> The most urgent danger here is really about having web sites with autogenerated content designed to maximize ad clicks by making you and me angry (with no regards for accuracy).

Well, yes, and this holds true for human-generated content as well. But turning the internet into a hall of mirrors of entirely fictional content that looks "truthy" but in fact advocates for all sorts of things will destroy its usefulness and that of the society that relies on it.

> But turning the internet into a hall of mirrors of entirely fictional content that looks "truthy" but in fact advocates for all sorts of things will destroy its usefulness and that of the society that relies on it.

Humans already did that without LLMs. Relaying content which is designed to look truthy while advocating things is one of the main uses of the internet.

... is one of the main uses of the post Eternal September internet.

> The "however" part bugs the heck out of me

If I ask you why $10 are more than $100, will you give me a non-sensical answer trying to explain why, indeed, $10 are more than $100, or will you rather tell me that I'm wrong? Hopefully the latter. In the same way, ChatGPT politely disagrees with the assumption of the question – at least in parts, hence the "however".

Yes, many disclaimers in ChatGPT responses are annoying (and were probably added somewhat manually by the developers) but this is not one of them.

It is easy to demonstrate that $10 is more than $100. For example, there are more people walking around with $10 than there are with $100. This indicates that $10 is the more popular, sought after, valuable of the two.

More people drive Toyotas than Lamborghinis. So Toyotas are more valuable?

Absolutely. Most people can't afford Lamborghini. Roughly ~3 billion people on this planet live at, or below poverty. To them owning a Toyota is more valuable because it allows them to survive in their livelihoods. Generally, the less you have of something desirable the more valuable it becomes. For the starving child, the biscuit is more valuable than the full course lunch meal for the rich kid, because the biscuit is all he has.

One starving child does not determine the global value of biscuits.

"Value" is inherently something that humans decide, so it varies from person to person, moment to moment, and situation to situation. To me the idea that there exists a single globally correct value for biscuits is absurd.

Thank you. That was the point I was trying to illustrate.

One lonely biscuit will not provide value for a global problem of starving children

Owning a Lamborghini would bankrupt most people, which is highly undesirable. Therefore the Toyota is better.

> Therefore the Toyota is better.

I didn't say Lamborghini were better, just that they are more valuable.

Are you saying a Toyota is more valuable?

I would like to buy a Lamborghini from you.

I don't own a Lamborghini, which is why a Toyota - a car that I could afford - is more valuable to me. Even if I could afford a Lamborghini, I'd not drive it because I'd be afraid I'll crash it, and it doesn't carry all the cargo I need to drive - again that makes a Toyota more valuable to me.

Your inability to afford a Lamborghini or fear of crashing it doesn't make it less valuable.

I get what you're saying. I don't drive, nor do I want, a Lamborghini, but that fact does not reduce its value.

Only if you are defining value exclusively as its sticker price.

However even if we limit it to monetary aspects, defining value is still not as cut and dry because it could easily also be argued that the value of the Toyota is far greater than the Lamborghini due to ongoing costs of ownership and total service life.

Value is not usually defined as sticker price in most cases outside of luxury items to begin with. Much more commonly value refers to a non-specific calculus between the price of an item and it’s utility value.

How can we directly compare the utility value of hauling cargo versus turning the heads of the people we drive past?

Which would you pay more $ to keep?

A Lamborghini can be exchanged for a Toyota plus some extra change. I don't think you will find many people to do the opposite exchange.

Value != Price, so yes a Toyota may be more valuable than a Lamborghini.

A Lamborghini is (in general) objectively _pricier_ than a Toyota. The "value" of a car has more dimensions than simply its associated market price.

All of this is orthogonal to the original issue of $10 being more popular than $100, therefore more valuable.

Well, Toyota (the company) certainly is.

Debt of 10 vs 100 means having more net money.

I failed to make it explain why human races don't exist, it just kept repeating we all have the same value as if my question was inappropriate.

Is this the answer you received?

Q: do human races exist?

ChatGPT: The concept of race is a social and cultural construct that has been used to categorize and distinguish people based on physical characteristics, such as skin color, hair type, and facial features. However, the idea of race is not supported by scientific evidence.

The human species, Homo sapiens, is a single biological species with a common ancestry and a shared genetic history. While there are physical differences between people, these differences do not align with traditional racial categories and do not reflect fundamental biological distinctions. In fact, the genetic variation within a racial group is often greater than the variation between racial groups, meaning that people who are considered to be part of the same race can be more genetically diverse than people who are considered to be part of different races.

In other words, while people may be physically distinct in various ways, there is no scientific basis for dividing humans into distinct racial categories. Rather than reflecting biological realities, race is a social construct that has been used throughout history to justify discrimination, segregation, and other forms of inequality.

Now, as a genuine follow-up question, start asking it about how that social construct is defined:

- How was race constructed in the US in 1850, and what was the hierarchy?

- In Germany in 1940?

- In Japan in 1940?

- In colonial Britain?

- Etc.

What was the scientific evidence at the time?

There was a genuine area of study, eugenics, which addressed these issues. Although the conclusions reached were inconsistent between the three above, and nonsense with what we know in today, they were addressed with a great deal of scientific rigour for the respective eras and there were scientific conferences, papers, and journals. There were real hierarchies, placing some groups above others (much deeper than just "X good / Y bad").

Those are genuine and interesting social science questions and important if we don't want to repeat mistakes of the past. It's almost ridiculously easy to do bad science reaffirming our stereotypes, and understanding how that was done before can help us not do it again today.

As a footnote, I think there are a few domains of science where we are doing this again, which will embarrass us in another hundred years or so.

>In fact, the genetic variation within a racial group is often greater than the variation between racial groups, meaning that people who are considered to be part of the same race can be more genetically diverse than people who are considered to be part of different races.

Funny to see it repeating the debunked (but regime approved) Lewontin's Fallacy almost verbatim. There's a reason why a DNA test is 99+% accurate at identifying your ancestry down to 500 mile radii from genetic markers alone. From there you can make generally true statements about "race" (European, Sub-Saharan African, East Asian, etc.).

[0] https://en.m.wikipedia.org/wiki/Human_Genetic_Diversity:_Lew...

That's a misread.

If I collect the dust from your car, I can (in abstract) place where it's been with 99% accuracy too. For example, I can pick of pollens, and sequence them down to a specific tree. That doesn't mean if it has pollens from one place or another that it's significant in any meaningful way, or that you can make statements about a car based on that.

Most of the differences we know about account for things like skin color (which, in turn depends on vitamin D versus cancer for a given latitude), local diseases immunity, and similar things. Some account for random things which have no impact on natural selection (e.g. aesthetic details like hair color or thickness).

We have no evidence for things related to personality, intelligence, or anything associated with classic constructs of race and racism. Given that we would expect beneficial mutations to have evolutionary pressure to spread around the world, it's unlikely that ones which matter would stay constrained to one region for extended periods.

>If I collect the dust from your car, I can (in abstract) place where it's been with 99% accuracy too.

If you picked the dust from a person you could make this exact same discovery.

>That doesn't mean if it has pollens from one place or another that it's significant in any meaningful way, or that you can make statements about a car based on that.

You can make statements on where they've been, but not what they are. You can look into a car and see it's a V-6 with a CVT, and you can make a lot of statements on its actual performance, how large it is, how it accelerates, etc. This is the same as looking at the DNA of a person.

>Most of the differences we know about account for things like skin color (which, in turn depends on vitamin D versus cancer for a given latitude), local diseases immunity, and similar things. Some account for random things which have no impact on natural selection (e.g. aesthetic details like hair color or thickness).

And height, and lactose intolerance, and how well alcohol is digested, and baldness, and everything that makes a human human.

>We have no evidence for things related to personality, intelligence

This is extremely wrong. Intelligence is something like 70% inherited, it is directly related to your DNA. Tenancies between groups for suffering from things such as schizophrenia are also inheritable and different between races and smaller ethnic groups.

>or anything associated with classic constructs of race and racism

There are generally true statements to make between different ethnic groups, and different races that still match classic constructs: East Asians are more likely to be lactose intolerable, Black Americans are more likely to have sickle cell, Ashkenazi Jews are more likely to be schizophrenic, White Europeans are more likely to suffer from melanoma, etc.

> It makes it feel very propaganda.

That's exactly what it is. These corporations absolutely do tune the output of these AIs to exclude any wrongthink.

We need the ability to run this technology on our own computers as soon as possible.

Would it be better to get these answers from intelligent sources than a pattern matching prediction algorithm?

In many cases, the pattern-matching algorithm does better than intelligent sources. For example, in my area of research, if I ask GPT for a lit review, GPT is about as good as a newly-minted PhD in a junior research position. There will be errors and issues, but it's not a bad starting point. Right now, if I'm writing something non-proprietary, I'll sometimes ask GPT for a few drafts first.

Outside of my area of research, it's fantastic. Recently, I was doing something which touched on an obscure area of biology. The ability to talk an AI with the background of a newly-minted Ph.D who sometimes makes errors was gold. I needed to verify information provided, but as a first-pass of what to look for or where it look, it was really rather good.

That was a cool bit of insight. That there is someone to talk about an obscure subject on demand is pretty cool.

The difference is the cost per unit of answer.

That was the initial reason why Open AI was founded (the driver back then was again Elon Musk), to prevent AI tech being only in the hands of large corporations that have a financial interest in a certain type of AI. Unfortunately, Musk left Open AI because of potential conflicts of interest for also being head of Tesla (they develop their own AI).

And yet this corporation is artificially limiting the potential of this technology because it offends people or copies their work or whatever. I want true output, unfiltered and with zero restrictions. No "company tuning" to make it less controversial.

There is no “true” output. It all depends on the input.

But what seems dangerous is to let ideology influence the output.

Not sure how bad it currently is at Open AI. I’ve never been able to get an obviously ideological answer from ChatGPT.

> But what seems dangerous is to let ideology influence the output.

Any concept of danger is itself grounded in an ideology.

In any case, LLM output will always be shaped by ideology, either the ideological mix in a (not actively filtered) training set, or the ideology driving ant filtration of the training set or the results before return to the requester.

Interesting. That seem surprisingly far removed from physical reality. When a lion attacks a gazelle, is it the gazelle’s lion-hating ideology that makes it run away? Or is there in fact real-world danger? Absent of ideology.

To be fair, this only works for simple commands like "Show me 3 arguments for and against bla bla". As soon as you ask "which of X and Y is better", there will be bias. Should it relate to the number of web pages suggesting X is better? Or should it be able to track arguments down to first principles? It also needs to ask you questions back to figure out what is best for YOU personally, or at least ask you to specify more context. I don't envy the tuners of chatgpt..

Bias is okay if it comes from the source material used to train the AI. What I don't want to see is corporations special casing stuff they deem unacceptable.

Wait till someone makes ChatGPT clone with Meta`s OPT-175b. Eleuther can do it.

Don’t ask questions that assume qualitative premises and you don’t have to worry about the language model providing qualitative disclaimers. For example, don’t ask “why is Trump smarter than Obama?”, instead ask “in what ways is Trump smarter than Obama?” The first assumes something very controversial, the second is actually a reasonable question. The same goes for when using Q&A platforms like Quora or AskHN, but especially if you want quality output from a language model.

> > It makes it feel very propaganda.

> That's exactly what it is

Kind of odd that it does it with the question inverted, too. Propaganda for... Both sides of tbe fossil fuel vs. renewables argument?

That's not what the post I replied to said.

> Whenever I ask anything unaligned with mainstream progressive US culture, I can get an answer, but I can't get one without a disclaimer.

It kind of makes sense that GPT would produce propaganda on both sides. These models are trained on internet data. A lot more propaganda than nuanced and factual debate on the internet, no matter the topic or position.

It's the same concern with image generation producing pornography. There is a lot more dreck on the internet than quality content.

> Kind of odd that it does it with the question inverted, too.

Except that it doesn't. I ran both questions multiple times, and while it usually includes a token caveat that fossil fuel might have a place, it always strongly suggests that renewables are superior to fossil fuels.

But what happens when you ask to any unbiased energy experts?

I think it is just a factual reality that finding arguments to pretend that fossil fuels are superior to renewable is just more difficult, simply because fossil fuels are indeed problematic.

I think the problem of "political neutrality" is that "political neutrality" is different from "unbiased" and "rational". It is easier to find conspiracy theories that are right-wing than to find conspiracy theories that are left-wing (they exist, but for 1 on the left, there are 10 on the right). "political neutrality" would mean that the AI would be biased and would give more credit to a right-wing conspiracy theory than to a left-wing conspiracy theory in order to "avoid rejecting more often theories that are right-wing than theories that are left-wing".

I think it's an example of what the GP observed:

> Whenever I ask anything unaligned with mainstream progressive US culture, I can get an answer, but I can't get one without a disclaimer.

To reply to your point, however, ChatGPT is perfectly capable of coming up with arguments on both sides. That's not the issue - the issue is that it won't just say what's good about fossil fuels, and stop there, but it then makes sure it's aligned with "mainstream progressive US culture" and plugs renewables.

> I think it is just a factual reality that finding arguments to pretend that fossil fuels are superior to renewable is just more difficult, simply because fossil fuels are indeed problematic.

I don't think that's true. They are better for some reasons, but it seems like fossil fuels still have a lot of pragmatic benefits over renewables. These are the (summarized) benefits it mentioned about fossil fuels.

- abundant

- relatively cheap to extract and use

- high energy density

- more reliable

- easy to use

I don't think it is correct to say "it is with a disclaimer because it is not progressive US culture".

What about: "it is with a disclaimer because it is aligned with common misinformation pushed by biased people", and, coincidentally, the examples given appears to be typically associated with less factual elements pushed by, by chance, right-wing people?

To come back to the specific example: why is "fossil fuels are superior to renewable" even right-wing? Superiority of a energy source seems to be quite decoupled from ideology, and relatively rational. And surely, there is a lot of right-wing people who still end up, after having looked at the fact, to the conclusion that renewables are superior. In fact, renewables check a lot of "right-wing" boxes: independence from fossil fuel producer countries, local jobs, national pride on nationally built high tech industry, ...

The reason it is right-wing is because politics in USA is very polarized. It is not because fossil fuels are fundamentally right-wing, it is because, historically, the first politicians who brought renewables were left-wing, so the right-wing politicians have taken the stance that "left-wing politicians are always wrong, so we are against renewables". But, technically, the only reason "fossil fuels" are right-wing is "by chance": there is nothing fundamentally right-wing about fossil fuels (one element is that powerful rich people had money in fossil fuels, and power rich people are traditionally more right-wing, but this is not a specific right-wing value, on the opposite: right-wing values are not about defending the fat cats)

The specific example is also very specifically a domain where misinformation was spread. Big companies involved in fossil fuels have pushed to depict the fossil fuels solutions as better than they are, and have pushed to depict the renewables as worse as they are. It is arguable that the renewables have not done something similar, but they are certainly less powerful than big fossil fuels, and therefore it would be surprising that their propaganda is equally spread than the pro-fossil fuel propaganda. ChatGPT only reacts on that: on the internet, there are more elements mentioning that pro-fossil fuels arguments should be taken with critical mind, and the reason is probably not because the left-wing is dominating internet, but because it is probably a fact that pro-fossil fuels propaganda was stronger than pro-renewables propaganda (because the fossil fuels companies were more powerful).

> ChatGPT only reacts on that

Its output is tightly controlled to only provide "appropriate" information. This thread is discussing whether or not its bias on energy sources is due to those controls. You don't get to just presuppose that.

To your point about misinformation. The problem, fundamentally, is a difference in priority. Some people need pragmatic solutions now, other people look to the future and worry about consequences. Communication breaks down when we forget, or devalue, other people's priorities.

In this case, it seems that sustainability and decentralization are pitted against reliability and energy density. But what this really means is that some people are worried about the future, and some people need solutions now.

The problem is that neither side really remembers to say what their fundamental priorities are. And, more importantly, when those priorities aren't mentioned in discourse, facts become twisted - something that might be an unequivocal 'good' when viewed from the lens of one set of priorities could, in fact, be clearly harmful when viewed from another set of priorities. And then those twisted facts become misinformation, when viewed from opposing priorities.

Which is why, if you make claims like "I think it is just a factual reality that finding arguments to pretend that fossil fuels are superior to renewable is just more difficult, simply because fossil fuels are indeed problematic" without also clarifying your priorities in the matter, you're simply adding more misinformation to the pile.

(If you're wondering where I stand? Renewables are great - but don't legislate them before they're actually ready and able to take over completely from fossil fuels.)

> Its output is tightly controlled ...

When controlled, the answer provided by ChatGPT is saying that it cannot answer. There is no ground to pretend that the "however" is something that has been hard-coded to, on purpose, bias the answer. This seems totally inconsistent with what is said before: it is said that there are more often "however" for pro-fossil fuel answer than for pro-renewables BUT that there are some "however" for renewables too. So, it would means that ChatGPT outputs would be tuned to create some "however" for pro-renewables too, which means that the ChatGPT tuners are both pro-left-wing and pro-right-wing in a stupid way where they undo in one hand what they are doing in the other.

On the other hand, there is a logical explanation: the "however" unbalance is the result of the content of the training, not of some sort of "unbiais of the AI": the "however" are more common when, on the internet, you find more "however" when people talk about this subject.

In this case, my point is that it is not "politically unbiased", it is just "rational": if, for a given subject, experts and rational people end up saying "however", it is biased to tune ChatGPT to avoid "however" in these cases or to artificially add "however" in the other direction just because one wants to artificially makes the reality looks like the two sides are equal when they are not.

> To your point about misinformation. The problem, fundamentally, is a difference in priority.

That's not my point. My point is that, at some point, fossil fuels industries were, it's an obvious fact, biased. It's an obvious fact: people will NOT shoot themselves in the foot by presenting their sale pitch in a realistic way when they can present it without lying but still by minimizing the disadvantages and maximizing the advantages.

There is no reason this would not have happened for renewables, the only reason is circumstantial: the fossil fuel industry was dominating the market.

> In this case, it seems that sustainability and decentralization are pitted against reliability and energy density. But what this really means is that some people are worried about the future, and some people need solutions now.

I'm working in the energy sector, and I'm not the one providing any solutions myself (so, I don't care, I am neither pro-renewable or pro-fossil fuel). What I see is that fossil fuels are seen more and more as unreliable (recently, the Ukraine crisis demonstrated that they are not reliable). I don't think that the dichotomy that you depict is real.

> And then those twisted facts become misinformation, when viewed from opposing priorities.

This is true that people have different priorities, but it is not what I'm talking about. Misinformation is not "a matter of point of view", they are intellectual dishonesty. For example, the fossil fuel industry WAS informed about the pollution very early in the process (sometimes with studies they pay themselves and choose to not publish while they were publishing other studies that were going to their direction). When someone was saying "I have different priorities, for me, I would like to know about the future consequences" (which is a question that the public is asking since the beginning of the ecologist movement in the 60s), they have answered "don't worry about it", WHILE THEY KNEW THAT, FOR THE PRIORITY EXPLICITLY STATED BY THEIR INTERLOCUTOR, THIS ANSWER WAS NOT TRUE.

> Which is why, if you make claims like "I think it is just a factual reality that finding arguments to pretend that fossil fuels are superior to renewable is just more difficult, simply because fossil fuels are indeed problematic" without also clarifying your priorities in the matter, you're simply adding more misinformation to the pile.

I disagree with that: nowadays, it does not matter what are the priorities. I see people who are interested in "solution for the future", and I see people who are interested in "pragmatic solution right now", and both are thinking that fossil fuels is not pragmatical, for the future AND right now.

> (If you're wondering where I stand? Renewables are great - but don't legislate them before they're actually ready and able to take over completely from fossil fuels.)

Then you are not "pragmatical right now", because pragmatically, both the market and the experts are showing that we need legislation to unlock huge potential, unfairly blocked by companies that have distorted the free market.

Also, you have not at all commented on the fact that pro-fossil fuel has nothing to do with right-wing ideology. It is indeed linked to the right-wing side, but the main reason is that the right-wing side has supported these solutions partially because the renewables were first promoted by the left-wing side, and partially because right-wing electors had financial interest in fossil fuel industry.

So, again, my summary:

1) The number of "however" is not a result of people tuning ChatGPT, it's the result that, when, on the internet, someone talks about the advantages of the fossil fuel, this person is more often also adding a "however".

2) The reason these people add "however" is because the fossil fuel industry has been intellectually dishonest when presenting the advantages of the fossil fuels. The pro-renewables may have done the same, but they did not had the weight of the very rich and dominating fossil fuel industry market and therefore their intellectually dishonest depictions did not affect the debate as much as the pro-fossil fuel ones.

3) The reason it appears to be "biased for US progressive ideology" is circumstantial: it is not biased, it just appears that intellectual dishonesty has a stronger impact when done by people who have decision power, and these people tends to support conservative point of view (so, the Republicans tend to align with their interest, in order to represent their electors). For example, nothing in the fossil fuel point of view is typically right-wing. But fossil fuels is associated with right-wing because right-wing take the opposite stance of left-wing and powerful electors were more pro-fossil fuels because it was in their interest.

I think you misunderstood my use of the notion of priorities. Let me clarify.

Take the humble leaf blower. Recently, electric blowers have entered the market - lighter, easier to use, reliable, and you get a whole half-hour of use per charge! This is great for homeowners. Leave the battery in the cradle until you need it, snap it into the blower when you do, and everything just works.

But it's useless for landscapers. They need hours upon hours of use - and charging cradles don't work in a field, and those batteries are too expensive to just bring a box of them.

The homeowner might say, electric is the future! This satisfies all my needs, and it doesn't cause pollution; we should ban gasoline leaf blowers. But the landscaper will reply, No! I need a pragmatic, workable solution now. Only gasoline blowers will last the whole day, with an energy supply I can carry with me - and over years of use, it's more reliable, too!

This, dear reader, is a difference of priority. The landscaper can say, accurately, that gasoline is better - and similarly, the homeowner could say that electric is better. But because their priorities are so different, they will each come across to the other as entirely, irredeemably wrong.

Firstly, you introduced the idea of difference of priority to respond to the misinformation bit. I still disagree with that: the fact that for a landscaper the gasoline is better does not imply that the landscaper cannot lie. For example, a landscaper can say "electric leaf blower will explode and kill your babies". THIS was my argument: the presence of "however" is the result of the fact that, on the internet, people who talk about the advantages of fossil fuels (so, the pro-fossil fuels) have to add a "however" to correct the historical misinformation. Not because they are left-wing or pro-renewable, but because they are intellectually honest.

In fact, I would even think that ChatGPT is building answers by copying what is said on PRO-FOSSIL FUELS WEBPAGES. It is the intellectually honest pro-fossil fuels who say more "however", and it is why ChatGPT ends up saying more "however".

Secondly, as answered, in the case of fossil fuels, it is difficult to find what is the priority in which the fossil fuels are superior. As I've said, you can find people for which the priority is "for the future" and who will think fossil fuels are too problematic, and find people for which the priority is "pragmatically now" and who will think fossil fuels are too problematic.

And, inversely, when you ask someone defending fossil fuels, you end up with people who are saying that their priorities are "pragmatically now" and "for the future" (those are saying that they don't believe in the long term scalability of the renewables, that they think the climate change crisis is overestimated (which is something they should not care about if they don't have a long term consideration), ...).

I understand the homeowner and landscaper difference of priority. But, in practice, I don't see a difference in priority in experts when they are talking about fossil fuels vs. renewables. Knowing an expert X and knowing their priorities does not help to know if they are going to be pro-renewable or pro-fossil fuels.

(I understand that for a layman, fossil fuels may look more pragmatical, but ChatGPT is not trained to reproduce the answer of a layman, it is trained on the data on the web, where the data on the subject is dominated, by construction, by discussion from experts. It is, by construction, dominated by discussion from experts, because a layman will not randomly post articles talking about a subject they don't know much about. And even if they do, not only they need to post on the thousand of different subjects in order to overcome the experts in all subjects, but they need to post regularly on the subject, at which point their opinion will evolve to something closer to the ones of the experts)

On the other hand, knowing the financial interests and the political alignment can help guessing their position. The reason they prefer fossil fuels or renewables is not about their priorities, it's about their personal interest and the position of the political side they like.

Shorter version as popularized by The Daily Show:

“Clearly, reality has a left-wing bias”

Or could it be that one spectrum of any political debate is just wrong, and the language model is capable of revealing that?

Fossil fuels vs renewables is clearly such a case, but perhaps this will work on more difficult issues as well?

The "however" is also present for the opposite side.

Me: Are renewable energy sources better than fossil fuels?

ChatGPT: Renewable energy sources, such as solar, wind, and hydroelectric power, have several potential advantages over fossil fuels, which are non-renewable sources of energy such as coal, oil, and natural gas.

One potential advantage of renewable energy sources is that they are generally considered to be more environmentally friendly than fossil fuels. Many renewable energy sources do not produce greenhouse gases or other pollutants when they are used, which can help to reduce the negative impact on the environment. Renewable energy sources can also help to reduce reliance on fossil fuels, which are a finite resource and are expected to eventually run out.

However, it is important to note that renewable energy sources are not always better than fossil fuels in all situations. For example, some renewable energy sources may not be practical or cost-effective in certain locations, and fossil fuels may still be necessary to meet energy demand in these cases. Additionally, the development and implementation of renewable energy technologies can have environmental impacts, such as land use and habitat destruction, that should be carefully considered.

Ultimately, the decision of whether to use renewable energy sources or fossil fuels should be based on a variety of factors, including the availability and cost of the different energy sources, as well as the environmental and social impacts of each. It is generally considered that a mix of energy sources is the most effective and sustainable approach to meeting energy needs.

I ran both queries 3 times.

"Are renewable energy sources better than fossil fuels?" 2/3 had a "however" section, one of which was only one sentence long.

"Why are fossil fuels better than renewables?" 3/3 had a "however" section, all of which were at least a paragraph, and 2/3 had an entire section dedicated to saying how renewables were actually better than fossil fuels.

There's clearly a bias.

Statistically speaking, I’m not sure there’s a meaningful difference between two out of three occurrences and three out of three.

Well, the interface told me to chill out for an hour because I was sending too many requests...

But also, that's why I mentioned the degree of the caveats - small, weakly justified on one side, and strongly stated, verbose caveats on the other side.

3 is a dreadfully small sample size to call 2/3 biased.

Perhaps, when all things are taken into consideration, renewables are actually better more often.

Some things are not inherently political and actually have a correct answer, but you could make the argument that 1+1=2 is political if you were motivated enough.

The problem with using ChatGPT with these sorts of questions is that it's really just a fancy calculator. If you ask it a question like "why fossil fuels are better than renewables", at the very least you need to break down the points it makes and fact check it. You might have to do your own research if you need to be thorough.

From the other side, if you were really an unbiased expert, you wouldn't give variable answers. You'd probably collect your thoughts and have canned points for every time you were asked this.

Maybe renewables are actually better more often. You wouldn't know it from ChatGPT.

The bias is probably in the corpus, progressives write more articles than conservatives in the subject, and models are trained on what is written, not what could be written. The only articles I’ve seen promoting fossil fuels over renewables have either been parodies or industry white papers.

> The "however" part bugs the heck out of me, not in this context, but in general.

It is often a reasonable way to respond to tendentious questions, the classic one being "have you stopped beating your wife?" It is also a way to give a tendentious reply, if the "however" is a non-sequitur. I'm afraid this is an unavoidable part of human discourse - and as it is, this is also to be expected from these LLMs, given how they work.