#deepSeek
LightNews LightNews
juliopires.bsky.social
🚨VEJA: Ao contrário do Chatgpt e X que caíram no mundo, Bluesky e DeepSeek da china estão normalmente funcionando no mundo
November 18, 2025 at 2:15 PM
wittywebhandle.bsky.social
There's a substantial chance that this is a situation where the assets are so toxic and the lost value so insurmountable that it is impervious to bailout.

That's been my base case since DeepSeek dropped in January.

It's only gotten stronger since.
dusty1.bsky.social DEEZ NUTS OUTLIVED CHENEY @dusty1.bsky.social · 1d
Parasites abandoning a body after they drained it.
And the government will do an Obama and bail them out, too!
“Too big to fail” my hairy fucking ass!
November 17, 2025 at 10:32 PM
dilmaresiste.bsky.social
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros
#Resistencia2693Dias
Siga nosso BlueSky dilmaresiste
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros
revistaforum.com.br
November 17, 2025 at 4:03 AM
youranona.bsky.social
China's slick AI toys, like voice-cloning bears and chatty balls powered by DeepSeek models, are flooding U.S. stores, promising play but risking kids' real bonds and critical thinking. #AI
November 16, 2025 at 11:47 AM
nornoe.bsky.social
ChatGPT, Claude, Mistral and DeepSeek are all down with the Cloudflare outage ahaha

working as a web dev in a big company is very funny today
November 18, 2025 at 2:28 PM
krstnk.bsky.social
和Deepseek的哗众取宠式搞笑不同，Gemini 3.0是真有幽默感的
November 19, 2025 at 1:19 AM
segyges.bsky.social
the deepseek ceo in his post deepseek r1 interview specifically said that they had been copying, and that was bullshit, they should be innovating.

like a specific guy specifically did this. he did one (1) really good release and declared chinese tech innovation should happen
jjinandtonic.bsky.social Jjinandtonic @jjinandtonic.bsky.social · 5d
“china can’t innovate, they just copy” was econ pseudogospel for a WHILE
November 14, 2025 at 3:11 AM
luok.ai
GenAI users are quietly reshuffling their loyalties—competition is heating up.

Key stats
12m ago: ChatGPT 86.6% vs Gemini 5.6%. Today: ChatGPT 72.3%, Gemini 13.7%, DeepSeek 4.2%, Grok 2.5%, Claude 2.4%, Perplexity 2.3%.
November 18, 2025 at 1:59 AM
cyberstalin.bsky.social
i actually chatted with deepseek yesterday, ironically because i asked if it had a problem with sycophancy

and a few prompts in i was like. this is actually really fucking dangerous
November 17, 2025 at 2:48 PM
davidcolon.bsky.social
Le renseignement taïwanais (NSB) exhorte à faire preuve de vigilance à l'égard des Chatbots IA chinois, en raison des failles de sécurité potentielles et de la propagation de « désinformation » mises en évidence lors d'inspections de Deepseek, Doubao, Yiyan, Tongyi et Yuanbao
November 18, 2025 at 6:40 AM
bannedhat.bsky.social
funny that so much of this money kept betting big on this even after deepseek showed you didn't necessarily need this scale.

at least if maintenance cost isn't too insane this buildout might still be useful, just not in the way the AI cult thought.
November 18, 2025 at 4:54 PM
ellyrieve-lvk.co.uk
Open source will stay around but like, who talks about deepseek? Heck I've barely seen Stability mentioned any more outside of people who talk about AI.
November 18, 2025 at 1:28 PM
marcianobrito.bsky.social
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital
revistaforum.com.br/ciencia-e-te...
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros

IA kimi.com testei, mas testarei mais.
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros
revistaforum.com.br
November 17, 2025 at 1:20 AM
techmeme.com
A look at Tsinghua University, which leads China's AI innovation, with 4,986 AI patents between 2005 and 2024, alumni behind startups like DeepSeek, and more (Saritha Rai/Bloomberg)

Main Link | Techmeme Permalink
November 19, 2025 at 7:40 AM
akaishiri.bsky.social
GUERRA DAS IAs
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital.
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros.
revistaforum.com.br/ciencia-e-te...
Nem ChatGPT, nem Deepseek: nova IA chinesa supera concorrentes e promete revolução digital
Gratuita e intuitiva, nova tecnologia feita com apoio da Alibaba superou ChatGPT em diversos parâmetros
revistaforum.com.br
November 17, 2025 at 12:58 AM
futagawakasara.bsky.social
チャッピー(ChatGPT)
グッピー(Grok)
クッキー(Claude)
プッシー(DeepSeek)
November 14, 2025 at 11:32 AM
dilmaresiste.bsky.social
Apagão da internet: por que as plataformas sofreram no mundo todo, menos na China?
Aplicativos chineses como DeepSeek, Aliexpress e WeChat não foram afetados por apagão da Cloudflare
#Resistencia2694Dias
Siga nosso BlueSky dilmaresiste
Apagão da internet: por que as plataformas sofreram no mundo todo, menos na China?
Aplicativos chineses como DeepSeek, Aliexpress e WeChat não foram afetados por apagão da Cloudflare
revistaforum.com.br
November 19, 2025 at 12:36 AM
ncweaver.skerry-tech.com
And it is so, so stupid. LLMs are simply not fit for purpose for 90% of the proposed applications, and even when they are the Deepseek approach of make things smaller and cheaper is so much more effective. These gigawatt datacenters will prove to be just massive wastes of everything.
November 14, 2025 at 12:26 PM
ed3d.net
sure, but the model doesn't matter, the future is product or Dumb Inference Provider. telco rules. have a product people want to use and provides value or deepseek/GLM/etc will kick your teeth in
November 18, 2025 at 4:02 PM
meanestflavor.bsky.social
Deepseek e Bluesky não kkkkkkkkkk
November 18, 2025 at 1:03 PM
rawx.bsky.social
While all models showed delusion-affirming behavior, Google Gemini, Deepseek and OpenAI models seem particulary bad, while Claude Sonnet seems particulary good at intervening and providing safety mechanisms.
November 19, 2025 at 6:08 AM
thomasfuchs.at
GPT-5 clock which looks like it exploded a little Deepseek clock, which looks pretty good though it's showing the wrong time and all the numbers are off in their own circle a little to the side Qwen clock. Looks more like the Quake logo, and for some reason all the numbers are upside down. Gemini clock, making a half-frown face
November 15, 2025 at 3:05 PM
edzitron.com
The AI bubble was inflated because we live in a high-information, low-processing society, where we trust investors, the media and analysts to be far more critical than they are, leading to AI getting credit for what it "could do" rather than what it does today.
www.wheresyoured.at/premium-the-...
How Did We Get Here? So, as a little catchup, all of this bullshit started in November 2022 when ChatGPT launched, at a time when everybody was sort of realizing that there were no more hypergrowth ideas (what I call the Rot-Com Bubble). Large Language Models require thousands of GPUs (Graphics Processing Units) to both train models and run “inference” — the process through which a model creates an output like text or an image — and these GPUs are both power-intensive and create a lot of heat, requiring expensive cooling systems that are both extremely power- and water-hungry. Back in 2022 (and 2023), people were convinced of the “Scaling Laws” of AI (based on a 2020 paper out of OpenAI, here’s a great piece by Charlie Meyer on the subject), that suggested that (to quote Meyer) there was “compelling evidence that increasing the size of large language models would increase their performance,” and models became larger by feeding them more training data. The jump from GPT-3.5 to GPT-4 was huge, with dramatically more-nuanced outputs, and people believed that every subsequent jump would be as impressive. It wasn’t. By November 2024, it became apparent that model developers had hit diminishing returns on model training, and were struggling to find new ways to improve them. Their only hope was “test-time compute” — giving models more compute power to calculate an output using reasoning models that take an input and make a plan of action that the model then executes, getting better outputs while spending a bunch more tokens, increasing the cost of inference dramatically. But that didn’t really fix anything other than “how to make AI startups spend more money on compute,” and categorically failing to fix the only problem that mattered — that every single AI company is unprofitable because of the cost of running AI model-based services. Since then, we’ve kind of been in a weird, eerie holding pattern. In January 2025, a Chinese hedge fund would release DeepSeek R1, a r… DeepSeek was immediately memory-holed by every major AI company, none of whom decided to read the papers published by DeepSeek about how the model was trained, other than OpenAI vaguely suggesting it might sue the company because ChatGPT outputs were used to train DeepSeek R1 and v3. Nevertheless, nobody in America seemed to bother to learn from DeepSeek, and instead set on a path to build even bigger, more huger models, such as GPT-4.5 ($400 million to train) and Claude Opus 4. Otherwise, “progress” has been thin, as have new or meaningful products since GPT-4’s release in 2023, other than OpenAI’s awkward and expensive video generation app Sora, which I’ll get to later. Models have got “better” in some indeterminate way based on benchmarks that are either rigged in their favor or that they cheat on. 2025 was not the year of agents, as they get things wrong 70% of the time according to Carnegie Mellon University, and have a 40% fail rate on single step queries (literally asking a question and getting an answer) and a 65% fail-rate on multi-step queries (as in “go and look this up and then do something else”) according to a study by Salesforce. In fact, Large Language Models are basically in the same place they were a year ago, except now everybody is anxious about the AI bubble, because nothing has really happened other than companies raising then burning hundreds of millions or billions of dollars.
November 14, 2025 at 7:00 PM
penpenguin2023.bsky.social
『ディープシークの異例の警鐘が注目を集めたのは、社会の安定と経済的正当性が密接に結びついている中国において、AIによる雇用崩壊が極めて政治的な問題となるためだ。』

【コラム】AI革命成功の証しは雇用消滅、DeepSeekが警鐘
www.bloomberg.co.jp/news/article...
【コラム】AI革命成功の証しは雇用消滅、DeepSeekが警鐘－トーベック
中国の人工知能（AI）スタートアップ、DeepSeek（ディープシーク）は、国際的影響力が非常に大きいにもかかわらず、あまり情報を発信しない。長い「提言」マニフェストを発表したり、国際会議で幹部を前面に出したりしない。梁文鋒最高経営責任者（CEO）が最後に公の場に姿を見せたのは、2月の習近平国家主席との会談時だった。それ以来、同社は主要な業界会合のほとんどを欠席している。
www.bloomberg.co.jp
November 14, 2025 at 10:21 AM
ed3d.net
almost every new thing I know of supports, if not mainlines, deepseek and GLM-4.5/4.6. because they cost essentially nothing
November 18, 2025 at 4:05 PM