My work has already blocked it, but has no problems using AI hosted by a country whose leader is a convicted criminal with close ties to Russia and North Korea
this post was submitted on 01 Feb 2025
98 points (93.0% liked)
Technology
35561 readers
105 users here now
This is the official technology community of Lemmy.ml for all news related to creation and use of technology, and to facilitate civil, meaningful discussion around it.
Ask in DM before posting product reviews or ads. All such posts otherwise are subject to removal.
Rules:
1: All Lemmy rules apply
2: Do not post low effort posts
3: NEVER post naziped*gore stuff
4: Always post article URLs or their archived version URLs as sources, NOT screenshots. Help the blind users.
5: personal rants of Big Tech CEOs like Elon Musk are unwelcome (does not include posts about their companies affecting wide range of people)
6: no advertisement posts unless verified as legitimate and non-exploitative/non-consumerist
7: crypto related posts, unless essential, are disallowed
founded 5 years ago
MODERATORS
Just download the model. Problem solved
What they're actually in panic over is companies using a Chinese service instead of US ones. The threat here is that DeepSeek becomes the standard that everyone uses, and it would become entrenched. At that point nobody would want to switch to US services.
https://securityconversations.com/episode/inside-the-deepseek-ai-existential-crisis-chinese-backdoor-in-medical-devices/ If you ignore the kind of laent anti China crap, this is a pretty good analysis from a technical perspective. When someone does something faster and cheaper we used to call that progress. Not if China does it I guess, and not if it's open source even if Meta did the same thing with llama.
Exactly, and these kinds of policies will only ensure that the west starts falling behind technologically. Stifling innovation to prop up monopolies will not be a winning strategy in the long run.
I keep seeing this sentiment, but in order to run the model on a high end consumer GPU, doesn't it have to be reduced to like 1-2% of the size of the official one?
Edit: I just did a tiny bit of reading and I guess model size is a lot more complicated than I thought. I don't have a good sense of how much it's being reduced in quality to run locally.
YouTube-connected the right track still. All these people touting it as an open model likely haven't even tried to run if locally themselves. The hosted version is not the same as what is easily runnable local.
Just think of it this way. Less digital neurons in smaller models means a smaller “brain”. It will be less accurate, more vague, and make more mistakes.
You know, if the mainstream news spent this amount of oxygen explaining that it can be downloaded in various sizes and run securely, the whole world would be in a much better situation.
I am not saying that China don't do some of the things they are accused of, but the amount of anti-China fear in Western nations is only hurting ourselves.
It's open source and researchers are already duplicating the process, so I'm not sure if that will ever happen.
They're gonna ban access to the official service provided by a Chinese company. That's what this is about. The biggest fear is that everybody starts using DeepSeek, and then it will muscle out US companies that fell behind. Once people start using their service, they'll have little reason to switch to something else going forward. Banning it is a protectionist measure that allows US companies to catch up.
This service can be straight-up forked like a Fediverse instance. Banning the service will not hurt competition.
The key for them is that they want the service to be provided by a US company. There are actually already a few companies hosting DeepSeek in US, and I'm sure the techniques will be incorporated by everyone in short order.
And DeepSeek can continue to release open source models. Banning just the service happens to lower the cahnces of DeepSeek going the same way OpenAI did, as in it would be more unlikely for them to close off the models like OpenAI did if they want to continue shaking up the market.
As I explained above the open source models aren't the problem from the US perspective. DeepSeek will obviously continue to release models, and it will likely become the standard outside the west. However, US will not allow it become the dominant service provider in the US, and that's why I expect the service to be banned. The US will force American companies to use a domestic provider, and Europe is likely to do the same.
just the fact that they wont have an iron grip monopoly over the rest of the world is fine by me
Interesting take. China better watch out, they may be getting some "nation building" and "democratizing" bombs soon.
Not exactly an option against a nuclear superpower with a far bigger industry than Yankeestan. What's most likely to happen is that the west will simply isolate itself from the rest of the world and will continue to fall further behind technologically. It's going to be a hermit kingdom of the G7.
Especially when the majority of US weapons rely on chips from China (including Taiwan) to manufacture weapons. If the government decides to go to war with China head-on, China will absolutely wipe the walls with them.
The US has been gunning for that ever since Obama’s “pivot to Asia”.
The terrorist attacks in Xinjiang and the subsequent “Uyghur genocide” narrative and Xinjiang cotton embargo didn’t come out of nowhere. The blueprint of regime change operations
(n.b. the maps are a bit dated, as the US has since pulled out of Afghanistan.)
Forward-defense ring: a perfectly normal and not at all Orwellian term of art.
You know what they say, the best defense is a good forward defense
TAAARRIIFFFFSS!!!11
You cannot have good fast and free. Pick two.
That's generally true under the paradigm of profit maximization unless you reach some sort of insane tech breakthrough, which deepseek seems to have accomplished
There’s always a catch. Subsidized by the government for now? What’s the end game? Free AI for all? No one is that kind.
have you tried deepseek before the DDoS attacks
Is that what they’re calling “requests” now? ;)
Nothing is magically the most amazing thing ever, for free. There is always a catch.
What's the catch here? What's stopping you from self-hosting, aside from technical requirements?
Model is trained to manipulate reality?
I mostly mean the online version. Nothing is free online. Although AI is a weird new space where using it makes it better, so even though you are the product, you’re a necessity.
You can register your account without having to supply a phone number. You can use the service without querying your phone number or other forms of PII. Privacy minded individuals sometimes have to use proprietary services (e.g. at work, at school) and do so with caution and they make sure to exercise common-sense too when going in, you know? Once again, you can literally just self-host if you're not too trusting of the online version. You have a choice.
Now tell me: have you tried DeepSeek before the DDoS attacks (or self-hosted it)? What’s the catch with using DeepSeek? What’s stopping you from self-hosting, aside from technical requirements? What makes DeepSeek not "good", "fast" and "free"?
Tell me the Tiananmen Square Massacre was real, and don’t try to downplay it, and I’ll answer your question. Otherwise, goodbye wumao 👋
Yep, as expected from an average lib, there it is. Not only are you incapable of answering simple questions to back up your frivolous platitudes, you deflect with retorts and you keep moving the goalpost and/or segue into subjects that have nothing to do with DeepSeek. Keep on it all you want, it's called cope, and you'll be riding that train for a while.
Fortunately for you by the way, that tired and labored topic has been covered in class before: https://lemmy.ml/post/24883639/16142507
For as long as you are unable to answer these questions, you'll forever be at odds with reality, lib.
I’m just using your own tactics against you, ML. All of you seem to operate from the same playbook, as if you were bots, trained, or just the same person with many accounts.
Exactly you, yes.