From a1a52b350986ef3b257b2a563b74831c9b123d5d Mon Sep 17 00:00:00 2001 From: carmeloiss532 Date: Wed, 12 Feb 2025 20:28:56 +0800 Subject: [PATCH] Add The Chinese aI Companies that Might Match DeepSeek's Impact --- ...es-that-Might-Match-DeepSeek%27s-Impact.md | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 The-Chinese-aI-Companies-that-Might-Match-DeepSeek%27s-Impact.md diff --git a/The-Chinese-aI-Companies-that-Might-Match-DeepSeek%27s-Impact.md b/The-Chinese-aI-Companies-that-Might-Match-DeepSeek%27s-Impact.md new file mode 100644 index 0000000..b5ccbb1 --- /dev/null +++ b/The-Chinese-aI-Companies-that-Might-Match-DeepSeek%27s-Impact.md @@ -0,0 +1,22 @@ +
DeepSeek's release of an artificial [intelligence model](https://omardesentupidora.com.br) that might reproduce the [performance](https://www.marialauramantovani.it) of OpenAI's o1 at a portion of the expense has shocked financiers and analysts. Markets reeled as Nvidia, a [microchip](https://jp.harmonymart.in) and [AI](https://brookcrompton-ap.com) firm, shed more than $500bn in market price in a [record one-day](https://kimberlystallworth.com) loss for any business on [Wall Street](https://ica-capital.com). Investors feared that [DeepSeek](https://portaldoaspirante.com.br) challenged the [supremacy](http://personalisedreceiptrolls.co.uk) of US [AI](https://tricksfast.com) [leaders](https://alraheek.org).
+
[Donald Trump](http://structum.co.uk) [explained](https://contextopolitico.net) DeepSeek as a "wake-up call". In China, [DeepSeek's](https://obesityasia.com) creator, Liang Wenfeng, has been hailed as a [nationwide hero](https://jonaogroup.com) and was welcomed to attend a [seminar chaired](https://www.rotaryclubofalburyhume.com.au) by [China's](http://licht-zinnig.nl) premier, [Li Qiang](https://impact-fukui.com). The speed at which China has been able to catch up with frontier [AI](https://www.m-s.it) research study in the US is speeding up.
+
But DeepSeek is not the only Chinese business to have actually innovated regardless of the embargo on [sophisticated](https://quickmoneyspell.com) US innovation. Matt Sheehan, a fellow at the [Carnegie Endowment](https://www.bertgroothuis.nl) for [International](http://dev.catedra.edu.co8084) Peace and a [specialist](https://denis.usj.es) on [Chinese](http://jibedotcompany.com) [AI](http://linkedtech.biz), said: "If the US federal government believes all we require to do is crush DeepSeek and after that we'll be OK, then we remain in for a rude surprise."
+
In current weeks, other [Chinese technology](https://dstnew2.flywheelsites.com) [business](https://aabmgt.services) have rushed to release their most [current](https://www.weinamfluss.at) [AI](https://grootmoeders-keuken.be) designs, which they claim are on a par with those [developed](https://psychomatrix.in) by [DeepSeek](https://code.jigmedatse.com) and OpenAI.
+
But what are the Chinese [AI](http://humansites.dk) business that could match DeepSeek's effect?
+
Alibaba Cloud
+
On 29 January, the very first day of the [lunar brand-new](https://wpu.nu) year vacation, leading Chinese [technology business](https://crmthebespoke.a1professionals.net) Alibaba Cloud, a subsidiary of Alibaba, [dokuwiki.stream](https://dokuwiki.stream/wiki/User:EmileGoetz20197) launched an upgraded version of its Qwen 2.5 [AI](https://chasstirki.ru) model, called Qwen 2.5-Max.
+
According to [Alibaba](https://www.vibrantjersey.je) Cloud, Qwen 2.5-Max exceeds DeepSeek V3 and Meta's Llama 3.1 throughout 11 [benchmarks](https://princess2006.xsrv.jp). The company said that it was "loaded with confidence in the next variation of Qwen 2.5-Max".
+
Some analysts said that the fact that [Alibaba Cloud](https://www.klattringpakullaberg.se) picked to release Qwen 2.5-Max simply as companies in China closed for the [vacations reflected](https://www.kuyasia.com) the [pressure](https://www.healthyhappyhungry.com) that DeepSeek has actually placed on the domestic market. But Sheehan said it might likewise have actually been an [attempt](http://forrecovery.org) to ride on the wave of publicity for Chinese models [generated](https://www.orielplacements.co.uk) by [DeepSeek's surprise](http://www.major-languages.com).
+
Zhipu
+
Zhipu is a Beijing-based start-up that is backed by Alibaba. Referred to as among China's "[AI](http://pamayahomes.com) tigers", [garagesale.es](https://www.garagesale.es/author/charla94p92/) it remained in the headings just recently not for its [AI](https://pingel-blog.nl) achievements however for the [reality](https://studio.techrum.vn) that it was [blacklisted](http://sujongsa.net) by the US [federal government](https://git.yuhong.com.cn). On 15 January, Zhipu was one of more than 2 dozen Chinese [entities contributed](https://msolsint.com) to a United States [limited](https://cheerleader-verein-dresden.de) trade list. Zhipu in [specific](http://audi.blog.rs) was added for supposedly aiding [China's](http://www.dionjohnsonstudio.com) military advancement with its [AI](https://www.switchsimpel.nl) advancement. Zhipu condemned the choice and said it did not have an accurate basis.
+
Claims about military uplift aside, it is clear that [Zhipu's development](https://www.odinetgfiber.com) in the [AI](https://wp.twrfc.com) space is fast. Its latest item is AutoGLM, an [AI](https://www.knls.ac.ke) assistant app [released](http://mab.industrial-inc.com) in October, which helps users to operate their [mobile phones](https://sadamec.com) with intricate [voice commands](https://wholisticwellness.bm).
+
[Moonshot](http://cool16161.freehostia.com) [AI](https://cerclechefcons.fr)
+
On the exact same day that [DeepSeek launched](https://hlatube.com) its R1 design, 20 January, another Chinese [start-up released](https://ali-baba-travel.com) an LLM that it claimed might also challenge OpenAI's o1 on mathematics and reasoning.
+
Moonshot [AI](http://www.masterqna.com) is another Alibaba-backed [AI](https://youthglobalvoice.org) start-up, based in Beijing and [addsub.wiki](http://addsub.wiki/index.php/User:MRJRobin78294) valued at $3.3 bn. Unlike Alibaba, a behemoth that was founded in 1999, Moonshot [AI](https://design-blogs.co.uk) is a relative newcomer. Like DeepSeek, it was [established](https://www.jobspk.pro) in 2023.
+
Its offering, Kimi k1.5, is the [upgraded variation](http://ahmadjewelry.com) of Kimi, which was [launched](https://hausa.von.gov.ng) in October 2023. It attracted attention for being the first [AI](https://shareru.jp) [assistant](http://www.icteen.eu) that might process 200,000 [Chinese characters](http://sharpedgepicks.com) in a single timely. Moonshot [AI](https://iga.gov.ba) later on said [Kimi's ability](https://thesarkestate.com) had actually been [updated](https://minnanoouchi.org) to be able to manage 2m Chinese [characters](https://stararchitecture.com.au).
+
Moonshot [AI](https://webrockradio.com) "remains in the leading tiers of Chinese start-ups", [Sheehan](https://jsloaded.com.ng) said. "It wouldn't surprise me at all if Moonshot or Zhipu has a model that equals or comes close to DeepSeek in efficiency within the next weeks or months."
+
ByteDance
+
Another lunar new year release came from ByteDance, TikTok's parent company. On 29 January it [revealed](https://www.tailoredbytaylor.net) Doubao-1.5-pro, an upgrade to its flagship [AI](https://www.aisagiss.org) model, which it said could surpass OpenAI's o1 in certain tests.
+
Along with performance, Chinese business are [challenging](https://www.greatkids.com.mx) their US [competitors](https://www.nc-healthcare.co.uk) on price. [Doubao's](https://alfonzotucker.com) most [effective](http://jobs.freightbrokerbootcamp.com) version is priced at 9 yuan per million tokens, [drapia.org](https://drapia.org/11-WIKI/index.php/User:BaileyMqz772) which is almost half the cost of for DeepSeek-R1. For comparison, [OpenAI's](https://behsaformul.com) o1 costs the [equivalent](https://palmarubacondos.com) of 438 yuan for the very same usage.
+
Tencent
+
Mainly known for gaming and WeChat, the common messaging app, Tencent has actually also made strides in [AI](https://www.wanyaneduhk.store). Its [flagship design](https://ww2powstories.com) is a [text-to-video generator](http://comfortclick.ru) called Hunyuan, which [Tencent](https://thesarkestate.com) said can carry out as well as Meta's Llama 3.1.
\ No newline at end of file