Wow! it is like open sourcing the technology and allow all of contribute and modify prevent said software to enshitify the the future
Cough
Cough
Microsoft Windows
Airbnb boss Brian Chesky told Bloomberg in October his company relied “a lot” on Alibaba’s Qwen to power its AI customer service agent.
He gave three simple reasons - it’s “very good”, “fast” and “cheap”.
Going into 2025, the consensus was despite billions of dollars being spent by US tech firms, Chinese companies were threatening to pull ahead.
“That’s not the story anymore,” Boudier said. “Now, the best model is an open-source model.”
A report published last month by Stanford University found Chinese AI models “seem to have caught up or even pulled ahead” of their global counterparts - both in terms of what they’re capable of, and how many people are using them.
Interesting stuff. I hope Sam Altman loses everything.
I use a lot of models and Chinese ones are the best. They are a bit restricted for certain fields, but not more than Microsoft phi models. And they are imcredibly cheaper even when using inference APIs.
China is winning the race just because they release open weights of a models. The main question is: why do they do that? They don’t earn money, so there must be something else that they gain. Political influence maybe? Spreading propaganda? Has anyone checked if those models answer questions about Tiananmen Square?
People using good chinese stuff makes their material conditions better, making them “china good”, so they will automatically disbelieve made up “china bad” narratives like yours. And there’s nothing you can do about it.
Because when something is open-source, reseaechers are more prone to adopt them. Companies pay less, and the market will be your. Instead of centralizing trhough open ai añi or microsoft azure, the let you run those models on your stack. You domt need to pay fees there. More people get involved, the models grows, you have free qa and developers working on the product. Alibaba is getting a lot of revenues just because people are switching to their infrastructure.
except there’s no fascist rule, taiwan claims that the mainland belongs to them and viceversa, and the only reason why they’re not unified is because of western imperialism. yeah, it’s literally the same. malvinas belongs to uk also /s
except there’s no fascist rule, taiwan claims that the mainland belongs to them and viceversa, and the only reason why they’re not unified is because of western imperialism. yeah, it’s literally the same. malvinas belongs to uk also /s
“China bad” aside. Chinese companies are entering a space dominated by US companies with endless money, and this is one way they can gain a foothold. Also in China a working solution is preferred over intellectual property. An open model allows more people and companies to produce actual products and systems. There is no emphasis on value for shareholders or stock prices.
There is no such thing like “running model with your training data”. To change model’s behavior you need to fine-tune it, which means: to continue training it on your own data set. For this to happen you need to have your own dataset, computing power and knowledge how to do it because you may as well make your model performing worse. It is not an easy task.
It’s not easy, but it’s done all the time. New models, new LoRAs, and in some cases, the training data doesn’t even need to be very large for a specific task.
You don’t need the entire training dataset that the model was built from.
All AI models have some boundaries implemented. A lot of them of ideological nature.
China has a big interest to challenge the US’s hegemonial status. If “they” win the AI arms race "they"later can charge actual money when having established themselves as THE winner and only supplier of consumer AI. It’s how most western monopolies became to be one.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
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
Wow! it is like open sourcing the technology and allow all of contribute and modify prevent said software to enshitify the the future
Cough
Cough
Microsoft Windows
Interesting stuff. I hope Sam Altman loses everything.
The whole AI as a service business model is cooked now. My prediction is that even stuff like coding will soon work well enough with local models. There are going to be very few cases to justify paying subscription for AI services either for companies or individuals. And this stuff is moving so incredibly fast. For example https://dev.to/yakhilesh/china-just-released-the-first-coding-ai-of-2026-and-its-crushing-everything-we-know-3bbj
Can’t wait for the bargain bin data center hdds
and cheap GPUs :)
The latest GLM 4.7 Flash is insanely good. Just sayin…
I use a lot of models and Chinese ones are the best. They are a bit restricted for certain fields, but not more than Microsoft phi models. And they are imcredibly cheaper even when using inference APIs.
China is winning the race just because they release open weights of a models. The main question is: why do they do that? They don’t earn money, so there must be something else that they gain. Political influence maybe? Spreading propaganda? Has anyone checked if those models answer questions about Tiananmen Square?
People using good chinese stuff makes their material conditions better, making them “china good”, so they will automatically disbelieve made up “china bad” narratives like yours. And there’s nothing you can do about it.
Because when something is open-source, reseaechers are more prone to adopt them. Companies pay less, and the market will be your. Instead of centralizing trhough open ai añi or microsoft azure, the let you run those models on your stack. You domt need to pay fees there. More people get involved, the models grows, you have free qa and developers working on the product. Alibaba is getting a lot of revenues just because people are switching to their infrastructure.
Taiwan moment
except there’s no fascist rule, taiwan claims that the mainland belongs to them and viceversa, and the only reason why they’re not unified is because of western imperialism. yeah, it’s literally the same. malvinas belongs to uk also /s
Yeah, Chinese models have the same issue with Taiwan
except there’s no fascist rule, taiwan claims that the mainland belongs to them and viceversa, and the only reason why they’re not unified is because of western imperialism. yeah, it’s literally the same. malvinas belongs to uk also /s
Aww look at you still malding your color revolution failed.
“China bad” aside. Chinese companies are entering a space dominated by US companies with endless money, and this is one way they can gain a foothold. Also in China a working solution is preferred over intellectual property. An open model allows more people and companies to produce actual products and systems. There is no emphasis on value for shareholders or stock prices.
The same reason Facebook’s llama was “open”: when you’re behind, try to get adoption as open source.
Yes a long time ago and they don’t, but the AI models and training data are two different things.
Also as they are open source, there’s nothing stopping anyone from running these AI models locally with your own training data.
There is no such thing like “running model with your training data”. To change model’s behavior you need to fine-tune it, which means: to continue training it on your own data set. For this to happen you need to have your own dataset, computing power and knowledge how to do it because you may as well make your model performing worse. It is not an easy task.
It’s not easy, but it’s done all the time. New models, new LoRAs, and in some cases, the training data doesn’t even need to be very large for a specific task.
You don’t need the entire training dataset that the model was built from.
All AI models have some boundaries implemented. A lot of them of ideological nature.
China has a big interest to challenge the US’s hegemonial status. If “they” win the AI arms race "they"later can charge actual money when having established themselves as THE winner and only supplier of consumer AI. It’s how most western monopolies became to be one.
The good thing about open models is that you can train the biases and rules out of them. The weights are not concrete.