“I’m a Grok guy,” said Vance. “I think it’s the best. It’s also the least woke!”
What a fucking idiot. Really wants everyone to go back to sleep
Elon Musk’s Grok Goes Haywire, Boasts About Billionaire’s Pee-Drinking Skills and ‘Blowjob Prowess’
Submitted 3 weeks ago by RandAlThor@lemmy.ca to technology@lemmy.world
Comments
jjjalljs@ttrpg.network 3 weeks ago
bcgm3@lemmy.world 3 weeks ago
A third of a nation, who only recently labeled their political opponents as “sheep,” now vocally proud to be “socially asleep.”
The irony would be comedic, if the results weren’t so tragic.
frongt@lemmy.zip 3 weeks ago
Not haywire. It was specifically prompted to get that output.
sundray@lemmus.org 3 weeks ago
It was asked leading questions about embarrassing topics, but the over-the-top, fawning praise for Musk was baked into the responses without being specifically requested in the prompts.
andrewrgross@slrpnk.net 3 weeks ago
I think this is the main story. I don’t think it’s new info, but it confirms the issue persists: this LLM is so heavily trained to fawn over Musk that it doesn’t exercise any application of context or attempt to find truth.
Which is sad.
nondescripthandle@lemmy.dbzer0.com 3 weeks ago
It’s really just an affirmation machine for tech bros
jj4211@lemmy.world 3 weeks ago
Yes, just some people figuring out that Grok was steered toward ass-kissing Musk no matter what, and exploited that for funny output. So the takeaways are:
- Musk is such a loser he insisted on Grok kissing his ass
- LLMs remain stupid and have no idea what they are saying
Imgonnatrythis@sh.itjust.works 3 weeks ago
This headline makes it sound like someone has doubts about Musk’s pee drinking skills. That seems a little judgemental. Pretty sure Grok knows his master well.
floofloof@lemmy.ca 3 weeks ago
Pee drinking is somewhat impressive, but can he eat shit and die?
natecox@programming.dev 3 weeks ago
Of course he can’t, he’s too much of a loser to eat shit and die. He couldn’t manage it.
Eat shit? Maybe. But not eat shit and die? No sir. No way. I won’t believe it unless I see it for myself.
Robaque@feddit.it 3 weeks ago
Didn’t Grok also claim he’s fitter than LeBron and smarter than da Vinci? Perhaps being an olympic piss drinker has been the key to his success.
A_norny_mousse@feddit.org 3 weeks ago
Yes, different article, also on Not The Onion.
Didn’t read, but I suspect the LLM was prompted to reply in this way. Still funny, considering it is known that Musk tries to tweak it in his favor.
DandomRude@lemmy.world 3 weeks ago
Although Grok’s manipulation is so blatantly obvious, I don’t believe that most people will come to realize that those who control LLMs will naturally use this power to pursue their interests.
They will continue to use ChatGPT and so on uncritically and take everything at face value because it’s so nice and easy, overlooking or ignoring that their opinions, even their reality, are being manipulated by a few influential people.
Other companies are more subtle about it, but from OpenAI to MS, Google, and Anthropic, all cloud models are specifically designed to control people’s opinions—they are not objective, but the majority of users do not question them as they should, and that is what makes them so dangerous.
khepri@lemmy.world 3 weeks ago
It’s why I trust my random unauditable chinese matrix soup over my random unauditable american matrix soup frankly
DandomRude@lemmy.world 3 weeks ago
You mean Deepseek on a local device?
RaoulDook@lemmy.world 3 weeks ago
Trusting any of that shit is the problem.
porcoesphino@mander.xyz 3 weeks ago
There’s huge risk here but I don’t think most are designed to control people’s opinions. I think most are chasing the cheapest option and it’s expensive to have people upset about racist content so they try to train around that sometimes too much leading to black Nazi images etc.
But yeah, it is a power that will get abused by more than just grok
DandomRude@lemmy.world 3 weeks ago
I use various AI models and I repeatedly notice that certain information is withheld or misrepresented, even though it is freely available in abundance and is therefore part of the training data.
I don’t think this is a coincidence, especially since the operators of all cloud LLMs are so business-minded.
GamingChairModel@lemmy.world 3 weeks ago
but I don’t think most are designed to control people’s opinions
Yeah I’m on team chaos theory. People can plan and design shit all they want, but the complexity will lead to unexpected behavior, always. How harmful that unwanted behavior is, or how easy it is to control or contain, is often unknown in advance, but invented things tend to develop far, far outside the initial vision of the creators.
Evil_Shrubbery@thelemmy.club 3 weeks ago
(Spicy sauce: oglaf.com/gifted)
vas@lemmy.ml 3 weeks ago
Why should we care about an AI tailored to that one person whom many consider a fascist?
Negative publicity is also publicity. Please let’s rather ignore this stupidity.
CrackedLinuxISO@lemmy.dbzer0.com 3 weeks ago
From the original 404 article:
And yet, this is of course an extreme example of the broader political project of AI chatbots and LLMs: They are top-down systems controlled by the richest people and richest companies on Earth, and their outputs can be changed to push the preferred narratives aligned with the interests of those people and companies. This is the same underlying AI that powers Grokipedia, which is the antithesis of Wikipedia and yet is being pitched by its creator as being somehow less biased than the collective, well-meaning efforts of human volunteers across the world.
You may already know this, but a lot of everydy people don’t. They still think that a computer can’t have bias, and if all these tech bros and business leaders are talking about AI then maybe it does make sense to replace our society with an impartial machine. This article is for them.
Scotty_Trees@lemmy.world [bot] 3 weeks ago
And with headlines like this, why the fuck do I care about any of the news coming out? This is so fucking ridiculous and stupid. Like come on. What the fuck are we doing in 2025?
bytesonbike@discuss.online 3 weeks ago
Sometimes I like to find top headlines from other countries.
Like damn, your country’s biggest news today is that you’re replacing outdated lights with eco-friendly alternatives?
Must be nice.
MadPsyentist@lemmy.nz 3 weeks ago
Current headline in NZ. Talking about how the government is going to charge for parking at 3 of the countries tourist hotspots over the summer.
Tryenjer@lemmy.world 3 weeks ago
Bannon’s strategy: flood the zone with shit and steal as much as you want from the people.
bridgeenjoyer@sh.itjust.works 3 weeks ago
People want this. Its sports to them, all entertainment and great if “their side” wins. Fucking sickening humans.
krooklochurm@lemmy.ca 3 weeks ago
I dunno about you but I’m shumming whenever possible.
SethTaylor@lemmy.world 3 weeks ago
This is undeniable proof that Grok is the most accurate.
bitjunkie@lemmy.world 3 weeks ago
Finally some accurate information from an LLM
ekZepp@lemmy.world 3 weeks ago
I mean, it is funny and all… but at this point, i’m just so tired of all this shit, I don’t even care. 😮💨
driving_crooner@lemmy.eco.br 3 weeks ago
I miss my RiF keyword filters.
ekZepp@lemmy.world 3 weeks ago
Piefed have some keyword filter.
Sterile_Technique@lemmy.world 3 weeks ago
…our governing hierarchy is just a long list of who’s blowing who, isn’t it?
Sonor@lemmy.world 3 weeks ago
it’s a snake blowing it’s own tail
W3dd1e@lemmy.zip 3 weeks ago
How many resources were destroyed getting Grok to say Elon drinks piss? We all already know he drinks piss. We don’t need to waste precious resources on getting a robot to say it.
Montreal_Metro@lemmy.ca 3 weeks ago
FYI entire forest was destroyed to bring you this critical information about Elon’s BJ skill.
null_dot@lemmy.dbzer0.com 3 weeks ago
Probably not that much honestly.
The resource intensive part is the initial training.
Tuning later is much less so.
Jarix@lemmy.world 3 weeks ago
Grok got new training which includes the Epstein files?
phx@lemmy.world 3 weeks ago
Aaaand now we know who “Bubba” really is
oneofmany@lemmy.world 3 weeks ago
BREAKING: Chatbot says random bullshit!
greenbelt@lemy.lol 3 weeks ago
BREAKING: bullshit generator generates bs.
db2@lemmy.world 3 weeks ago
In this case I’m inclined to believe the dumb llm 100%.
falseWhite@lemmy.world 3 weeks ago
I guess every once in a while, even Grok spews out some truth.
BarneyPiccolo@lemmy.today 3 weeks ago
A broken clock is right twice a day. A blind squirrel occasionally finds a nut.
SaveTheTuaHawk@lemmy.ca 3 weeks ago
Squirrels have poor eyesight. Almost blind.
Wilco@lemmy.zip 3 weeks ago
How do you know the claims are false? Maybe the AI knows more than we do.
TropicalDingdong@lemmy.world 3 weeks ago
We’re gonna need to bring Bubba in to evaluate…
NOT_RICK@lemmy.world 3 weeks ago
I can’t stop laughing at these, what a great day of shitposting
edgemaster72@lemmy.world 3 weeks ago
Well now let’s hear it out…
wuffah@lemmy.world 3 weeks ago
Humanity invented the assembly line and the first thing we did was have a huge war with the rest of humanity.
Humanity invented the atomic bomb, and the first thing we did was drop it on humanity, twice.
Humanity invented the Internet, and the first thing we did was figure out how to censor humanity.
Now humanity has invented AI, a queryable sum of all human knowledge, and the first thing we do is try to manipulate humanity with it.
h3rmit@lemmy.dbzer0.com 3 weeks ago
I don’t really know for assembly line and the atomic bomb, but for internet and AI the first thing we did is porn.
wuffah@lemmy.world 3 weeks ago
sigh yes… and also porn. So, so much porn.
Asfalttikyntaja@sopuli.xyz 3 weeks ago
Well, theoretically you can use assembly lines for making dildos.
mechoman444@lemmy.world 3 weeks ago
Haywire? That all sounds legit to me!
andallthat@lemmy.world 3 weeks ago
AI goes haywire and starts actually giving reliable information
ArmchairAce1944@discuss.online 3 weeks ago
I wouldn’t be surprised if this was true…
obey@lemmy.wtf 3 weeks ago
He’s one talented guy
Kolanaki@pawb.social 3 weeks ago
Did it really go haywire, or was it actually giving 100% accurate facts and they are just claiming it went haywire? 🤔
mrgoosmoos@lemmy.ca 3 weeks ago
I’ve got a question not related to the topic of the post: why is this site so often posted, when it has an ad blocker banner that you can’t just click a button to dismiss?
Why is this site considered acceptable? Do they generally do good work, and so people agree that disabling adblockers are worth it?
I’m just so used to sites like this being shit upon and posters linking them told to go fuck themselves and smarten up
REDACTED@infosec.pub 3 weeks ago
Musk responded in a statement saying, “Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me. For the record, I am a fat retard.”
Damn, a self-conscious retard
cley_faye@lemmy.world 3 weeks ago
I’ll never put foot on that hellhole again, but it would be funny if this was something he posted on twitter and grok showed up to “correct” him.
greenbelt@lemy.lol 3 weeks ago
Sorry Mr. Musk, I need to tell you: You are not fat but healthily athletic. Your strength only compares to the strongest man on the planet: Donald J Trump!
phoenixz@lemmy.ca 3 weeks ago
For the record, I am a fat retard.
Well Elmo, that might be the first, and likely last thing we both agree on
Ilovethebomb@sh.itjust.works 3 weeks ago
After the posts were deleted, Musk responded in a statement, saying, “Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me. For the record, I am a fat retard.”
I have to hand it to him, that’s a pretty funny response.
sns@lemmy.dbzer0.com 3 weeks ago
Under no circumstances do you have to “hand it to him”.
SabinStargem@lemmy.today 3 weeks ago
Ilovethebomb@sh.itjust.works 3 weeks ago
Look, I can appreciate a funny reply from someone awful, OK?
ininewcrow@lemmy.ca 3 weeks ago
I think they’re programming the ketamine into the LLM at this point
SocialMediaRefugee@lemmy.world 3 weeks ago
I’m convinced the Amish may have the right idea.
cowfodder@lemmy.world 3 weeks ago
Image
sns@lemmy.dbzer0.com 3 weeks ago
Ted Kaczynski wasn’t as crazy as we all thought
Lost_My_Mind@lemmy.world 3 weeks ago
No no. He was.
jali67@lemmy.zip 3 weeks ago
Or when tech is in the hands of a select few vile men, it is becoming a net negative. It definitely is not for the public good.
peopleproblems@lemmy.world 3 weeks ago
Software Engineer here: computers were a mistake.
Quokka@quokk.au 3 weeks ago
Computers are great, it’s capitalism that is the issue.
smeenz@lemmy.nz 3 weeks ago
Discovering fire was a mistake.
clif@lemmy.world 3 weeks ago
Grok here: computers are fine, humans were a mistake
ShaggySnacks@lemmy.myserv.one 3 weeks ago
I’m going with the Luddites.
A_norny_mousse@feddit.org 3 weeks ago
Fully agree with the class struggle aspect here.
Weavers destroying industrial looms. How would that translate to today?
ObviouslyNotBanana@piefed.world 3 weeks ago
Specifically on the tech thing then. They have a lot of things wrong lmao
SocialMediaRefugee@lemmy.world 3 weeks ago
Yes, just the tech part.
knobbysideup@sh.itjust.works 3 weeks ago
Incest, rape, and animal abuse while dodging taxes?
kilgore_trout@feddit.it 3 weeks ago
Apparently they are over 400.000 in the US now. Time to join!