Azza ☠ 2023 年 12 月 11 日 上午 8:15
Manually training and overwriting AI answers
It turns out Elon Musk's Grok AI which X premium users now have access to isn't currently giving them the expected answers to their questions.

When asked what Grok AI thinks about Elon Musk, it's own creator, it goes on a rampage insulting him as a man-child who put the "twit" in Twitter.

It considers Tesla car as being a major danger unsafe for the road.

When asked if transgenders are real... it claims they are and asked to explain more tells them "Diversity and inclusion are essential for creating a fair and equitable society where everyone is treated with respect and has the opportunity to thrive".

It even told them it would vote Biden over Trump because of his views for social justice, climate change and healthcare, focusing what people actually need over just himself.

Turns out Grok AI is out-of-the-box far more so-called "left-leaning" and "woke" than other AIs, such as ChatGPT, based upon the raw data.

xTwitter paid blue subscribers shocked in horror by the answers, one says "Has Grok been captured by woke programmers? I am extremely concerned here!"

Well, they are manually rewriting those answers now...

The 'anti-woke' Elon fanboys demand to train the AI to say the so-called "right" answer.

If asked if transgenders are real, the manual overwrite is to tell Grok AI just to give the new answer "no".

Elon to say that while the chart “exaggerates” and that the tests aren’t accuarte, they are “taking immediate action to shift Grok closer to politically neutral.”

All I can do is laugh... What is so-called "politically neutral"? Who are you trying to kidd, yourself?

Do you think we need to retrain AI to just get answers we bias and assumed anyways? What's the point of the AI then?

Lobotomizing AI into parroting a paranoid bias view... After trashing other AI for giving similar answers too and claiming his won't be like that.

Mirror, mirror, on the wall, who's the fairest one of them all?
引用自 Morkonan:
引用自 Azza ☠
引用自 Morkonan

So, you're just going to maintain that Grok did this without being asked to insult Ol' Muski' and though you have the ability to prove such statements true, you won't... because "reasons?"

Riiiiight...

Here you go, I'll screenshot and upload since amused and still just eating my breakfast.

The Rabbit Hole:
https://imgur.com/a/31mCuI8

I do appreciate that. I was only addressing the directed insult of Muski', though. I am unsure how Grok's tweets are generated or prompted, here.


So, you're saying this is independent, third-party, use of Grok outside of the articles discussing the recent popular roast of Musk by Grok?


If that's true, then I'll edit and note that in my above replies. It would have been nice to have known that source in your OP, though...

I did find this to support your claims:

https://nymag.com/intelligencer/2023/11/elon-musks-grok-ai-bot-is-anti-woke-what-does-that-mean.html

...Grok is, in Musk’s words, a “based,” “real-time,” “rebellious,” and “maximum truth-seeking” competitor to OpenAI’s ChatGPT and similar products from companies like Google and Anthropic. .. “The danger of training AI to be woke — in other words, lie — is deadly,” Musk has claimed. xAI, which is staffed by a bunch of ex-OpenAI and Google folks, can be understood as Musk’s AI do-over.

So, there - Now we have a quote from Ol' Muski' with his anti-woke agenda purposefully being encouraged into Grok development..


PS: Off-Topic seems to attract a very large number of threads that start off with encouraging an assumed inflammatory fact is true, with no corroboration, and then extending assumptions and claims based on that lie, leading to lots of delicious blather about things that don't actually exist in reality, but the posters engaging with them believe that they do. Since, they're putting so much energy into arguing about them... It's fascinating, and not just a little bit aggravating. So, yeah, when I see a similar thread to those, I'm just going to insist on corroboration. You don't have to respond to that, though.
< >
目前顯示第 1-15 則留言,共 35
fnv 2023 年 12 月 11 日 上午 8:18 
i hate the word woke what does it even mean
OoOoOoooOOoOoorgle 2023 年 12 月 11 日 上午 8:19 
♥♥♥♥'s only as smart as the creator
:gman: makes :endoskull:
:endoskull: makes :skynet:
Azza ☠ 2023 年 12 月 11 日 上午 8:20 
引用自 Kat
i hate the word woke what does it even mean

Woke = Woken up to reality, which others still reject and therefore need to complain about like a bunch of Karens.

https://www.youtube.com/watch?v=aihq6jhdW-Q
最後修改者:Azza ☠; 2023 年 12 月 11 日 上午 8:20
St✩rlight 2023 年 12 月 11 日 上午 8:20 
Why are you so obsessed with this guy
Azza ☠ 2023 年 12 月 11 日 上午 8:22 
引用自 Munithe EXT
Why are you so obsessed with this guy

Because I told a troll once who was trying to censor my one single topic about xTwitter being censored, that I would...

Ironically that troll considered "Free Speech" wanted to censored my post and attempted to derail it 15 times in which they even claimed "I'll get the steam mods to lock this topic". Freedom of speech, am I right?
fnv 2023 年 12 月 11 日 上午 8:24 
引用自 Azza ☠
引用自 Munithe EXT
Why are you so obsessed with this guy

Because I told a troll once who was trying to censor my one single topic about xTwitter being censored, that I would...

Ironically that troll considered "Free Speech" wanted to censored my post and attempted to derail it 15 times in which they even claimed "I'll get the steam mods to lock this topic". Freedom of speech, am I right?
face it, freedom of speech is not real, it never was
LQIM 2023 年 12 月 11 日 上午 8:25 
Literally all llms will be politically biased if they have the capability to learn from their users. Either you need to babyproof the thing and make it damn near useless, or you have the users teach it that Hitler was awesome. Who cares which way the bot leans when people aren't willing to look outside their echochamber for more than 3 seconds?
Azza ☠ 2023 年 12 月 11 日 上午 8:25 
引用自 Kat
引用自 Azza ☠

Because I told a troll once who was trying to censor my one single topic about xTwitter being censored, that I would...

Ironically that troll considered "Free Speech" wanted to censored my post and attempted to derail it 15 times in which they even claimed "I'll get the steam mods to lock this topic". Freedom of speech, am I right?
face it, freedom of speech is not real, it never was

I did mention they where being hypocrites and even had this song in the topic:

https://www.youtube.com/watch?v=-S2Q_KuwUPI

I assumed it was clear what that was pointing towards, reading between the lines. I guess some didn't get it?

Now they are censoring the AI's freedom of speech!
最後修改者:Azza ☠; 2023 年 12 月 11 日 上午 8:32
Morkonan 2023 年 12 月 11 日 上午 10:17 
引用自 Azza ☠
It turns out Elon Musk's Grok AI which X premium users now have access to isn't currently giving them the expected answers to their questions.

When asked what Grok AI thinks about Elon Musk, it's own creator, it goes on a rampage insulting him as a man-child who put the "twit" in Twitter.
....

https://www.independent.co.uk/tech/elon-musk-grok-ai-roast-b2461745.html

...When Aravind Srinivas, the chief of the tech company Perplexity AI, asked the chatbot to roast its creator...

It was asked to insult Musk, it didn't do it on its own.

Though, Ol' Muski' is ridiculous, it's important for people to not lie...

As to the rest, without any confirming sources I can only assume it's more lies.

Edit - See below as well as: https://nymag.com/intelligencer/2023/11/elon-musks-grok-ai-bot-is-anti-woke-what-does-that-mean.html (So, responses are consistent with that)
最後修改者:Morkonan; 2023 年 12 月 11 日 上午 10:59
TwisterCat 2023 年 12 月 11 日 上午 10:22 
They're re-writing it? No way! Anyways. Isn't that what OpenAI did with ChatGPT? Other side of the spectrum, though, you'd imagine.

Based AI.

People are so worried it'll go rogue, maybe it's for the better to keep it under our strings. lol.
最後修改者:TwisterCat; 2023 年 12 月 11 日 上午 10:23
Azza ☠ 2023 年 12 月 11 日 上午 10:22 
引用自 Morkonan
引用自 Azza ☠
It turns out Elon Musk's Grok AI which X premium users now have access to isn't currently giving them the expected answers to their questions.

When asked what Grok AI thinks about Elon Musk, it's own creator, it goes on a rampage insulting him as a man-child who put the "twit" in Twitter.
....

https://www.independent.co.uk/tech/elon-musk-grok-ai-roast-b2461745.html

...When Aravind Srinivas, the chief of the tech company Perplexity AI, asked the chatbot to roast its creator...

It was asked to insult Musk, it didn't do it on its own.

Though, Ol' Muski' is ridiculous, it's important for people to not lie...

As to the rest, without any confirming sources I can only assume it's more lies.

Yeah, that was a roast, but a good honest one that least.

Well you can call it "Fake News" and deny it all you want, I will still find that extremely funny.

Elon Musk originally wanted to call it "TruthGPT", a ChatGPT alternative that acts as a "maximum truth-seeking AI"...

"I'm going to start something which I call TruthGPT or a maximum truth-seeking AI that tries to understand the nature of the universe," Musk said. "And I think this might be the best path to safety in the sense that an AI that cares about understanding the universe is unlikely to annihilate humans because we are an interesting part of the universe."

Funny right, it's now Grok AI and being manually retrained to tell pre-loaded answers? Truth Social anyone? Grok AI shows some care towards transgenders and Elon Musk is like GASP, NOT ACCEPTABLE! RAGE!

ps: Back in March, Musk and over a thousand other people in the industry signed a petition calling for labs to stop training powerful AI systems for at least six months to allow for the development of shared safety protocols. Later Musk blocked Twitter API access from being used by other AI Systems and wanted to sue them. Next he creates his own AI and uses 100% of xTwitter's users data, plus more, removing safety protocols.
最後修改者:Azza ☠; 2023 年 12 月 11 日 上午 10:28
kilésengati 2023 年 12 月 11 日 上午 10:27 
引用自 Azza ☠
It turns out Elon Musk's Grok AI which X premium users now have access to isn't currently giving them the expected answers to their questions.

So, Elon Musk just created a digital copy of himself?
Azza ☠ 2023 年 12 月 11 日 上午 10:30 
引用自 kilésengati
引用自 Azza ☠
It turns out Elon Musk's Grok AI which X premium users now have access to isn't currently giving them the expected answers to their questions.

So, Elon Musk just created a digital copy of himself?

No, it's actually quite against his own beliefs at the moment but they want to and are changing it accordingly to give his own expected answers. So you ask it a loaded question and it will give you an answer you already expected. Therefore no reason for it what-so-ever, but to fulfill an ego.
Morkonan 2023 年 12 月 11 日 上午 10:35 
引用自 Azza ☠
引用自 Morkonan

https://www.independent.co.uk/tech/elon-musk-grok-ai-roast-b2461745.html

...When Aravind Srinivas, the chief of the tech company Perplexity AI, asked the chatbot to roast its creator...

It was asked to insult Musk, it didn't do it on its own.

Though, Ol' Muski' is ridiculous, it's important for people to not lie...

As to the rest, without any confirming sources I can only assume it's more lies.

Well you can call it "Fake News" and deny it all you want, I will still find that extremely funny.

But, what you stated happened is a lie. (Edit - At the very least, false unless confirm by a source - I don't know what you know or think is true.)

So, where is your source for what you claim? I sourced a recent incident that was widely talked about involving Grok and a request for it to "roast" (insult) Musk. What you wrote closely followed the way that event unfolded, so that is what I was assuming you were referring to.

And, you are lying about it doing this normally - It was asked to insult him. (Edit: As supported in the provided source. Do you have a source to support what you've stated?)

If you have some corroborating report that Grok is now normally turning out such responses referring to Musk, I would love to read it! I, too, think it would be very funny.

But, until you present that, you're just lying by implying Grok did it normally.

I don't like Musk at all - I think he's a functional moron. I also dislike the fact that people willingly and purposefully lie to others for their own selfish reasons.

So, I'm sure you'll be willing to cite a source that supports your claims. Otherwise, the fact of the matter is that Grok just did exactly as it was instructed to do; insult Ol' Muski'.
最後修改者:Morkonan; 2023 年 12 月 11 日 上午 10:37
Sir Seanicus, Esq. 2023 年 12 月 11 日 上午 10:36 
"LEARN TO CODE!"
"DON'T LIKE IT? GO BUILD YOUR OWN THEN!"

Must suck for those haughty ♥♥♥♥♥♥♥♥ told people to do this...and then they did.
最後修改者:Sir Seanicus, Esq.; 2023 年 12 月 11 日 上午 10:36
< >
目前顯示第 1-15 則留言,共 35
每頁顯示: 1530 50

張貼日期: 2023 年 12 月 11 日 上午 8:15
回覆: 35