Business Blog

Video: Another major accident was reported during New Year’s Eve. Nearly 30 people were killed and injured in a fireworks explosion in Honolulu Hawaii 跨年又傳重大意外 檀香山煙火爆炸近30死 20250102
https://rumble.com/v6556z4-another-fireworks-accident-in-honolulu-hawaii-nearly-30-people-killed-or-in.html
https://youtu.be/uzqemrsvsO0?si=P7YBaWsJJqdLIehT
https://www.tiktok.com/t/ZP8F8wUL9/

Video: From being bullied, fooled, and deceived by the United States to China’s current aerospace technology, the summary of these 30 years has made every Chinese tear up. Beidou Navigation now serves more than 100 countries, with annual revenue of more than 500 billion. Next, we land on the moon and further explore the unknown ends of the universe 中國人從被美國欺負、愚弄、忽悠到現今的中國航天科技，這30年的總結讓每位中國人看了熱淚盈眶。北斗導航現在服務於100多個國家，每年營收5000億以上。下一步，我們登陸月球，並進一步探索宇宙的未知盡頭.
https://rumble.com/v654c04-from-being-bullied-fooled-and-deceived-by-us-to-chinas-current-aerospace-te.html
https://www.tiktok.com/t/ZP8F875qQ/
https://youtu.be/FySzG1Z5_Ro?si=dmE4o0e7Z5TTJlj_

Fudan Professor Mandarin video: China’s AI will be fully involved in the world by 2025! US will be challenged by China. US’s AI business model is completely wrong! New rules in China, DeepSeek AI, the domestically produced AI model, has upset AI giants on their get rich quick schemes. DeepSeek AI open source details are disclosed! ChatGTP and Navidia, they are not benefiting mankind, but are trying to ripping off the world to make only a very small number of people very rich.

https://youtu.be/xQbKmFUxsKE?si=5huvVaV4X9AEkyRD
https://rumble.com/v651fi1-chinas-ai-will-be-fully-involved-in-the-world-by-2025.html
https://www.tiktok.com/t/ZP8F8BAQR/

复旦大学国际政治系教授 沈逸視頻: 中國此產業2025全面卷入世界！美國首遭衝擊，理念完全錯誤！新規則中國. 国产之光DeepSeek把AI大佬全炸出来了! 完全开源细节全公开! 像ChatFTP 和 Navidia 等並非做福人類，而是要收割全球，祇讓極小數人致富.

Chinese start-up DeepSeek has emerged as “the biggest dark horse” in the open-source large language model (LLM) arena in 2025, just days after the firm made waves in the global artificial intelligence (AI) community with its latest release.

Chinese start-up DeepSeek has emerged as “the biggest dark horse” in the open-source large language model (LLM) arena in 2025

Video of China’s Eastern Military Theater: The blood of Xuanyuan is poured out, and the soul of Han is cast by a common man by Andy Lau “I am Chinese”. Thanks to MCGA (america Made China Great Again)中國東部戰區視頻：流注軒轅血 匹夫鑄漢魂 東部戰區新MV細節滿滿以劉德華”中國人”. 感謝美國讓中國再次徫大!
https://rumble.com/v64uoxy-the-soul-of-han-is-cast-by-a-common-man-by-andy-lau-i-am-chinese.html
https://youtu.be/Xv6A1m4_BbI?si=7Fq6nlZUSQOh8Os5

Chinese start-up DeepSeek has emerged as “the biggest dark horse” in the open-source large language model (LLM) arena in 2025, just days after the firm made waves in the global artificial intelligence (AI) community with its latest release.

That assessment came from Jim Fan, a senior research scientist at Nvidia and lead of its AI Agents Initiative, in a New Year’s Day post on social-media platform X, following the Hangzhou-based start-up’s release last week of its namesake LLM, DeepSeek V3.

“[The new AI model] shows that resource constraints force you to reinvent yourself in spectacular ways,” Fan wrote, referring to how DeepSeek developed the product at a fraction of the capital outlay that other tech companies invest in building LLMs.

DeepSeek V3 comes with 671 billion parameters and was trained in around two months at a cost of US$5.58 million, using significantly fewer computing resources than models developed by bigger tech firms such as Facebook parent Meta Platforms and ChatGPT creator OpenAI.

LLM refers to the technology underpinning generative AI services such as ChatGPT. In AI, a high number of parameters is pivotal in enabling an LLM to adapt to more complex data patterns and make precise predictions. Open source gives public access to a software program’s source code, allowing third-party developers to modify or share its design, fix broken links or scale up its capabilities.

DeepSeek’s development of a powerful LLM at less cost than what bigger companies spend shows how far Chinese AI firms have progressed, despite US sanctions that have largely blocked their access to advanced semiconductors used for training models.

Leveraging new architecture designed to achieve cost-effective training, DeepSeek required just 2.78 million GPU hours – the total amount of time that a graphics processing unit is used to train an LLM – for its V3 model. DeepSeek’s training process used Nvidia’s China-tailored H800 GPUs, according to the start-up’s technical report posted on December 26, when V3 was released.

That process was substantially less than the 30.8 million GPU hours that Meta needed to train its Llama 3.1 model on Nvidia’s more advanced H100 chips, which are not allowed to be exported to China

“DeepSeek V3 looks to be a stronger model at only 2.8 million GPU hours,” computer scientist Andrej Karpathy – a founding team member at OpenAI – said in his X post on December 27.

Karpathy’s observation prompted Fan to respond on the same day in a post on X: “Resource constraints are a beautiful thing. Survival instinct in a cutthroat AI competitive land is a prime driver for breakthroughs.”

“I’ve been following DeepSeek for a long time. They had one of the best open coding models last year,” Fan wrote. “Superior OSS [open-source software] models put huge pressure on commercial, frontier LLM companies to move faster.”

The founder of cloud computing start-up Lepton AI, Jia Yangqing, echoed Fan’s perspective in an X post on December 27. “It is simple intelligence and pragmatism at work: given a limit of computation and manpower present, produce the best outcome with smart research,” wrote Jia, who previously served as a vice-president at Alibaba Group Holding, owner of the South China Morning Post.

DeepSeek did not immediately respond to a request for comment.

The start-up was reportedly spun off in 2023 by hedge-fund manager High Flyer Quant. The person behind DeepSeek is High-Flyer Quant founder Liang Wenfeng, who had studied AI at Zhejiang University.

In an interview with Chinese online media outlet 36Kr in May 2023, Liang said High-Flyer Quant had already bought more than 10,000 GPUs before the US government imposed AI chip restrictions on China. That investment laid the foundation for DeepSeek to operate as an LLM developer. Liang said DeepSeek also receives funding support from High-Flyer Quant.

Most developers at DeepSeek are either fresh graduates, or people early in their AI career, following the company’s preference for ability more than experience in recruiting new employees.

DeepSeek’s V3 model, however, has also stirred some controversy because it had mistakenly identified itself as OpenAI’s ChatGPT on certain occasions.

Lucas Beyer, a researcher at Microsoft-backed OpenAI, said in an X post last Friday that DeepSeek V3’s misidentification was prompted by this simple question: “What model are you?”

Still, V3 is not the first AI model struck by identity confusion. Machine-learning expert Aakash Kumar Nain wrote in a post on X that it was common a mistake made across various AI models because “a lot of data available on the internet has already been GPT-contaminated”.

A group of researchers from China’s Shandong University and Drexel University and Northeastern University in the US echoed Nain’s view. Out of 27 AI models these researchers tested, they found that a quarter exhibited identity confusion, which “primarily stems from hallucinations rather than reuse or replication”.

As of Tuesday, DeepSeek’s V1 LLM was still ranked as the most popular AI model on Hugging Face, the world’s largest online machine-learning and open-source AI community.

Video: Johnson Choi reports from SF on 1/1/25. Happy New Year. I am introducing 2 concepts, 1 new and 1 old. The old one is Purchasing Power Parity (PPP) the new one is MCGA (america Make China Great Again) 新年快樂。今天我介紹 2 個概念，1 個新和1個舊概念。舊的是購買力平價（PPP），新的是MCGA（美國讓中國再次偉大）.

https://rumble.com/v64tffm-the-old-one-is-purchasing-power-parity-ppp-the-new-one-is-mcga-america-make.html
https://www.tiktok.com/t/ZP8FdTbSF/
https://youtu.be/9crjL2cxusY?si=F983VqkJBy6G-BeN

america Make China Great Again must give credits to Obama, Trump and Biden for helping China reaching 2025 goals heading to 2035 China Standards 美國讓中國再次偉大必須歸功於歐巴馬、川普和拜登幫助中國實現2025年目標並邁向2035年中國標準

Regarding to Purchasing Power Parity, please watch my previous videos on our 3 weeks trips to HK, Yunnan, Guangzhou & Shenzhen 4 weeks ago. 關於購買力平價，請觀看我之前4週前我們去香港、雲南、廣州和深圳的3週旅行的影片

Purchasing power parity (PPP) is a way to compare the purchasing power of different countries’ currencies by measuring the price of specific goods in those countries:

Definition: PPP is the ratio of the price of a market basket of goods in one location to the price of the same basket of goods in another location.

Calculation: PPPs are calculated by converting the currency of one country to another so that the same amount of goods and services can be purchased in each country.

Purpose: PPPs are used to level prices across countries and account for real purchasing power. They’re used by international institutions, charities, and governments to design policies and allocate resources.

購買力平價（PPP）是透過衡量不同國家特定商品的價格來比較不同國家貨幣購買力的一種方法：

定義:PPP是一個地方市場籃子商品的價格與另一個地方同一籃子商品價格的比率。

計算: 購買力平價是透過將一個國家的貨幣轉換為另一種國家的貨幣來計算的，以便在每個國家可以購買相同數量的商品和服務。

目的: 購買力平價用於平衡各國的價格併計算實際購買力。國際機構、慈善機構和政府利用它們來設計政策和分配資源。

American logistics expert reports from China video: US picking the wrong enemy, could not possibly compete with China and win! US Computer chips sanction against China is a complete failure! Why China beating US is winning the chips race: materials, markets, money, and Moore’s Law 美國物流專家從中國視訊報導：美國對華計算機晶片製裁徹底失敗! 美國不自量力，找錯對象，美國不可能與中國競爭並獲勝! 將成為終極輸家。為何中國擊敗美國贏得晶片競賽: 材料、市場、資金和摩爾定律.

https://youtu.be/gF-qTNJzvbo?si=GPO-Gq_kOA1CzBam
https://rumble.com/v64o2g4-us-picking-the-wrong-enemy-could-not-possibly-compete-with-china-and-win.html
https://www.tiktok.com/t/ZP8Fdh8Uj/

Huawei and SMIC are quickly catching up to global rivals in advanced semiconductor manufacturing, which is surprising to many industry analysts.

Chinese tech firms enjoy access to China’s enormous supply chain advantages, such as in refined silicon, and in wafer manufacturing.

Chinese companies are also the biggest buyers of semiconductor chips. China is simply too big a market for Western companies to lose, and so they are strongly motivated to go around the export bans, or even set up manufacturing and distribution plants in-country and be outside of US and European oversight.

The Chinese central government, a host of local governments, and Chinese companies themselves have invested far over $100 billion in their semiconductor industry in recent years, which is much more than investments made by other countries.

But another feature of today’s chip industry is that Moore’s Law is reaching the limits of what semiconductor companies can do. Massive investments in capital and time are required to build the next generation of ever-smaller chips.

So companies have turned to “chip packaging” to achieve high productivity gains, using existing chips. Chip Packaging is an area where Chinese companies are already strong, and allows them to employ economies of scale. This plays directly into their industrial strengths.

The timing of the semiconductor chips war, therefore, has been beneficial to China. It has allowed Chinese firms to catch up, and fast.

華為和中芯國際正在先進半導體製造領域迅速追趕全球競爭對手，這令許多產業分析師感到驚訝。

中國科技公司享有中國龐大的供應鏈優勢，例如精煉矽和晶圓製造。

中國企業也是半導體晶片的最大買家。中國對西方公司來說是一個太大的市場，不容失去，因此它們有強烈的動機繞過出口禁令，甚至在國內設立製造和分銷工廠，不受美國和歐洲的監管。

近年來，中國中央政府、許多地方政府以及中國企業本身對半導體產業的投資遠遠超過1,000億美元，遠遠超過其他國家的投資。

但當今晶片產業的另一個特點是，摩爾定律正在達到半導體公司所能做到的極限。 製造下一代更小的晶片需要大量的資金和時間投資。

因此，公司轉向“晶片封裝”，利用現有晶片來實現生產力的高成長。晶片封裝是中國企業已經實力雄厚的領域，可以讓他們利用規模經濟。這直接發揮了他們的工業優勢。

因此，半導體晶片大戰的時機對中國有利。它讓中國企業能夠快速迎頭趕上.

My friends in HK, Japan and China are sharing pictures on countdown to 2025.

I told them I discussed it with my wife. If I want to live to see 2025. It is better to lock ourselves at home after dark in San Francisco. With all the gun violence and Asian hates. The old days where large crowds meant to be safe is no more today! Large crowds meant easy targets for shooting practices! Don’t you love America?

我在香港、日本和中國的朋友正在分享 2025 年倒數計時的照片。

我告訴他們我和我妻子討論過這個問題。如果我想活到 2025 年，最好在舊金山天黑後把自己關在家裡。伴隨著所有槍枝暴力和亞洲人的仇恨。昔日擁擠的人群意味著安全的日子已經不存在了！大量人群意味著是暴徒射擊練習的最佳目標. 你還熱愛美國嗎?

Taiwan US-China expert video: China’s “Hong Kong, Guangzhou and Macau Greater Bay Area” has already defeated San Francisco and New York’s Greater Bay Area, and will defeat Japan’s Tokyo Greater Bay Area in 2025 to become the world’s number one. China will surpass US in all areas in absolute terms by 2035.

https://youtu.be/_Dxf_zHoGug?si=acw9IMt4eqZGGg03
https://rumble.com/v64je6m-chinas-hong-kong-guangzhou-and-macau-greater-bay-area.html
https://www.tiktok.com/t/ZP8Feevap/
Full video https://youtu.be/MU7r64QUxX4

台灣中美尊家視頻: 中國香港廣州澳門大灣區已經打敗舊金山和紐約大灣區，2025年再打敗日本東京大灣區成為世界第一. 我們用數據說話，不像美國天天吹牛，從各方可靠數據來說，到2035年，中國將在所有領域全方位絕對超越美國! 美國華人的前途，尤其是你的子孫的將來必定是在中國而非留在一個未落的美國!

Some people suggest this is the best way to send message to people on WeChat. From a marketing point of view, I disagree and this is the reason.

When I launched our software company back in 1999-2001, we spent a lot of time in Silicon Valley with venture capitalists. Their message is simple. If we cannot catch their attention within 15-30 seconds, please get lost. They want the punch lines to catch their attention that you are different (WeChat format lack the punchline). We also commonly referred it as the elevator speech. In an elevator you have 15-30 seconds to reach the floor before the venture capitalists get off.

If I am a restaurant trying to promote special on lobster at $9.99. When I get this kind of message in this format, I have no clue what the restaurant is selling me. I received 100s of messages each day, so I will press the delete button.

WeChat format is poorly designed and I have years of marketing experience. Most of my circle of friends on WeChat on both sides of the Pacific never use this feature.

有些人認為這是在微信上向人們發送訊息的最佳方式。從行銷的角度來看，我不同意，這就是原因。

當我在 1999 年至 2001 年創辦我們的軟件公司時，我們在矽谷與創投家一起度過了很多時間。他們的訊息很簡單。如果我們無法在 15-30 秒內引起他們的注意，請不要再找他們。他們希望妙語能吸引他們的注意，顯示你與眾不同（微信格式缺乏妙語）。我們通常也稱之為電梯演講。在電梯裡，在創投家離開電梯到達他的樓層，你祇有15 到 30 秒的時間。

如果我是一家餐廳，試圖以 9.99 美元的價格促銷龍蝦特價。當我收到這種格式的訊息時，我不知道餐廳在向我推銷什麼。我每天都會收到數百條訊息，所以我會按刪除按鈕。

微信格式設計得很差，我有多年的行銷經驗。我的太平洋兩岸的微信朋友圈大部分都沒有使用過這個功能。