Will ai become an instrument to push American/western centric views to other countries?

I have been playing with bing chat, and been really impressed about how Human like it is.

But i have also been disappointed about a couple of things.

1. it acts literally like a mad man, endorsing completely opposite views of things and giving opposite answers to the same questions in a Matter of minutes without remember its original answers.

2. It very often endorses western views of social topics and more specifically urban liberal western views. It does so through 2 mechanisms:

The first one is that it mostly uses English speaking sources regardless of which lenguage you use to speak to it (i have spoken to it in English, Spanish, Portuguese, Chinese and Japanese and it always uses English speaking sources and then translates them to the language you use ).

In this case, even if the langua

ge model doesn't directly express an opinion, it reflects the views of western English speakers the most.

Second, sometimes it will actually express "opinions". Once i was asking it about a socially sensitive topic and it told me it's opinion about it. It was based on an article from the new York time that provided some anecdotal evidence about the topic. I asked if it had statistically relevant opinion to back that opinon, but it told me he didn't have and didn't need it, for some things, numbers were not necessary.

I was shocked to hear this and told it, you are language model, you don't have any opinions, you are just reproducing opinions from open ai employees. He told me, no i am taking opinions from the users i interact with and my users prefer me to have opinions.

Long story short, ai social opinion was extremely American and very different from how a latin American would see things.

The issue with it, is that this will gradually become a work instrument, and a writer assistant for journalists and small media around the world.

Once it stops being crazy it will just more consistently endorse it's creators views and will do so from an apparent air of objectivity being that it's basically a non emotional computer.

All your base are belong to us!

@mattbbpl: for great justice

@mattbbpl: you mean my English is not correct?

@nirgal: Your English is fine, buddy. That phrase is a classic meme based on a bad translation in a video game. I thought it was ironic given the subject matter of both the OP and the quote.

@mattbbpl: oh all right, but I'm serious with what i am saying.

It feels lately that all social topics that start in the us tend to ripple to other countries.

I feel this tool will accentuate this trend.

I remember reading something about this, not necessarily about AI, but programming and sort of "default answers" to things is always more or less skewed towards white, Western, and male.

It's not a malicious or even conscious effort in most scenarios, it's just a lot of people making these things are either white, western, and male or educated by white western males, or things like that.

No, I am not trying to make this about being "woke" it's just something I've read about that said it's objectively identifiable. I'll try to link the article if I can find it.

It's one of the reasons women have such a tough time breaking into tech. Yes, there is of course some discrimination, but it's also systemically biased in favor of men.

It's like, if you design a car in a country where everyone is 3'10", your cars are going to be made for people that are 3'10". IF you want to export, you need to make extra effort to make cars for people larger than 3'10". Shitty example but I hope you can get what I am trying to say. It's harder for you to go outside your norm; likewise, it's harder for people to adapt to your norm

@mrbojangles25: i don't know, i wish this was not done on purpose it but it feels like it is.

And it could be solved simply by giving greater weight to not English speaking sources or at least adding a "western culture English speaker" bias disclaimer.

Even within English speaking countries people should be concerned about this, since it clearly favours Urban views over rural views on social issues.

Also since this software is being programmed by elite universities graduates from mostly the USA, it stands to assume that when the employees train the ai by favoring some output, their own culture and world views leak through.

@mrbojangles25: You've stretched the original claim a bit from "Western" to "white and male." One doesn't indicate the other.

If by western you mean broad based liberal democratic values then my answer is: yes, I hope so.

If you mean something else then you need to be clearer because pretty much everything else is going to be a niche view that is contraindicated by competing views.

@SUD123456: I bordered on asking, "is that a good thing?" Apparently some other people think so as well.

@nirgal: this will be why AI make terminators to kill us. Someone like Zariya will start calling the chatbot alt right and cry to the mods. Few weeks later T800s walking around

@SUD123456: i will give you an example.

I asked chat gpt to list it's sources.

It gave me a small list of all English speaking, center to left leaning sources.

I asked the criterion used for selecting sources. It said reputation, factuality and reliability

@SUD123456: i will give you an example.

I asked chat gpt to list it's sources.

It gave me a small list of all English speaking, center to left leaning sources.

I asked the criterion used for selecting sources. It said reputation, factuality and reliability.

So i gave it a lot of sources and asked it to classify them.

Among them were china daily and Russia today. The software identified them as state funded, bias, representative of their respective governments.

Among others i also asked them about DW news and the BCC. They were identified as factual, reliable and of great reputation.

I asked them why don't you also identify these sources as state backed and bias If they are funded by their respective states.

It said it considered that due to their better reputation, they were not bias.

I asked about how you can tell they have a better reputation. It said it's information it gets from it's sources.

So basically it uses wester sources that regard western sources as more reliable, to justify giving greater weight to western sources.

By no means i am saying that china daily and Russia today are great sources, but not identifying western state backed sources as such and using them as independent sources is very concerning.

@mattbbpl: it's a good thing only for those that agree with points of views held and pushed by the ai.

That is until they don't anymore.

That's because we both understand that it is the only broad values framework that can accommodate everyone and the reason it can do so is because it is flexible enough to change and make room for differences. Yes it can be messy at times but it is that mess that allows for difference and change.

I know you already understand this...but maybe someone else will read and pause to think.

Also, as much as I respect Rousseau, Hobbes, Locke, etc. we ought to annoint Gene Roddenberry as our symbolic overlord.

@SUD123456: the funny thing is that western values have changed, but the deeply ingrained sense of superiority of western values has not changed.

You have failed to understand my post.

I said if by western you mean liberal democratic values...

You respond with western. You create your own wind.

Lets try again. It is true that liberal democratic values largely originated in western thought. However, liberal democratic values exist in many countries around the world, not just those commonly associated with western countries.

Second, liberal democratic values are rooted in socio-political constructs that represent a balance between individuals and their liberty, the overall body politic, governance, and the means by which these forces interact.

In other words, the general balance of how society operates....not the time bound specifics of norms and behaviors, nor cultural specific attitudes. Diversity is a perfect example of how the liberal democratic governance system and society can deal with change much more readily than any other.

You obviously confuse mere differences of cultural and social perspectives (whether that is timebound, location or ethnicity based) with that of the fundamental underpinnings of socio-political structures

The opposite of liberal democratic values is authoritarianism.

The beauty of liberal democratic values is that they can accommodate all sorts of people with all sorts of different attributes. There is a reason why oppressed people try to escape their shit reality to liberal democratic countries.

So, I think you should be more careful, make fewer assumptions, and don't let your ethnocentric hangups get in the way.

As to your original question I think your other post clarifies a bit. I have never interacted with AI but I imagine that the language it is constructed in and the attributes of the programmers likely have some relationship to bias but I don't see how that is different than interacting with a human.

The original claim of the post or the article I was referencing?

In either case, I know I did, I was just using it as an example of how bias can be systemic based on who is writing the rules. I guess maybe it wasn't 100% relevant but whatever 😋

@mrbojangles25: I get it. I just wanted to point out "Western" is more diverse than what was painted there. As a classic example, Japan is considered Western.

how would a completely unbiased system even function if you're asking it opinion based questions?

"1. it acts literally like a mad man, endorsing completely opposite views of things and giving opposite answers to the same questions in a Matter of minutes without remember its original answers."

that seems like the only possible way it could act if it were truly without bias.

if they goal is to create a human-like system that learned from evidence and formulates ( or regurgitates ) thoughts and opinions based on its experience, then by default you're going to get bias because ALL people have a very limited scope of experiences when compared to the wide world.

maybe the systems will be improved so you can ask it to respond from different perspectives. like "what are your thoughts on the Ukraine war as farm worker from Angola?" or "as a auto mechanic in Singapore, how should the public transportation system in Madrid be run?"

@comp_atkins: well it most often just reproduces opinions from the sources it considers reliable.

After all, even though it gathers information from all around the internet, it doesn't weight it the same way .

That's why I was complaining about western bias. I imagine the open ai leadership is the one chosing the sources.

@SUD123456: look you are just referring to a more specific subset of western thought. It may exist to an extent in other countries (less than you think though), but it exists as result of media influence, it does not mean it has a joint origin in other countries.

well, that's part of the problem with any of these systems, particularly the ones generated by corporations out to make money. there is little incentive or motivation to disclose how they're trained and on what data. secret sauce and all.

@nirgal: hopefully people are smart enough to not trust robots on everything.

Probably not tho lol yeah it's concerning but I'm sure China will have it's own AI bots as well. So will be biased AI depending on who created it.

Basically how we have media now with humans from different areas. The AI will have it's bias.