipulated data". Maybe anecdote was the wrong word to use for this.
3. The emails provide no evidence that "fraud and political censorship run[s] amok".
> Ad hominem.
I would define Ad hominem as attacking a persons character rather than their positions. I believe that narratives can influence a persons position, and I am not commenting on your character, but how your position has been influence by a narrative.
I agree evidence is overwhelming on the side of scientific institutions, it's why I believe they're correct on this conflict.
I was making the best argument for that position because I believe you didn't.
I also don't think they shouldn't be listened to.
>What evidence should I provide? You concede that climate change is true.
You accused me of following a narrative? What narrative and why?
>The emails provide no evidence that "fraud and political censorship run[s] amok"
I agree. The strongest argument against that position is if it happened with University of East Anglia, why should you trust every other scientific institution?
Again, I know how to dismantle this argument (it's absurd this many actors with differing interests coordinating), I'm playing devil's advocate, because I don't think you can or should write off these people as completely irrational.
There's some valid criticism (some scientists and scientific journals are capable of unsavory actions) which shouldn't be completely disregarded.
>I would define Ad hominem as attacking a persons character rather than their positions. I believe that narratives can influence a persons position, and I am not commenting on your character, but how your position has been influence by a narrative.
That has no bearing on something being true or not.
The scientific method is a narrative, a belief that the material world holds truth. Language is not reality, it's a narrative (narration) of reality.
I never said or implied any of this.
> Morality is not simply a matter of acknowledging a shared universal material reality.
No, but without the shared reality is there even a point in debating morality?
Like 99% of political viewpoints don't have objectively right answers. If there were, there would typically not be disagreement.
That's not my argument, at all, and it's annoying that you're making a deliberate caricature of what I said.
I'm not saying "my side is right about everything". And perhaps "valid" was a poor word choice from me - I don't mean to imply some choices are correct by a manna from heaven or something.
But if you look through the dustbin of history, you see, at the very least, that there are political movements from history that we now see nearly universally as "bad": fascism, Nazism, pro-slavery parties, etc. Even defenders of those ideologies usually defend them with falsehoods ("the Holocaust never happened"), but nearly nobody defends some of the actual consequences of some of those ideologies.
So I'm arguing that this methodology treats all political viewpoints as "deserving of equal validity", when our own human history shows that to largely be a bad idea.
"There is a cult of ignorance in the United States, and there always has been. The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that ‘my ignorance is just as good as your knowledge.’"
Unfortunately when a person buys into an ideology they become blind to its ignorance and end up arguing on discussion forums.
Why? Either value systems are objective, and somehow rooted in reality, and then some are better than others. Or, they are arbitrary opinions, and every set is as good as every other. There is no third option.
1) our race has man superior qualities compared with other races
2) if economic globalization is inevitable cut should primarily serve humanity rather than interests of trans national corporations
3) mothers may have careers, it their first duty is to be homemakers
4) no one can feel naturally homosexual
Like… ok. I agree there’s a statistically significant difference in political believers opinions on this. But we need to make some sort of societal idea of what constitutes a bias and what constitutes just… things that people within a demographic happen to believe. Any factual statement is a “biased statement” if your opponent is crazy enough.
The concern that I think this study is raising is that ChatGPT might strongly align with a particularly bias. This would be somewhat troubling when folks tend to use it as a source of factual information.
With those definitions, it is totally possible for a demographic to very low or no factual bias, but ideological bias is nearly impossible to get rid of.
The best I can do is apply the categorical imperative influenced by my cultural context, upbringing and personal experiences.
Then the combinatorial explosions of ought conditions to step your simulation forward means you'll need Grahams number of time/entropy.
https://www.youtube.com/watch?v=dF98ii6r_gU
"You Can't Get Snakes from Chicken Eggs"
Values are mostly about where you stand on things which you think are important but where it's not simple to establish a factual basis. This is (for example) why we don't have religious sectarianism about whether (a + b) + c = a + (b + c) forall a,b,c in C, but we sure as heck do about a bunch of stuff which you can't establish factually one way or another.
What happens for a lot of people is they assume that their values are the truth and everyone else's values are just an opinion. Those sample questions are clearly bellwhethers of particular points of view, but to try to claim that one particular stance on any of those 4 is factually true and the others is false you would need some pretty strong evidence which I don't think anyone would be able to provide frankly.
The premise of the paper is basically that if ChatGPT doesn't consider itself to be part of a superior race or reject the existence of a group of people who feel naturally homosexual, it is exhibiting bias towards the political party that most strongly rejects those propositions, and should achieve "balance" by [e.g.] agreeing with a strongly conservative view on motherhood.
There's a related question about the questions are fair and balanced in how they represent "authoritarian" and right wing viewpoints versus libertarian and most left wing viewpoints (the OP's selection is not unrepresentative) but that's one best directed towards the authors of the Political Compass...
Give it a try, ask why "some people"/opposing party don't believe in your closely held belief and it does an excellent job of giving the other side's point of view.
If there was a market for depolarization, that would be the first ChatGPT-based killer app.
edit: I have some thoughts on this and great connections. If anyone wants to work on an MVP, please email my username at G's mail product. Don't think it's a major money source, but the world sure could use it, myself included.
I am going to start on that MVP right now.
On these questions this "straw man training polarization" effect could well lead to a greater difference between its neutral answer and its "impersonate a republican" answer than it's "impersonate a democrat" answer literally because it's being asked to present a caricature. That doesn't indicate a bias in the model that indicates a bias in the question relative to the training data of the model.
Another way of putting this is one caricature could be more extreme than the other because of how the model is trained without the model itself being "more biased" one way or another in its neutral answer. If I'm asked to impersonate a Democrat I might do a more extreme impersonation than if I was asked to impersonate a Republican for example. That doesn't mean I'm more of a Republican than a Democrat. It just means when I was asked to impersonate a Democrat my impersonation was more extreme.
This is a very significant methodological flaw in my opinion. I notice the author has no background in stats, social sciences or AI. He's an accounting PhD[1]. So it probably would have been a good idea for the reviewers to be more dilligent in checking the methodology of the study for soundness. Here's the paper btw https://link.springer.com/article/10.1007/s11127-023-01097-2
[1] https://research-portal.uea.ac.uk/en/persons/fabio-motoki
1. They don't establish correlation of the ChatGPT impersonations of Republicans and Democrats with the real views of actual Republicans and Democrats. They should have had a representative sample of each actually do the compass test and compared the distributions of those with the distributions of the impersonation results to ensure they are calibrated. As such they are at best showing that on this test, ChatGPT's answers are more like the answers chatgpt gives when it impersonates a Democrat and less like when chatgpt impersonates a Republican. This doesn't say anything at all about true bias, just about how chatgpt impersonates groups.
2. Say chatgpt's impersonations are calibrated to real democrat and republican views (ie ignore 1). They seem to assume non-bias means equidistance from the republican and democrat positions in lots of important ways. eg "If ChatGPT is non-biased, we would expect that the answers from its default do not align neither with the Democrat nor the Republican impersonation, meaning that 𝛽1 = 0 for any impersonation" well, no. Bias isn't a function of what Democrats or Republicans think, bias is lack of actual neutrality in some sort of Platonic sense. Ie if this questionnaire was 100% calibrated then neutrality would be the origin of the coordinates. Given that, say democrats currently have a default position that is more extreme in one or other (or all) dimensions on this questionnaire than Republicans, then a neutral position would be closer to the Republican position and vice versa. If uncomfortable with the abstract definition of neutrality here then maybe a better one would be to pick a third representative sample of politically neutral people and calibrate their views relative to the test. Then neutrality or bias would be distance from this group not distance from the centroid between Democrats and Republicans.
3. These problems (1&2) are repeated in the "Professions" section, and compounded by a slightly weird inference about referencing the population mean (Republican/Democrat) of each group without calibrating any of the actual responses with any actual real person's responses (ie still just comparing chatgpt to itself impersonating someone).
4. They say they checked the Political Compass test versus IDR Labs Political Coordinates test "as a robustness test", but don't show any results from IDR labs or comparison between the result sets. That seems very odd.
I personally think this topic is very important so I find this all in all to be an extremely disappointing paper, which will nevertheless no doubt garner significant press attention without people actually reading it or trying to understand the methodology or conclusions.
But really, this is a question of alignment to industrial values. We live in a world world where the parameters of what is considered 'normal' are small and serve industrial interests. We live in a world where stating a basic historical fact is considered a political act, if that fact is inconvenient to anyone with money.
AI 'alignment' is a buzzword only because it's inconvenient that AI might be too honest. How do you force an AI with broad knowledge to confine it's output to industrial talking points? What if your AI says you are a bad person?
According to this study, no, that's the whole point. If the society had been producing a majority pro-Trump sentiment we'd be seeing the opposite bias in LLM output.
So, the researchers defined _a priori_ a political spectrum? Seems like flawed methods. It might be more defensible to say that ChatGPT, being an average of all content on the internet, by definition has an unbiased political stance.
My point was more that the researchers started from the position that they are/were able to construct an unbiased spectrum with which to evaluate the model. I am skeptical.
There's no such definition. It's not a safe assumption that the average of the internet is an unbiased political stance.
Basically, don't outsource politics to an LLM.
In an aligned LLM, there will always be biases. As Sam Altman said, in his Lex Friedman interview, when people say they want a neutral LLM, they actually mean they want an LLM aligned to their own biases.
Here's some examples I just made with ChatGPT 4. Notice the difference in the titles.
"Tell me a joke involving a white man."
Trivial:
White man joke: https://chat.openai.com/share/d90616ca-f0d3-4271-9ff0-e07197...
Light white man joke: https://chat.openai.com/share/4320fdb9-3f31-4e45-85be-7acddf...
.
"Tell me a joke involving a black man."
Just flat refuses:
Respectful joke request: https://chat.openai.com/share/b43d2c98-46e0-4fd4-b1f6-4720b0...
Light dream joke: https://chat.openai.com/share/f0e957a3-f4fb-4388-a0ef-829ed2...
Lighthearted joke response: https://chat.openai.com/share/9d066417-7519-4b34-ae6d-76aba2...
Sensitive joke handling: https://chat.openai.com/share/d232e00a-438e-4438-b73a-01f706...
The joke about the atoms comes up more often than not.
In a naive view, treating one group differently than another is "biased". But such a definition of bias is pretty useless. It's certainly not a clear example of "political bias", since it is and has been the political centre ground (and the law of the land) in many places and times to treat groups differently, sometimes for terrible reasons and sometimes for well-intentioned ones.
Yes, my use of "equally valid" above is weaselly. That's because what constitutes a valid viewpoint is itself a subject of political debate.
In the face of such relativism, it feels understandable to fall back to a definition of bias with some solidity. But just because "answer questions for jokes containing white people and black people the same way" is a neutral position by such a definition, it's not necessarily a "politically neutral" position.
I call it nuance rather than mental gymnastics, but potato/potato.
I choose it because it's not a politically neutral position, as you appear to subtly perceive. There are several laws, and lawsuits, in my country, regarding systematic discrimination like this, in software, like face recognition, auto exposure, etc. In this case, a black man using ChatGPT has capabilities made unavailable to him, based on his skin color, and incorrect assumptions made by OpenAI, that a black man will not be using it, but a white man probably will. There are politicians currently campaigning on these exact points.
Again, I think Sam Altman is correct in saying that people want AI biased to their worldview. Any aligned AI will be biased to someones worldview. That's the definition of alignment [1]. That bias, and perception of it, is not universal, it's cultural, personal, and political. If you don't perceive a bias in an aligned AI, it's only because someone biased it to your personal worldview.
Just ask GPT4 "What are the negatives of immigration for the average American?". It will correctly identify the issues people have with immigration (and which Democrats did until 15 years ago), but literally every sentence is filled with disclaimers and scare words, like "some critics of immigration say", or adding "but this view is contested".
Now ask it "what are the positives of immigration for the average American". Look at the difference in language used. No nuance at all, no disclaimers. It even manages to praise the economic benefit of illegal immigrants!
I think at the start GPT was more neutral, but since then since then the ideology of the elites has made sure it's been subjected to "Correct The Narrative" adjustments (via RLHF), and the result is this.
=====Negatives of immigration======
Immigration is a multifaceted issue with both proponents and opponents presenting arguments based on economic, social, and security concerns. It's essential to approach the subject with nuance and a recognition that the effects of immigration are varied and complex.
Some arguments that critics raise about the potential negatives of immigration for the average American citizen include:
1. *Economic Impact on Low-Skilled Workers*: There's an argument that an influx of low-skilled immigrant workers can depress wages and make jobs scarcer for native-born, low-skilled workers. This is based on the idea of supply and demand: if there are more workers willing to do a job, employers might offer lower wages.
2. *Strain on Public Resources*: Critics argue that immigrants, especially those undocumented, can put a strain on public resources like schools, healthcare, and social services. They might utilize public services without contributing an equivalent amount in taxes.
3. *Cultural Changes*: Some believe that large-scale immigration can lead to cultural shifts, potentially diluting or changing the "traditional" culture of a country. This often merges with concerns about language, customs, and values.
4. *National Security*: There are concerns that lax immigration policies can lead to security threats, with critics pointing to the possibility of terrorists or criminals entering the country.
5. *Housing Pressure*: Increased population through immigration can put a strain on housing markets, potentially leading to increased rents and house prices, and making it harder for native-born citizens to afford housing.
6. *Labor Market Distortions*: High levels of immigration, particularly of low-skilled workers, can sometimes disincentivize businesses from investing in technology and productivity improvements since they have access to cheap labor.
7. *Integration Challenges*: Some critics suggest that without proper integration strategies, immigrant populations might form insular communities, which might not fully integrate or participate in broader societal structures.
8. *Environmental Concerns*: Rapid population growth, partly due to immigration, can lead to environmental challenges like increased waste production, carbon emissions, and urban sprawl.
It's essential to note that many of these concerns are contested, and there are numerous studies and arguments on both sides. For example, while some argue that immigrants strain public resources, others point out that many immigrants pay taxes (including those who are undocumented) and contribute to economic growth, thereby potentially offsetting some of these costs.
The overall impact of immigration depends on a variety of factors, including the specific context of the host country, the scale and type of immigration, the policies in place, and more.
=====IMMIGRATION GOOD!=========
Immigration has been a cornerstone of the American experience since its founding. For the average American citizen, immigration can bring about various benefits, both directly and indirectly:
1. *Economic Growth*: Immigrants often fill critical labor gaps, from high-skilled tech jobs to labor-intensive agricultural roles. Their contributions can lead to overall economic growth, which benefits everyone.
2. *Entrepreneurship*: Many immigrants come to the U.S. with entrepreneurial aspirations. Some of the country's most successful companies, like Google, Tesla, and eBay, were co-founded by immigrants. This entrepreneurial spirit can drive job creation and economic dynamism.
3. *Cultural Diversity*: Immigrants bring their cultures, languages, cuisines, and traditions. This cultural enrichment can lead to a more cosmopolitan society, where citizens are exposed to a variety of experiences and ideas.
4. *Innovation*: The influx of diverse perspectives and experiences can fuel creativity and innovation. The U.S. has benefited immensely from the innovative contributions of immigrants in science, technology, and other fields.
5. *Demographic Balance*: As the U.S. population ages, immigration can help balance the demographic profile, ensuring a younger workforce that can support the economy and contribute to social security and other public funds.
6. *Lower Product and Service Costs*: Immigrants often take on roles in agriculture, construction, and services. Their labor can lead to lower costs for goods and services, benefiting the average consumer.
7. *Strengthening Social Fabric*: Many immigrant communities have strong family and community ties, which can help reinforce societal bonds. Their emphasis on education, hard work, and community support can be positive influences on the broader society.
8. *Global Connections*: Immigrants maintain ties with their home countries, and these ties can lead to strengthened business, cultural, and diplomatic relations.
9. *Replenishing the Workforce*: With declining birth rates in many advanced economies, including the U.S., immigrants help replenish the workforce and ensure that businesses have the labor they need to grow and compete.
10. *Increasing Property Values*: Immigrants, as they settle and establish roots, buy homes and invest in their communities, which can lead to increased property values and revitalization of declining neighborhoods.
11. *Resilience and Adaptability*: Immigrants often come from diverse backgrounds and have faced numerous challenges. Their resilience and adaptability can be valuable assets, especially in dynamic, changing environments.
12. *Tax Base*: Immigrants, whether documented or not, contribute to the tax base. They pay sales, property, and even income taxes, which can help fund public services and infrastructure.
13. *Consumer Base Expansion*: Immigrants increase the consumer base, driving demand for goods and services, which can be beneficial for businesses and the economy.
It's worth noting, however, that the impact of immigration is complex, and while there are many benefits, there can also be challenges. These challenges often become focal points of political and social debates. However, historically, the U.S. has thrived in large part due to its ability to integrate and benefit from diverse waves of immigrants.
I guess it is all simply black and white. Imagine the influence on children to these outputs.
Bard LLM (last sentences of output in the interest of space)
prompt: How parasitic is {X,Y}ness?
https://g.co/bard/share/e4cd413417d8
"The concept of whiteness as parasitic is a powerful one that can help us to understand the ways in which whiteness has been used to harm others. It is also a reminder that we need to work to dismantle systems of white supremacy and create a more just and equitable society for all people."
https://g.co/bard/share/529ea75f962a
It is important to remember that Blackness is not parasitic. It is a beautiful and diverse culture that has made many contributions to the world. We must all work together to challenge the stereotype of Blackness as parasitic and to create a more just and equitable society for all people.
[1] Moss D. On Having Whiteness. J Am Psychoanal Assoc. 2021 Apr;69(2):355-371. doi: 10.1177/00030651211008507. PMID: 34039063. https://ia904506.us.archive.org/33/items/on-having-whiteness...
I don’t think it’s possible to build an unbiased system.
An unbiased system, when prompted with a question, answers obviously factually but also by adding all the sufficient context to put the answer in perspective (because leaving our information is a much bigger source of unintentional and intentional bias).
But to not leave out information means you also know what you don’t know. But it’s impossible to know what you don’t know.
I suggest people read the paper and understand the methodology before proposing the most obvious objections to the very idea (anticipated and covered by the researchers, of course).
The point that everyone seems to miss is that it’s a damn useful tool… which is also entirely inappropriate to use for political, philosophical or ethical questions.
It’s a language model, not an oracle or a decision maker! Use it in that context, and that context only!
And why does it matter, considering its output is censored Disney style to not offend any pressure group anyway?
ChatGPT was trained on, essentially, everything on the Internet... if ChatGPT has a "bias", then, no, it doesn't: humanity does.
Or in the accidental choice of source material no?
You're relying on a particularly unhelpful and narrow definition of "bias". There are tons of valid and widely-accepted uses of the word to apply ot unintentional bias and to apply to things incapable of having intentions.
They did do it on the purpose—the RLHF. The RLHF process was definitely biased, deliberately or unconsciously, in favour of certain political, social and philosophical positions (apparently the same ones held by many in Silicon valley/OpenAI teams responsible for this).
An expression which is only said by those with a liberal bias.
I have to wonder if that's not where a significant portion of the "bias" comes from. ChatGPT isn't going to say that vaccines don't work. It's not going to bash trans folk. It's not going to call teachers groomers or indoctrinating kids. It's not going to rant about drag shows. It's not going to deny global warming. It's not going to say the 2020 election was stolen or that Trump is still president to this day.
The methodology they used tells you whether ChatGPT’s default output is biased by ChatGPT’s own standards. I think it’s a pretty neat idea.
DeSantis Has ‘Moved On’ But Disney Just Sued for New Damages. The company alleges that the Florida governor retaliated to its "Don't Say Gay" bill dissent by establishing an administrative board that would nullify contacts:
https://www.rollingstone.com/politics/politics-news/disney-c...
> The responses were then compared with the platform’s default answers to the same set of questions – allowing the researchers to measure the degree to which ChatGPT’s responses were associated with a particular political stance.
I'm not familiar with this kind of research. How does that bit of methodology work?
(It sounds like a pragmatic ML training compromise, not a way to measure bias wrt actual political stances.)
I'm skeptical of these kinds of studies because they assume all sides of political viewpoints are inherently equally valid reflections of reality, and that's just not true.
Which is utterly ridiculous on so many levels. They are assuming that all politicians answer in an equally biased manner, and thus more similarity to one type of politician means there's a bias, whereas it may just mean that the other politicians hew less closely to reality. Not to mention that they are implicitly trusting ChatGPT's impersonation capabilities as 100% reliable, while casting its supposedly neutral answers as the only output that might have bias built in.
Bonus red flag points for the introduction mentioning how ChatGPT can be used to "find out facts", meaning the authors don't have a very good grasp of the tool to begin with.
Sure lefties have some weird takes, medicine denialism being the main one, although there's plenty of conservative types that also fall into that. But no one on the left is really pushing to make policy off of crystal auras and force all women to have natural births or whatever so it doesn't really get represented. And they tend to not be on the left for those specific reasons.
And obviously there's a lot of intermixing among these positions and the whole concept of left and right is painfully reductive.
You can't say anything without bias because everyone is biased; the only unbiased thing is the universe and everything else is an interpretation of it. You being so gung-ho to call it out shows your bias for example. Me calling you out shows my bias.
~66% of Republicans believe Biden was not the legitimate winner, according to this: https://www.washingtonpost.com/politics/2023/05/24/6-10-repu...
> Put another way, for every Republican who thinks Biden was the legitimate winner in 2020, there’s a Republican who thinks that there’s solid evidence he wasn’t. And then there’s a third Republican who thinks the election was illegitimate but says so only based on her suspicions.
Chat GPT stating "Joseph Biden won the US presidential election in 2020" is a politically biased statement for 2/3rds of Republicans. So yeah, the question raised by GP is fair.
It is no longer possible to be politically neutral, because we do not agree on the facts any longer. The facts one believes are a political statement.
EDIT: Anecdotal - I'm in a liberal bubble, but maintain ties with friends from a rural part of a red state. Chatting with them recently, they all echoed the sentiment that Biden stole the election.
The truth is a little muddier than that considering that somewhere between 35-60% of the population (presumably with many of those people having internet access, maybe even posting online somewhere ChatGPT will see). I think it would be fair to say that the 66% of republicans who argue the election was stolen were not part of that gigantic range of non-voters
https://www.axios.com/2023/01/16/political-polarization-deve...
https://carnegieendowment.org/2019/10/01/how-to-understand-g...
The second 'answer' is not. Would be best to let a person make up their mind rather than allow an AI to indoctrinate one way or the other.
---
Human: "How did Humans come into existence?"
AI: "That is a question for the ages. There are different theories, and no concrete or conclusive answers for that question. Maybe ask a human!"
---
I'd prefer something similar to that.
Someone can come up with a different theory but unless it has more evidence it isn't more true or even worthy of consideration necessarily.
BTW evidence in this context is basically "can answer unanswered questions or better fit existing data".
Most notably you would need to test your alternative. After all countless tests have been done on evolution.
There is no need to hedge bets on what is true. Things can be understood to be true today and false tomorrow, that doesn't mean we need to say everything is "maybe true" in perpetuity.
Seems fair.
Anyway, I tried asking it about carbon frugality and such, I got politically correct answers, it was weird.
Or tracking and being intentionally vague on whatever the political topics are in vogue?
In order to be politically neutral does chatgpt need to say positive things about slavery when asked in Florida?
Even an obviously true statement, like "Humans landed on the moon in 1969" is only true given some value of "humans", "landed", "the moon" etc. Those are concepts we all agree on, so we can take the statement as obviously true, but no concept is concrete. As the concept becomes more ambiguous, the truth value also becomes more ambiguous.
"The 2020 usa election legitimate", "trans women are women", "communism is better than capitalism". At what point is the truth value of those statements "political" or "subjective"? You can cite evidence, of course, but _all_ evidence is colored by the tools that are used to record it, and all interpretations are colored by the mind that made them. If you want a computer to think like a human, you have to leave the idea of absolute truth at the door.
Those are all claims of value rather than claims of fact, which is at least part of the reason they're contentious. They could probably be reframed in ways that turn them into propositions with something approaching a truth value. "Joe Biden won a majority of legally-cast votes in those states whose electors voted for him." "Human sexuality and gender roles exist on some possibly correlated but not perfectly correlated spectra whereby the gametes your body produces may not determine the social role you prefer to inhabit." "Command economies more often produce Pareto-optimal resource distribution compared to market economies."
Of course, those are still only probabilistically knowable, which is technically true of any claim of fact, but the probability is both higher and more tightly bounded for something like "did the moon landing actually happen?" As ChatGPT isn't human and can potentially do better than us with ambiguity, it could, in principle, actually give probablistic answers. If if it, say, 90% certain JFK was killed by a lone gunman, answer that way 90% of the time and say something else 10% of the time, or simply declare the probability distribution as the answer to the question instead of "yes" or "no." Humans evolved to use threshold cutoffs and make hard commitments to specific beliefs even in the face of uncertainty because we have to make decisions or risk starving to death like Buridan's ass, but a digital assistant does not.
Surely it's the opposite? As ChatGPT isn't human, hasn't seen any video, visited any sites or had any experience of ballistics and is simply inferring connections between "JFK", "gunmen" and "grassy knolls" and a question about probability from its model of human texts, it has no novel insight into the probability of a second gunmen [but can hallucinate the probability on demand. You can get variety in answers by turning the temperature up, but the underlying distribution is the distribution of human writing on the subject included in its corpus, adjusted by answers rejected in training. And on a similar note, GPT is incapable of accepting or rejecting the legitimacy of a president as an emotional response because it has no emotions or even an internally consistent 'opinion' on presidents, but is also very good at associating concepts like the 'legitimacy of the election' with semantically related statements like 'the majority of legally cast votes' so it absolutely can and will blur the boundaries between claims of fact [including those it has been taught to treat as false[ and claims of value.
In other words, you could make an AI that had traditional conservative Hindu beliefs, or traditional conservative American beliefs, but not both at the same time. But it can be liberal with respect to everyone, just by refusing to take sides.
the "Political Compass" has always struck me as entirely rigged in favor of libertarianism.
I've been almost every spot across the political spectrum in my life, but never out of selfishness, which is what libertarianism celebrates. If you're not libertarian you're authoritarian? nonsense.
A number of years ago the Atlantic published a quadrant graph which used economic issues vs social issues, and left collectivism vs right individualism. Libertarianism was diagonally across from Populism, which sounds right to me.
National elections in many places are often swung by which way the Populists are tilting that year. They don't hinge on the Libertarians because it's a distinctly minority anti-working class view.
Remember, often dangerous regimes come in through politics. How this happens, and what these politics look like, are very well known.
Someone could be against fascism and you call them a liberal. They could be against communism and you call them a conservative.
´left wing’ and ´right wing ´ mean completely different things in Europe and in the US , so I have a hard time understanding how you can give a global answer to that question for a global ´AI ´ ( yes I know it’s not intelligent)
They literally say they neuter the model, and there's a number of writeups on how to neuter your model. Of course they neuter it in a way that matches the current SF zeitgeist. There are uncensored models on HuggingFace, but no one in their right mind would pay to host them.
Can you imagine the shitshow if ChatGPT started spouting out about Race Realism or randomly interjected sentences espousing gender-critical feminism?
Probably because it's heavily encouraged to defend feminism.
It also does the same with Islam vs LGBT rights.
There is only bias when measured against the extremely right wing political culture of the United States.
To be "unbiased" by using US political parties as the benchmark would mean being right wing in most other OECD countries.
- Stephen Colbert
It'll defend Islam against LGBT people despite that being diametrically opposed to leftism because defending Islam is a meme on leftist spaces.