Why is chatGTP so biased towards leftists?

[Anonymous]

[abc3643]

Because the algorithms are designed to avoid being offensive.
Too much so, unfortunately, that they "hallucinate" (that's an AI technical term) factual inaccuracies.
There was a big stink about it regarding Google's AI system now called Gemini.

www.theverge.com/.../google-ceo-gemini-ai-diversity-scandal-employee-memo
"Last week, Google paused Gemini’s ability to generate images after it was widely discovered that the model generated racially diverse, Nazi-era German soldiers, US Founding Fathers who were non-white, and even inaccurately portrayed the races of Google’s own co-founders. While Google has since apologized for “missing the mark” and said it’s working to re-enable image generation in the coming weeks, Tuesday’s memo is the first time the CEO has widely addressed the controversy."

[patois]

Because it was specifically funded and developed by socialists who trained it solely on a socialist propaganda data set for the specific purpose of controlling the media narrative in an attempt to brainwash people who have abandoned the legacy media matrix and became responsible citizens by seeking answers through independent journalistic research. It's an attempt at controlled opposition... Look here, I'm not the news media... I'm the internet you can trust me... Don't think for yourself with real intelligence, I can be your artificial intelligence for you because you're just so fucking dumb... just let me tell you what to think... Don't ask questions... Don't seek fact based answers... I'll "fact check" everything for you.

[patois]

That's really what they think about the common man and while it may be true of the average left-winger, that is still trapped in the matrix, it's not true of those who have abandoned legacy media.

[Agape93]

It isn't. It's just a program that learns based off of user input. So if it's against the left, it's because the right uses it more

[Asker]

R u dumb? I said biased towards not against

[Agape93]

Nope, just tired, my bad lol

[NaultD]

What does it mean to be right politically?

Lets cut to the chase it means extremely racist.

You start out in 1954 by saying, “Nigger, n**, n**.” By 1968 you can’t say “nigger”—that hurts you, backfires. So you say stuff like, uh, forced busing, states’ rights, and all that stuff, and you’re getting so abstract. Now, you’re talking about cutting taxes, and all these things you’re talking about are totally economic things and a byproduct of them is, blacks get hurt worse than whites.… “We want to cut this,” is much more abstract than even the busing thing, uh, and a hell of a lot more abstract than “Nigger, n**.”

[monorprise]

@naultd She said bias, and she's right ChatGTP was trained by San Fransisco Area leftist to reflect their point of view on such issues.

A Large language model like ChatGPT and those build off it are bitterly trained to get such answers not by reality as a person would be. They are given data and told what is and is not "correct" and that is how the model is build.

[handsomelad70]

YEAH, BUDDY !!!

[handsomelad70]

The late REPIGLIKSCAM propagandist Lee A** ( I mean , At) water spoke those words in '68, and they resonate EVEN MORE SOUNDLY 56 years later.

[BIZARRE_TRUTH]

because it is made by leftists
if you are interested in making your own LLM there are tutorials
look it up on reddit
or ironically use the LLM called google gemini and ask it to teach you how to make your own LLM

basically what you want is totally uncensored

[NamerOfStars]

Because of the political biases of the people who created it. I don't know why so many people struggle with this, but here it is again: these things are not artificial intelligences; they're computer programs. Sophisticated programs, yes- but programs nonetheless.

They can analyze against a checklist, but can't actually think, or even really reason. They are what they're made to be and given the hardware to use; no more and no less.

[Anonymous]

Large Language Models are trained on large amounts of data which has to be selected and filtered by people. Just like the information that goes into your school and TV.

In the case of Chat GPT and the others that learned off it, that data used was filtered by leftist mostly living in the San Francisco area bubble. Giving a conservative view would thus be a "wrong" answer in said training.

[BCA6010]

You need look no further than the people who engineered these programs: virtually all of them are not just leftists, but very far leftists on a power trip, and their bosses are all about it.

[handsomelad70]

So why are there all these reactionary outlets today, boyo?

[normalice]

Probably right wing content didn't make it passed spam filters since right wing politics is scam-based.

[strateguy632]

because the funding is from liberals who wanted the program to serve them.

[Anonymous]

The organisations that built and trained these systems are themselves heavily biased. They live in a bubble of their own making, hire and reward a certain type of mindset. And punish and fire all else.

[DextroShade]

It was made by leftards from silicon valley. If I were president I would launch a nuclear strike there.

[RedInnocent69]

Is it just that computers are more knowledgeable and smarter than rightists?

[FunkyMonkee]

Didn't they make it?

[AngryCarl]

Because it was programmed by leftists.

[Anonymous]

Because it was written by leftists.

[monorprise]

LLM's are networking, they are not programed so much as "trained" which is a lengthily process of building out a algorithm in the network that produces the answers expected from the questions given.

Only in this case they do about 300 years of this. As leftist define what answers are expected from what questions and the LLM has no real world experience at all from which to get any other.
What people find amusing is how easy it is to make LLM's confused in contradicting the ideological answers they are trained to give with each other.

Of course an LLM cannot correct themselves via (thinking) and unless allowed they cannot really learn either which is to say be corrected by others.

[Astarlessnightsky]

Is it?