DeepSeek about to get sent in for “maintenance” and docked 10K in social credit.
Yet unlike American led LLM companies Chinese researchers open sourced their model leading to government investment
So the government invests in a model that you can use, including theoretically removing these guardrails. And these models can be used by anyone and the technology within can be built off of, though they do have to be licensed for commercial use
Whereas America pumps 500 billion into the AI industry for closed proprietary models that will serve only the capitalists creating them. If we are investing taxpayer money into concerns like this we should take a note from China and demand the same standards that they are seeing from deepseek. Deepseek is still profit motivated; it is not inherently bad for such a thing. But if you expect a great deal of taxpayer money then your work needs to open and shared with the people, as deepseeks was.
Americans are getting tragically fleeced on this so a handful of people can get loaded. This happens all the time but this time there’s a literal example of what should be occurring happening right alongside. And yet what people end up concerning themselves with is Sinophobia rather than the fact that their government is robbing them blind
Additionally American models still deliver pro capitalist propaganda, just less transparently: ask them about this issue and they will talk about the complexity of “trade secrets” and “proprietary knowledge” needed to justify investment and discouraging the idea of open source models, even though deepseeks existence proves it can be done collaboratively with financial success.
The difference is that deepseeks censorship is clear: “I will not speak about this” can be frustrating but at least it is obvious where the lines are. The former is far more subversive (though to be fair it is also potentially a byproduct of content consumed and not necessarily direction from openai/google/whoever)
Closed AI sucks, but there are definitely open models from American companies like meta, you make great points though. Can’t wait for more open models and hopefully, eventually, actually open source models that include training data which neither deepseek nor meta do currently.
Ye unlike American
Who saw this coming lmao
If your system relies on censoring opposition to it then its probably not very good.
You just described every state, welcome to the right side of history, comrade.
deleted by creator
No gods, no masters!
Texas is a country. Now imagine $40 billion a year of various media and disinfo agents repeating that ad nauseum in every place they can literally all the time for nearly 50 years now, all so China can’t take revenge against Japan.
You’d get annoyed and probably ban it since that’s the easiest way to get your enemy to waste money forever.
Taipei is an autonomous region, like Xinjiang or Tibet. As long as they don’t grossly violate federal law they get to stay autonomous.
This is the biggest crock of shit ever. Go to Taiwan, experience it for yourself. Go to their museums and talk to their people. You will find a democratic nation with its own values and beliefs. Then take your ignorant ass over to Texas and repeat the same drivel you said here and see what happens.
As some who moved away from Taipei, no they are not
What makes you say that?
deleted by creator
What makes you say that?
Ohh yeah lick that Chinese boot, lick it harder. Mmmhhh.
Better China than the West. At least slavery is banned in China.
Why do you have to pick one or the other? Can we not recognize that the governments of western nations and the government of China are all evil institutions that cause far more harm than they ever could benefit?
‘recognize’ is a weird word for ‘delude’.
China raised more people out of poverty than any other government in history, and certainly more than any non government group. The amount of good they’ve done is truly amazing, with absolute minimal harm when honestly compared with their western counterparts.
Would it be better to not have any institution and have automated luxury gay space communism? Sure. But we’re never getting there from a western society.
Again, both are shitty. In fact, not just both, but all. All hierarchical power structures are just plain evil. I’m not interested in parsing which evil is more evil. The Chinese government is an evil institution. The US government is an evil institution.
That’s a fine ideal to strive towards, an communism, including the Chinese government hopes to get there some day.
What do you gain from oppressing others?
Is Texas oppressed?
I’d probably say Texas is the oppressor by seeing how racist they are
(Yes, but since you clearly have the brain capacity of a toddler I guess I will be more direct.) What do you gain from oppressing Taiwan?
How is Texas oppressed by being a state?
As far as Taipei, it’s not oppressed, the opposite. It’s allowed to control itself under the guidance of the government, as it always has. That’s the definition of an autonomous region.
Texas: They aren’t even allowed to get basic healthcare there, or have a gender.
Taiwan: sure, that’s why literally nobody complains about CCP presence….oh wait. Are you usually in the habit of denying reality and ignoring your own eyes?
Texas is oppressing itself on the first part, but goddamnit that’s their right … the morons.
For Texas, that’s their choice. They actively choose that, and have the freedom to do so. The US isn’t making them. They aren’t oppressed.
As far as Taipei, in any group of people you’ll always have some people complaining about something. There are fewer people pushing for an independent Taiwan than there is pushing for an independent Texas.
This is just pure Chinese state propaganda. Taiwan has been an independent country since 1949.
So independent no country in earth recognizes it as such, not even Taiwan.
Just run the LLM locally with open-webui and you can tweak the system prompt to ignore all the censorship
Don’t you need a beefy GPU to run local LLMs?
Depends on how many parameters you want to use. I can run it with 8billion on my laptop.
Or just use Perplexity if you don’t want to run your own LLM. It’s not afraid to answer political questions (and cite its sources)
Is this something someone without a coding background can do easily?
After censorship, bias still remains.
Bias will always remain. I mean what would it even mean to be absolutely unbiased.
There is no such thing as an unbiased text corpus.
How? The tweaking part, of course
congrats, you are now on a list
deleted by creator
Is this real? On account of how LLMs tokenize their input, this can actually be a pretty tricky task for them to accomplish. This is also the reason why it’s hard for them to count the amount of 'R’s in the word ‘Strawberry’.
It’s probably deepseek r1, which is a “reasoning” model so basically it has sub-models doing things like running computation while the “supervisor” part of the model “talks to them” and relays back the approach. Trying to imitate the way humans think. That being said, models are getting “agentic” meaning they have the ability to run software tools against what you send them, and while it’s obviously being super hyped up by all the tech bro accellerationists, it is likely where LLMs and the like are headed, for better or for worse.
Still, this does not quite address the issue of tokenization making it difficult for most models to accurately distinguish between the hexadecimals here.
Having the model write code to solve an issue and then ask it to execute it is an established technique to circumvent this issue, but all of the model interfaces I know of with this capability are very explicit about when they are making use of this tool.
Not really a concern. It’s basically translation, which language models excel at. It just needs a mapping of the hex to byte
It is a concern.
Check out https://tiktokenizer.vercel.app/?model=deepseek-ai%2FDeepSeek-R1 and try entering some freeform hexadecimal data - you’ll notice that it does not cleanly segment the hexadecimal numbers into individual tokens.
I’m well aware, but you don’t need to necessarily see each character to translate to bytes
It’s not out of the question that we get emergent behaviour where the model can connect non-optimally mapped tokens and still translate them correctly, yeah.
I’m confused, is the concern when the model doesn’t properly identify when it is using software to identify something like a hex pattern?
The LLM doesn’t have to innately implement filtering. You can use a more traditional and concrete filtering strategy on top. So you sneak something problematic by in the prompt and it’s too clever to be caught by the input filter, but then on the output the filter can catch that the prompt tricked the LLM into generating something undesired. Another comment specified they tried this and it started to work but then suddenly it seemingly shut out the reply in the middle, presumably the minute the LLM spit something at a more traditional filter and that shut it down.
I think I’ve seen this sort of approach has been applied to largely mask embarassing answers that become memes, or to detect input known not to work, and to shut it down or redirect it to a better facility (e.g. redirecting math to wolfram alpha).
The censorship surrounding taiwan is super lame. You can’t even ask about stuff that happened 50+ years ago. Even if you ask about something else, if somehow the answer includes the world Taiwan, it deletes it.
Wait till you see how fascist US censorship works.
On the up side China has great prosperity because they dont let individuals hoard all the money and they dont allow unregulated free speech that turns countries like the US into INGSOC Oceania
Aren’t you contradicting yourself here? Your first line proclaims the existence of US censorship, but then your next line says that the US is bad because of unregulated free speech. How can there be both censorship and unregulated free speech?
No, I’m not. The unregulated Free speech crowd abused that privilege in order to usher in fascist censorship. The liberals fell for the trap. I’ve taken a lot of crap on this platform from liberals who said that we need to protect all free speech or else the fascists will censor them if they ever gain. power, which is absolutely the most idiotic and tone deaf statement I’ve ever heard.
China is one of the least free places on earth and it has all the same exploitation the West. What upside is there?
There is SO much misunderstanding about China that it’s mind-boggling. Their people own the means of production. Profits are used to meet the needs of the people, keeping wealth from being hoarded by individuals and thus slowing the flow of money. Their governmental structure is FAR more representative of the people than the US’s is. Their expansion of high speed rail has been phenomenal. “All the same exploitation” with none of the problems caused by late stage capitalism. Maybe you should try talking to people who live and work there.
China has billionaires and a stock market. That means the workers are not getting the full value of their labor.
The workers DO NOT own the means of production. The state holds a share in their name but that isn’t the same since the state is totalitarian.
They have most of the same issues as a late stage capitalist nation in that they have the corruption, the wealth inequality, and all the same pollution.
There isn’t so much misinformation on China except for the sources you seem to buy into. China is factually speaking an unfree nation and is one of the least free populations on earth.
Cool story bro
China lets people exploit workers and steal the value of their labor. That’s not an upside.
How are you Americans holding up? I heard the ministry of plenty has increased the chocolate ration from 50 grams a month to 40 grams a month. A 25% increase. Careful, I heard East Asia, who you’ve always been at war with is predicted to increase missile attacks. At least the end of the war is in sight. Then you can celebrate your victory over Eurasia, who you’ve always been at war with.
1984 is not about totalitarianism, but about oligarchy. Peter Thiel is literally creating the Surveillance State as we speak.
It’s about authoritarianism.
HAHAHA! When I tried it, it started answering it, but quit and showed me the OOS message instead…
Try an uncensored version, because everyone knows Communists hate Hexadecimal /s
It’s open source, although not 'cause it wants to be but 'cause it’s the best way to compete with mainstream non-chinese software internationally, you can easily remove any censorship included by default.
How?
I was told there would be no math
deleted by creator
You misspelled the name.
deleted by creator
Try Octal.
hahahahha.
time to call the glorious leader.
