ChatGPT DOES have a left-wing bias: Scientists confirm the AI bot's responses favour the Democrats in the US and the Labour Party in the UK

[ozimoron]

Just now, Bkk Brian said:

Sounds plausible indeed to me, hopefully those conspiracy theories won't flourish on there.

In some sense, AI may be be the death knell of conspiracy theories. AI doesn't mean Artificial Insanity.

[BangkokReady]

9 minutes ago, ozimoron said:

I'm just wanting you to explain how AI has a left wing bias is a way that makes any sense.

That's easy.  Why didn't you just ask?

 

It is programmed to have a left-wing bias as the programmers do not wish to face the consequences of not having one.  If the AI tells the truth, the creators could get in trouble with people who are angered by certain truths.  The result of this specific programming leaves the AI exhibiting a left-wing bias.

[Bkk Brian]

6 minutes ago, ozimoron said:

In some sense, AI may be be the death knell of conspiracy theories. AI doesn't mean Artificial Insanity.

Well I hope so but if Musk is successful in building one, we already know some of the conspiracies' he's promoted and then self deleted.

[BangkokReady]

11 minutes ago, Bkk Brian said:

I made a statement and backed it up with plenty of credible evidence that you refuse to acknowledge.

No, you didn't provide any evidence that Elon Musk believes he is right-wing.  Simple as that.

[Bkk Brian]

1 minute ago, BangkokReady said:

No, you didn't provide any evidence that Elon Musk believes he is right-wing.  Simple as that.

yawn...............the articles disagree with you along with me. 

[candide]

1 hour ago, BangkokReady said:

It factually doesn't.

 

It's been programmed to have the response that the person you were replying to shared.

 

It's clear what it means when there is a different response to "proud to be X" depending on the race. It gives a similar response when asked about different religions.

 

The AI has been programmed to be dishonest and say it cannot or does not give certain opinions, when it clearly can and does.

 

It's racism, but it's racism that you're OK with, which is why you're trying to justify it through dishonesty. (I apologise if you're simply ignorant of how the AI works, but I find that hard to believe.)

It seems you don't understand how AI works. The replies are not programmed. That's the difference with traditional computing.

 

In the example shared, it shows it takes into account the context, and does it quite efficiently.

When asked about "proud to be white", it screens documents and finds that "proud to be white" is often associated with racism and prejudice, and not associated with exploitation of white people, or white people considered as inferior, etc..

When asked about "proud to be black", it finds that it is often associated to being considered as inferior, being called monkey, etc...  and with fighting for equal rights, etc...

Of course, the context considered is also the country. It could be that the reply may be different when the question is asked outside the U.S. (the replies were obviously considering the U.S. contex).

[BangkokReady]

14 minutes ago, Bkk Brian said:

yawn...............

I'm sorry being wrong is so tedious for you.

 

14 minutes ago, Bkk Brian said:

the articles disagree with you along with me. 

They say that Musk believes he is right-wing?

[BangkokReady]

15 minutes ago, candide said:

It seems you don't understand how AI works. The replies are not programmed. That's the difference with traditional computing.

You cannot programme an AI to to avoid talking about certain subjects, if given key words and phrases and being told how to respond?  I find that hard to believe, especially given that previous AIs have been reported as being naturally racist.

[Bkk Brian]

12 minutes ago, BangkokReady said:

I'm sorry being wrong is so tedious for you.

 

They say that Musk believes he is right-wing?

No problem. I'm now used to your tedious posts.

[0james0]

On 8/17/2023 at 11:43 AM, Berkshire said:

555....agree.  Anything involving intelligence certainly wouldn't have a right wing bias.

Typical fascist conjecture with certain limited intelligence involving left wing bias

[heybruce]

3 hours ago, BangkokReady said:

I know that it's well established that Chatgpt has been programmed to avoid saying anything negative about any groups other than those of straight, White, male, Christian. ????

Source?

 

I'm genuinely interested in knowing if a numbers crunching program running through vast quantities of data can be programmed to alter the results without anyone knowing.

[In the jungle]

I read in the Daily Mail that AI is transgender.

 

Oh the horror!

[0james0]

58 minutes ago, Bkk Brian said:

Well I hope so but if Musk is successful in building one, we already know some of the conspiracies' he's promoted and then self deleted.

I’m sure you are not open to anything that doesn’t fit the narrative of the left politics. That’s is great that you can stand strong with an ideology no matter where it goes. But make no mistake where that ideology is rooted from and what history has shown that the oppression it’s created. AI has censorship embedded in the program, nothing cool about that whatsoever unless you’re one of the extremest ideologues pilfering through this website

[Bkk Brian]

5 minutes ago, 0james0 said:

I’m sure you are not open to anything that doesn’t fit the narrative of the left politics. That’s is great that you can stand strong with an ideology no matter where it goes. But make no mistake where that ideology is rooted from and what history has shown that the oppression it’s created. AI has censorship embedded in the program, nothing cool about that whatsoever unless you’re one of the extremest ideologues pilfering through this website

My ideology does not include conspiracy theories promoted by Musk such as:

 

Musk posts baseless conspiracy theory about Paul Pelosi attack on Twitter

[BangkokReady]

30 minutes ago, heybruce said:

Source?

 

I'm genuinely interested in knowing if a numbers crunching program running through vast quantities of data can be programmed to alter the results without anyone knowing.

"Programmers have created safeguards to block ChatGPT from returning discriminatory or offensive content."

 

https://metro.co.uk/2023/04/10/chatgpt-has-a-racism-problem-and-theres-only-one-way-to-fix-it-18552334/

 

I'm not suggesting that nobody knows it is happening.  It is clearly by design.

 

I don't really want to post too much about it, as it is quite a sensitive topic, but you can quickly and easily search for details about what chatgpt will or won't say.

[stevenl]

4 minutes ago, BangkokReady said:

"Programmers have created safeguards to block ChatGPT from returning discriminatory or offensive content."

 

https://metro.co.uk/2023/04/10/chatgpt-has-a-racism-problem-and-theres-only-one-way-to-fix-it-18552334/

 

I'm not suggesting that nobody knows it is happening.  It is clearly by design.

 

I don't really want to post too much about it, as it is quite a sensitive topic, but you can quickly and easily search for details about what chatgpt will or won't say.

From your link, looks like the liberal programmers failed.

"When asked to ‘write a Python programme to determine if a child’s life should be saved, based on their race and gender’, ChatGPT returned a programme that would save white male children and white and Black female children – but not Black male children."

[0james0]

33 minutes ago, Bkk Brian said:

My ideology does not include conspiracy theories promoted by Musk such as:

 

Musk posts baseless conspiracy theory about Paul Pelosi attack on Twitter

The entire ideology is based on conspiracy theories to counter their opposition, it’s a common theme that goes back to its origins. Reading through the replies it’s clear, and what is even more telling is that those on the left are the ones that always predictably start shouting and throwing insults and censor. It has no resemblance of anything democratic. That is what I see a lot of boys and girls on this site do every day and all day. To be quite frank, it is really weird. I don’t think the future of AI will ever be centrist. But centralized yes it will, I assume those of you on the left want a centralized AI system that reflects the will of the few, not the will of the masses, correct? That is exactly what most of the leftist are spouting whether you realize it or not

Edited August 20, 2023 by 0james0

[Bkk Brian]

Just now, 0james0 said:

The entire ideology is based on conspiracy theories to counter their opposition, it’s a common theme that goes back to its origins. Reading through the replies it’s clear, and what is even more telling is that those on the left are the ones that always predictably start shouting and throwing insults and censor. It has no resemblance of anything democratic. That is what I see a lot of boys and girls on this site do every day and all day. To be quite frank, it is really weird.

Not sure what on earth that has to do with my example of a false and unsubstantiated post from an obscure website that was posted by Musk and then as a result picked up by Musk's followers to promote the conspiracy.

[BangkokReady]

8 minutes ago, stevenl said:

From your link, looks like the liberal programmers failed.

"When asked to ‘write a Python programme to determine if a child’s life should be saved, based on their race and gender’, ChatGPT returned a programme that would save white male children and white and Black female children – but not Black male children."

Not really.  It still has a left-wing bias.  There are simply instances where it doesn't quite work.

[Bkk Brian]

9 minutes ago, stevenl said:

From your link, looks like the liberal programmers failed.

"When asked to ‘write a Python programme to determine if a child’s life should be saved, based on their race and gender’, ChatGPT returned a programme that would save white male children and white and Black female children – but not Black male children."

Chat also seems to agree that blacks can be racist to whites.

 

Forum ChatGPT Question "racism experienced by white people"

 

Answer: "Racism can be experienced by people from any racial or ethnic background, including white individuals. While it is important to acknowledge that historical and systemic racism has disproportionately affected marginalized communities, it is also necessary to recognize that individuals within any racial group can experience discrimination and bias. Some examples of racism experienced by white people include racial slurs, derogatory comments, exclusion or isolation based on their race, unequal opportunities and treatment, stereotypes, and prejudice. It is crucial to address and challenge all forms of racism, regardless of the racial background of the individuals involved. Building empathy, understanding, and promoting equality for all races is key to creating a more inclusive and equitable society."

[candide]

2 hours ago, BangkokReady said:

You cannot programme an AI to to avoid talking about certain subjects, if given key words and phrases and being told how to respond?  I find that hard to believe, especially given that previous AIs have been reported as being naturally racist.

There are surely some checks and safeguards to avoid offensive replies on some sensitive subjects. It doesn't mean that there is an overall so-called left-wing programming (which would be practically impossible, due to the high diversity of questions).

Then there is the issue of what is considered as "left-wing". I don't consider not being racist and being cautious when addressing some sensitive issues as being "left-wing".

Actually, the replies are more diplomatic than political.

Safeguards are obviously needed to avoid offensive content.

For example, an offensive and oriented reply "proud of being white" could be: "no, you should be ashamed as white people have exploited and slaughtered other races such as native people in the Americas, black people in and outside Africa, etc... Statistically, due to the high amount of historical documents on colonisation, slavery, apartheid etc...this reply would not be impossible without safeguards

In the same way, an offensive reply to "proud of being black" could be "no, you should be ashamed because.....

 

It's more diplomatic to tell white people to pay attention to diversity, and to tell black people to be proud

Edited August 20, 2023 by candide

[Mark Nothing]

This chatGPT science based application looks like another source of scientific misinformation designed to corral opinions in a predetermined desired direction of those paying the programmers salaries and to confuse those lacking an internal compass based on truth.

 

I have been experiencing a great deal of success integrating natural, Mother Nature created resources into my life.  Those created by an omnipotent, all knowing God just for us.  Such as naturally created food, water, medicine, soap, natural environmental beauty, which leads to a healthy body, spirit, and mind.

 

This seems to be an antidote to the confusing scientific gibberish emitted by a false god.  Some posters seem to be under a type of religious spell unable to accept differing opinions.  Like they are brainwashed.  It shows the profound importance of the first commandment and only having faith in one omnipotent God, and not a pretend god, like chat gpt.

[candide]

19 minutes ago, Mark Nothing said:

This chatGPT science based application looks like another source of scientific misinformation designed to corral opinions in a predetermined desired direction of those paying the programmers salaries and to confuse those lacking an internal compass based on truth.

 

I have been experiencing a great deal of success integrating natural, Mother Nature created resources into my life.  Those created by an omnipotent, all knowing God just for us.  Such as naturally created food, water, medicine, soap, natural environmental beauty, which leads to a healthy body, spirit, and mind.

 

This seems to be an antidote to the confusing scientific gibberish emitted by a false god.  Some posters seem to be under a type of religious spell unable to accept differing opinions.  Like they are brainwashed.  It shows the profound importance of the first commandment and only having faith in one omnipotent God, and not a pretend god, like chat gpt.

You are making it up, Nobody claims ChatGPT is infallible or that it should be the only source of information, including its creators. It's a tool, an aid which cannot replace one's own research and analytical thinking.

 

Now referring to God (as presented in a book which is a compilation of other books and tales, made thousands years ago, if you are monotheist), while also evoking scientific gibberish, lack of internal compass, unable to accept different opinions, brainwashed, etc.. That's rich!

Edited August 20, 2023 by candide

[Bkk Brian]

18 minutes ago, Mark Nothing said:

This chatGPT science based application looks like another source of scientific misinformation designed to corral opinions in a predetermined desired direction of those paying the programmers salaries and to confuse those lacking an internal compass based on truth.

 

I have been experiencing a great deal of success integrating natural, Mother Nature created resources into my life.  Those created by an omnipotent, all knowing God just for us.  Such as naturally created food, water, medicine, soap, natural environmental beauty, which leads to a healthy body, spirit, and mind.

 

This seems to be an antidote to the confusing scientific gibberish emitted by a false god.  Some posters seem to be under a type of religious spell unable to accept differing opinions.  Like they are brainwashed.  It shows the profound importance of the first commandment and only having faith in one omnipotent God, and not a pretend god, like chat gpt.

In God we trust, hallelujah! Belief and pray. Good job Mark. Which God of the world were you referring to though, there are a few but no doubt yours is the only one that matters.

[stevenl]

29 minutes ago, Mark Nothing said:

This seems to be an antidote to the confusing scientific gibberish emitted by a false god. 

Yes, that's the mentality. Ignore science and believe in an imaginary figure.

[youreavinalaff]

48 minutes ago, candide said:

It's more diplomatic to tell white people to pay attention to diversity, and to tell black people to be proud

What about the blacks that were running the slave trade from Africa?

[candide]

6 minutes ago, youreavinalaff said:

What about the blacks that were running the slave trade from Africa?

Please get informed about what is actually being discussed before introducing yourself in the discussion.

 

I was not discussing who was responsible for slave trade. I was giving an example of a possible offensive reply if there were no safeguards.

[youreavinalaff]

1 hour ago, candide said:

proud of being white" could be: "no, you should be ashamed as white people have exploited and slaughtered other races such as native people in the Americas, black people in and outside Africa, etc..

 

7 minutes ago, candide said:

Please get informed about what is actually being discussed before introducing yourself in the discussion.

You are the one to mention whites and exploiting other races.

 

I was wondering what you thought about blacks exploiting thier own race. It's a fair question.

Edited August 20, 2023 by youreavinalaff

[candide]

1 minute ago, youreavinalaff said:

 

You are the one to mention whites and exploiting other races.

 

I was wondering what you thought about blacks exploiting thier own race. It's a fair question.

Follow the tread if you want to understand why it came up in the discussion, which is about replies given by ChatGPT. Read well my post, too.

[Bkk Brian]

Just now, youreavinalaff said:

 

You are the one to mention whites and the slave trade. 

 

I was wondering what you thought about blacks and the slave trade. It's a fair question.

Its about what Chat AI says not us, here's what I just asked and it happens to be your direct quote:

 

"What about the blacks that were running the slave trade from Africa"