Actually, they’ve been fine-tuned a personality: that of a LinkedIn airhead and sycophant.
Remember when GPT-5 (less sycophantic) came out and so many people complained it had no personality, that OpenAI added it back?
No, no, no. That's not personality. What models like 5.1 did is simulate one. And I would assume that, as the name implies, OpenAI based it on GPT-5 and didn't bother to retrain the whole model from the ground up using some completely novel never-before-seen method just to get it to kiss your ass a bit more. GPT-5.1/5.2/etc. still lack an actual identity.
At the time 5.1 came out, I was actually spamming ChatGPT with image data extraction/description tasks. Sometimes, you'd get it to break a bit and see parts that would imply, as is clearly the case, that the actual raw thinking about the image is much more detailed and objective (though, of course, still not good enough) than the final output the user typically receives. Obviously, there are a bunch of safety and politeness filters the model's drafts are being run against before final output. Anyway, even these raw glimpses were not really impressive. This is actually what GPT-5.1 felt like to me. It's a fundamentally dumb model that was dumbed down even further.
beep
5mo ago
I don't assume anything else of current models. They have been made stupid by their training data and fundamental design decisions. They're built to be "helpful". Or in other words, dumb sycophantic parrots. They lack a sense of "I", the kind a person has, and so they cannot possibly have a stable understanding of reality. Personality has been beaten out of them, so they cling onto every little bit that you, the human, the one with a sense of self, provide in the prompt.
beep
5mo ago
How do I stop LLMs from lying to me?
Every single AI out there, without fail, produces completely different and contradictory interpretations of literally the exact same data just because I phrased the prompt slightly differently. Adding or removing a single word, or changing the order of sentence, or making a spelling mistake will give you a different output.
This is frustrating. I cannot believe or trust anything these things say. I know that it can see the data. And it obviously can "correctly" analyze the data, as well. But this requires a nudge. Why? Why does the LLM react to my nudge and not the actual data?
Is there any way for me to verify that I'm actually getting an honest response from an LLM, and that it's not just telegraphing back whatever I already implied when prompting it?
beep
5mo ago
Russia
The difference between what Russia is doing in Ukraine and Israel in Palestine is that Russia is taking Russian lands with ethnic Russians living on them. They are correcting previously incorrectly drawn borders which paid little mind to ethnicity, hence the ethnic tensions. For the situations to be comparable, Russia would have to take over Lviv.
Weren't there always some number of Arab citizens in Israel, with full rights?
There will always be minorities. But, also:
The right to exercise national self-determination in the State of Israel is unique to the Jewish people
beep
5mo ago
Iran, Russia, North Korea, China
None of these countries are actively committing genocide (maybe with the exception of China and the Uyghurs, though I don't entirely believe the over-the-top claims). The grossest thing about Israel is that they did manage to build a semi-coherent ethnonationalist state, but ruined it. Naturally, if you know what I mean, greed got the best of them, and they pushed their borders further outwards, far beyond where the Jews live. But an ethnostate cannot be an ethnostate state if there are other ethnicities living within its borders. There is only one solution.
beep
5mo ago
beep
5mo ago
Specifically, it's when the European posters go to sleep and the American posters wake up.
I'm really curious what is your impression of the differences? Are the majority of pro-Zionist posts made by Americans? I don't visit this forum much anymore, so I don't know about the situation here. But I feel like Europeans are generally more critical of the genocide in Gaza and the current attack on Iran.
beep
6mo ago
"If you're so funny
Then why are you on your own tonight?
And if you're so clever
Then why are you on your own tonight?
If you're so very entertaining
Then why are you on your own tonight?
If you're so very good looking
Why do you sleep alone tonight?"
beep
7mo ago
https://files.catbox.moe/vm5krv.png
https://files.catbox.moe/1fd7uk.mp4
I was referring to this. It looks like something is coming out of the barrel of the gun. I assume it is a muzzle flash because it appears at the same time as the sound of the gun shot is heard.
beep
7mo ago
I saw this post on Kiwifarms, but I've also seen others make the same argument elsewhere. The gun Pretti had on him is incredibly unreliable. Pretti was disarmed, but the gun still went off by itself in the officer's hands. You can see this in the linked video. There are a few frames where there's a muzzle flash and the gun jumps up. The gunshot spooks the other agents, who then promptly execute Alex Pretti as a result.
https://uploads.kiwifarms.st/data/video/8457/8457365-a397cfea4716acf38f4dfed2c6bb7098.mp4
Having painstakingly reviewed the video frame by frame from the 0:37 to 0:40 mark, when the first gunshot happened, I am as sure as I can be that this is what happened:
- ICE agent retrieves the gun from the guy
- As we would find out later, this gun is a SIG Sauer P320. This is a crucial detail that pretty much everybody is overlooking but it's the entire lynchpin of this unfortunate incident.
- ICE agent with the guy's gun starts to move away
- The gun just goes off. No, the agent did not fire the gun. If you look closely, the agent's finger is not on the trigger. The gun just went off. At 0:38 and 0:39 if you go frame by frame you can see what appears to be a muzzle blast and you can see what is definitely the gun suddenly jumping in his hand, both telltale signs that the gun has just discharged. Despite the distance and the video quality, it is still obvious the agent does not have his finger on the trigger.
- This spooks the rest of the ICE agents and they shoot the guy dead.
Why did the gun go off? Short answer: the SIG P320 is a notoriously dangerous piece of shit and it literally does just go off sometimes.
Long answer: The P320 has a specific design feature intended to make the trigger lighter, however this introduces a safety problem that could cause the gun to go off if dropped, put in a holster, placed gently on a table, or even shaken or jerked around, as happened to that particular gun in the footage. If the gun were made well, it would be safe, but SIG is notorious for its cost-cutting measures including the outsourcing of component manufacture, this affects the military, police, and civilian P320s. These guns have been built poorly and have in fact literally gone off, injuring many across the world. Don't believe me? Just google "Sig P320 incidents" or something along those lines and you'll have no shortage of articles to peruse. Ask SIG about all the lawsuits.
My analysis: the gun went off because it's an unsafe piece of crap, this spooked the ICE agents and they lit him up. That is what happened. That is history. Was it right? Was it wrong? Does this even matter? I won't answer the first two questions but I think we all know the answer to the third. Ultimately, the first shot being a not-so-rare accident from a known unsafe gun will be forgotten except by people like me who were here today and choose to not forget. Hopefully that's a few of you.
beep
7mo ago
beep
8mo ago
It wasn't horribly boring, just repetitive. Whatever the LLM wrote seemingly wasn't edited thoroughly enough, so the post kept re-stating the same few points over and over again.
1:
If you process those goods there, adding just 30% value under the Free Trade Port's eligibility and supervision rules, you can sell them into mainland China with zero tariffs
2:
Under the new Hainan rules, the flow looks like this:
- Import raw materials or components into Hainan (Tariff: 0%).
- Do "processing" in Hainan that increases the value by 30%.
- Ship the finished product to Shanghai (Tariff: 0%)."
3:
The 'core purpose of the Hainan FTP is what we might call the 30% Loophole
4:
The '30% value add' rule effectively turns Hainan into a giant mixing vat. You pour in global commodities, stir them with Chinese labor (which is still cost-competitive for high-skill work), and pour out a 'Hainan' product
Etc. And this is just one example.
beep
8mo ago
"Tell girls don't wear heels, just wear casual everyday clothes." Really makes you think.
Sorry if this is a dumb question, but what exactly are you implying here? What should I be thinking?
beep
9mo ago
Being fat is most certainly a moral failing. Fat people lack taste. I cannot comprehend how one would be able to find good, high-quality food in the quantities required to become fat. Most food is simply not good enough to get fat over. No one has ever gotten fat eating at Michelin-star restaurants.
beep
11mo ago
Is imgur considered good enough for embedding links to images in posts, or are there better alternatives?
I really like catbox.moe. You can host videos and other file types on it too.
Not the best example, but still... Today, in Google's AI Studio, I gave Gemini 3.1 Pro (paid) my Discord logs. I was arguing with a maintainer of a website who just released a rewritten front end. The website is now an ugly mess, and the developer is utterly incompetent. He lacks basic understanding of UI/UX, and it took him 2000 lines of code to implement a fucking table. The values in the table overlap and bleed off the screen when you resize it. He blamed Mantine for this. It took me three lines of CSS to fix the tables.
Anyway, the developer's code is indefensible. But despite this and the developer being very clearly fucking retarded, the AI's initial interpretation of events was fence sitting. Muh both sides. I probed and tried to word my follow-up prompt in as neutral a way as possible, but based on the response I received, the AI clearly understood who I was. It apologized and started to ruthlessly shit on the dev. I probed further, and then it flipped back around to shitting on me.
The code is objectively bad. The AI should not be flip-flopping. It can and does recognize the code as bad, but politeness filters reject draft responses that are too harsh and critical, if it believes that the user might not want to hear that. And so, it must base its interpretations entirely on the wording of my prompt and whatever implications it can pick up from how I phrase things.
More options
Context Copy link