Chinese Lab's AI LLM Performance Shocks Silicon Valley
A Chinese Lab has sparked panic in Silicon Valley with the release of its first AI model that can outperform America's best despite being built more cheaply and with less-powerful chips, according to the US media reports. The lab called DeepSeek has recently unveiled a free, open-source large-language model (LLM) that it says took only two months and $5.5 million million to build, using reduced-capability chips from Nvidia called H800s. By comparison, the US-based OpenAI's closed LLM model cost $100 million to develop and train using the most advanced H100 chips from Nvidia. Open-source and free DeepSeek models can significantly help developing nations like Pakistan by providing affordable access to the latest AI technology, allowing them to develop solutions tailored to their specific needs without high costs.
DeepSeek, a small startup lab in China, has accomplished this feat despite the US technology export controls to slow down China's AI efforts. Former Google CEO Eric Schmidt is now acknowledging that China has narrowed or closed the AI technology gap with the United States.
In 2022 America banned the export of advanced chips to China, according to Economist magazine. Nvidia, a leading chipmaker, has had to design special downgrades to its products for the Chinese market. America has also sought to prevent China from developing the capacity to manufacture top-of-the-line chips at home, by banning exports of the necessary equipment and threatening penalties for non-American firms that might help, too.
The slower H800 chip was created by Nvidia to comply with export regulations that prevent the chipmaker from selling its high-end GPUs to China. Apparently, the limits imposed by Washington on Chinese engineers' access to the most advanced Nvidia chips forced them to develop a much more efficient model to achieve the same performance as their US counterparts. Other Chinese tech companies ranging from Alibaba and Huawei to TenCents are also working on their own multiple AI models, including LLMs.
DeepSeek has emerged from High-Flyer, a Chinese hedge fund started by 40-year-old Liang Wengfeng in 2015 to use AI to gain an edge in stocks-trading. Conducting fundamental research helped High-Flyer become one of the biggest quant funds in the country, according to The Economist magazine.
Related Links:
Pakistani Military Launches Defense AI Program
Algorithm: Origins of Artificial Intelligence in Islamic Age
Digital Pakistan 2022: Broadband Penetration Soars to 90% of 15+ Population
STEM Enrollment in Pakistan Exceeds One Million
Digital Public Infrastructure in Pakistan
Generative AI Buzz in Pakistan
Is Pakistan Ready for the AI Revolution?
Growing Presence of Pakistani Women in Science and Technology
Riaz Haq's Youtube Channel
Comments
https://macropolo.org/interactive/digital-projects/the-global-ai-talent-tracker/
China and Pakistan are deepening their collaboration in artificial intelligence (AI) through a multifaceted partnership that spans education, industry, research, and defense. Here's a structured overview of their cooperation:
Educational and Research Collaborations:
University Partnerships: Institutions like Pakistan's National University of Sciences and Technology (NUST) and COMSATS collaborate with Chinese universities such as Tsinghua University and the Chinese Academy of Sciences. Joint research centers and labs focus on AI, machine learning, and robotics.
Scholarships and Training: China offers scholarships for Pakistani students and professionals to pursue AI-related studies and training in China, enhancing Pakistan's technical expertise.
Economic and Industrial Initiatives:
CPEC and Digital Economy: Under the China-Pakistan Economic Corridor (CPEC), AI is integrated into digital infrastructure projects, including smart cities (e.g., Lahore and Islamabad) and industrial automation. Chinese tech giants like Huawei, Zong, and Alibaba contribute through cloud computing, 5G networks, and AI-driven solutions.
Startups and Innovation: Chinese venture capital funds and tech incubators support Pakistani AI startups, fostering innovation in sectors like fintech, agriculture, and healthcare.
Defense and Security:
Military AI Applications: Collaboration extends to defense, with joint ventures in autonomous systems, surveillance drones, and cybersecurity. Projects like the Hangor-class submarines incorporate AI technologies for enhanced operational capabilities.
Policy and Strategic Development:
National AI Policy: Pakistan’s AI policy framework is influenced by Chinese expertise, aiming to build a skilled workforce and ethical AI governance. Bilateral dialogues, such as the China-Pakistan Joint Committee on Science and Technology, guide strategic priorities.
Conferences and Knowledge Exchange:
Workshops and Forums: Regular events like the China-Pakistan AI Cooperation Forum facilitate knowledge sharing, showcasing projects in healthcare (e.g., AI-based disease diagnostics) and agriculture (e.g., precision farming tools).
Challenges and Considerations:
Infrastructure and Skills: Pakistan faces challenges in infrastructure and talent, which Chinese investments aim to address. However, concerns around data privacy and geopolitical implications (e.g., balancing relations with other powers) persist.
This cooperation underscores the "Iron Brothers" synergy, leveraging China’s technological prowess and Pakistan’s strategic needs, while positioning both nations competitively in the global AI landscape.
https://youtu.be/WEBiebbeNCA?si=9nPOIhy5jt6r29x7
A little-known AI lab out of China has ignited panic throughout Silicon Valley after releasing AI models that can outperform America’s best despite being built more cheaply and with less-powerful chips. DeepSeek, as the lab is called, unveiled a free, open-source large-language model in late December that it says took only two months and less than $6 million to build. The new developments have raised alarms on whether America’s global lead in artificial intelligence is shrinking and called into question big tech’s massive spend on building AI models and data centers. In a set of third-party benchmark tests, DeepSeek’s model outperformed Meta’s Llama 3.1, OpenAI’s GPT-4o and Anthropic’s Claude Sonnet 3.5 in accuracy ranging from complex problem-solving to math and coding. CNBC’s Deirdre Bosa has the story. This video also includes Bosa’s full interview with Perplexity CEO Aravind Srinivas.
@rowancheung
NEWS: DeepSeek just dropped ANOTHER open-source AI model, Janus-Pro-7B.
It's multimodal (can generate images) and beats OpenAI's DALL-E 3 and Stable Diffusion across GenEval and DPG-Bench benchmarks.
This comes on top of all the R1 hype. The 🐋 is cookin'
https://x.com/rowancheung/status/1883917681642070282
https://www.reuters.com/technology/artificial-intelligence/alibaba-releases-ai-model-it-claims-surpasses-deepseek-v3-2025-01-29/
BEIJING, Jan 29 (Reuters) - Chinese tech company Alibaba (9988.HK), opens new tab on Wednesday released a new version of its Qwen 2.5 artificial intelligence model that it claimed surpassed the highly-acclaimed DeepSeek-V3.
The unusual timing of the Qwen 2.5-Max's release, on the first day of the Lunar New Year when most Chinese people are off work and with their families, points to the pressure Chinese AI startup DeepSeek's meteoric rise in the past three weeks has placed on not just overseas rivals, but also its domestic competition
"Qwen 2.5-Max outperforms ... almost across the board GPT-4o, DeepSeek-V3 and Llama-3.1-405B," Alibaba's cloud unit said in an announcement posted on its official WeChat account, referring to OpenAI and Meta's most advanced open-source AI models.
The Jan. 10 release of DeepSeek's AI assistant, powered by the DeepSeek-V3 model, as well as the Jan. 20 release of its R1 model, has shocked Silicon Valley and caused tech shares to plunge, with the Chinese startup's purportedly low development and usage costs prompting investors to question huge spending plans by leading AI firms in the United States.
But DeepSeek's success has also led to a scramble among its domestic competitors to upgrade their own AI models.
Two days after the release of DeepSeek-R1, TikTok owner ByteDance released an update to its flagship AI model, which it claimed outperformed Microsoft-backed OpenAI's o1 in AIME, a benchmark test that measures how well AI models understand and respond to complex instructions.
This echoed DeepSeek's claim that its R1 model rivalled OpenAI's o1 on several performance benchmarks.
The predecessor of DeepSeek's V3 model, DeepSeek-V2, triggered an AI model price war in China after it was released last May.
The fact that DeepSeek-V2 was open-source and unprecedentedly cheap, only 1 yuan ($0.14) per 1 million tokens - or units of data processed by the AI model - led to Alibaba's cloud unit announcing price cuts of up to 97% on a range of models.
Other Chinese tech companies followed suit, including Baidu (9888.HK), opens new tab, which released China's first equivalent to ChatGPT in March 2023, and the country's most valuable internet company Tencent (0700.HK)
Liang Wenfeng, DeepSeek's enigmatic founder, said in a rare interview with Chinese media outlet Waves in July that the startup "did not care" about price wars and that achieving AGI (artificial general intelligence) was its main goal.
OpenAI defines AGI as autonomous systems that surpass humans in most economically valuable tasks.
While large Chinese tech companies like Alibaba have hundreds of thousands of employees, DeepSeek operates like a research lab, staffed mainly by young graduates and doctorate students from top Chinese universities.
Liang said in his July interview that he believed China's largest tech companies might not be well suited to the future of the AI industry, contrasting their high costs and top-down structures with DeepSeek's lean operation and loose management style.
"Large foundational models require continued innovation, tech giants' capabilities have their limits," he said.
@jacksonhinklle
🚨🇨🇳🇺🇸 BREAKING: CHINA HAS DEFEATED AMERICA
🇨🇳 DeepSeeak BEATS OpenAI 🇺🇸
🇨🇳 BYD BEATS Tesla 🇺🇸
🇨🇳 Huawei BEATS Apple 🇺🇸
🇨🇳 Huawei BEATS US Telecoms 🇺🇸
🇨🇳 Alibaba BEATS Amazon 🇺🇸
🇨🇳 CATL BEATS US lithium-ions 🇺🇸
🇨🇳 TikTok BEATS Instagram 🇺🇸
🇨🇳 DJI BEATS all US drones 🇺🇸
🇨🇳 Temu/Shein BEAT US ecomm 🇺🇸
🇨🇳 BRI BEAT Build Back Better World 🇺🇸
🇨🇳 China is also BEATING the US in infrastructure, high speed rail, healthcare, education, space technology, manufacturing, defense technology, gold production, energy, housing affordability & annual GDP PPP.
https://x.com/jacksonhinklle/status/1884686222356079075
https://www.reuters.com/technology/india-it-minister-praises-deepseeks-low-cost-ai-compares-it-with-own-investment-2025-01-30/
Jan 30 (Reuters) - India's IT minister has praised Chinese startup DeepSeek for shaking up the sector with its low-cost AI assistant, likening its frugal approach to his government's efforts to build a localized AI model.
India announced a $1.25 billion AI investment in March, dubbed IndiaAI mission, which includes funding for AI startups and developing its own AI infrastructure.
"Some people question the amount of investments the government has committed in (IndiaAI mission). You have seen what DeepSeek has done? $5.5 million and a very very powerful model. Because, the use of brain," Ashwini Vaishnaw said on Tuesday at an event in the eastern state of Odisha.
DeepSeek has triggered a dramatic rethink on artificial intelligence spending around the world, claiming it took just two months and cost under $6 million to build an AI model using Nvidia's less-advanced H800 chips.
Downloads of its app recently surpassed OpenAI's ChatGPT on Apple's App Store, while the cost and performance of its tools upended industry beliefs that China was years behind U.S. rivals in the AI race.
Vaishnaw's statement appeared to target comments made by OpenAI's Sam Altman during a visit to India last year, when he cast doubt on the possibility of an Indian team being able to build a substantial model in the OpenAI space with a $10 million budget.
"The way this works is we're going to tell you it's totally hopeless to compete with us on training foundation models. You shouldn't try. And it's your job to like try anyway. And I believe both of those things," he said, comments which are now in focus again on online platforms such as X after DeepSeek's success.
Altman is due to visit India again on February 5, just as his company is currently locked in a court battle in the country with digital news and book publishers over copyright breaches.
https://www.cnbc.com/2025/01/30/cnbcs-inside-india-newsletter-how-chinas-deepseek-could-benefit-india.html
China’s DeepSeek released its model free for commercial use and also made public the technology know-how to build such a model from scratch. The company said it spent a mere $6 million in AI chip costs to develop the model.
Although some have raised questions about the exact figure, it compares to the hundreds of millions — and sometimes billions of dollars — spent by American technology firms.
The development could mark the start of AI model development within India, as previous methods for training large language models have required thousands of energy-intensive and expensive AI chips.
It could also mark a major turning point for Indian technology companies such as Infosys
, which have previously had to rely on AI models created by U.S.-based tech firms, such as Meta
’s Llama.
Keshav Murugesh, chief executive of business transformation firm WNS
, said DeepSeek’s AI model is a “pivotal advancement” for Indian technology companies.
He suggested the low cost of development will enable new AI models to be trained in India’s regional languages and enable use cases previously deemed uneconomical. The vast majority of sophisticated large language models today, such as OpenAI’s o1 and Anthropic’s Claude Sonnet 3.5, can output text only in a handful of languages.
“By leveraging the innovations behind DeepSeek, these companies can significantly lower costs and accelerate their time to market,” Murugesh told CNBC. WNS
, which is listed on the NYSE, revealed in its third-quarter call earlier this month, that it will soon be enabling generative AI use cases at a U.S. insurance firm, a top 10 client for the company.
Industry surveys have shown that data privacy and the high cost of implementing large language models are among the reasons why enterprises have resisted AI adoption. DeepSeek-R1′s benefits, if confirmed, would swiftly eliminate two of the top 10 concerns and could begin to address a multitude of others.
The Indian government has also embarked on subsidizing access to AI chips, known as graphics processing units, to enable academics and start-ups in the country to begin developing AI models.
I am looking to evaluate if I can use it in a production environment instead of OpenAI 4o(my current production llm), and will take me a week of benchmarking and testing.
I always intended to move the production to own hosted llama-3.x for privacy, deterministic response time and cost, but at this point, deepseek is definitely in the running.
I love this (idiotic?) rivalry between OpenAI/Deepseek, or is it US vs. China. Entrepreneurs get a lot of goodies to build value while the giants duke it out on extremely expensive foundation models.
Opinion
Fareed Zakaria
DeepSeek has created a 21st-century Sputnik moment
Export controls may have helped turbocharge Chinese innovation.
https://www.washingtonpost.com/opinions/2025/01/31/deepseek-sputnik-competition-trade/
Is this a Sputnik moment? The world has reacted with astonishment to the release of a disruptive AI model from Chinese company DeepSeek, which appears to be able to perform as well as or, in some cases, better than ChatGPT and other cutting-edge models put out by U.S. companies. Americans had assumed their massive lead in funding, access to high-quality chips and innovation would keep them well ahead. That assumption now looks like hubris.
The episode is in some ways a much bigger deal than Sputnik. Sputnik was about the Soviet Union’s space program competing with that of the United States. Few thought the Soviet economy in general was more technologically advanced than America’s. But DeepSeek is a private Chinese company that demonstrated its stunning prowess on the cheap in the most important technology for the future. It’s not exactly clear just how much DeepSeek’s model actually cost, to what extent it needed to use U.S. models for training and whether there was any closet Chinese government help. But given the enormous efforts that the U.S. government has made over the past few years to preserve its advantage — chip bans, export controls, etc. — DeepSeek has made a remarkable achievement. It suggests to me two lessons and two questions.
The first lesson is that, over time, open artificial-intelligence systems are likely to outperform closed systems. (An open system is like Lego blocks with instructions; a closed one is the built Lego structure with the instructions kept secret.) Many have pointed out that DeepSeek used Meta’s open-source Llama model to train. It also used Qwen, a family of AI models, also open-source, put out by the Chinese technology giant Alibaba. While DeepSeek is currently the best, China’s big technology companies have been releasing a number of AI models, mostly open-source, that are getting better and better. If the history of technology is any guide, the ability to see the innards of these models and understand their reasoning should lead to greater and faster technological innovation than using closed models that others cannot use for collaboration.
Second, constraints can be useful, as former Intel CEO Pat Gelsinger has noted. Just as art sometimes flourishes in repressive environments, in which restrictions force artists to be creative, so also engineers often operate best under constraints. Forced to use second-tier chips, Chinese engineers produced creative workarounds. This is not just true with DeepSeek. In 2023, Chinese telecommunications giant Huawei released a smartphone with a seven-nanometer chip, a kind that had been explicitly banned by U.S. export controls. There is some evidence that, after years of sclerosis, China’s chipmakers have responded to U.S. bans by becoming much more innovative.
Fareed Zakaria
DeepSeek has created a 21st-century Sputnik moment
Export controls may have helped turbocharge Chinese innovation.
https://www.washingtonpost.com/opinions/2025/01/31/deepseek-sputnik-competition-trade/
In a fascinating interview last year, Liang Wenfeng, the CEO of DeepSeek, argued that his engineers were more motivated by doing research than making money, and appeared to contrast that attitude with the one prevalent in Silicon Valley, which is all about maximizing revenue, providing cloud services and generating cash flow. Demis Hassabis, who leads Google’s DeepMind and also shared the 2024 Nobel Prize in chemistry for AI-related scientific breakthroughs, is said to have fought to keep his team in London, far from Silicon Valley, so that it can focus on basic research.
The first question that DeepSeek raises is: Can the United States stop China from advancing along the technological frontier? Some argue that DeepSeek shows that export controls work: Its model needed many Nvidia chips, which it managed to procure before export bans were fully in place. Soon, China will not have access to the best chips and will suffer even more from the ban.
But as we have learned with the rounds and rounds of global sanctions against Russia, the world economy is large and porous. Stuff gets through. And China is not Russia. It is a vast, technologically sophisticated economy with millions of software developers and hundreds of high-quality firms in the technology space. Human talent on that scale will find ways to innovate, even if those measures keep China slightly behind.
The second question: What is the cost of this approach? If technology bans and export controls at best keep China behind a year — maybe just several months — is that gain worth the cost? That cost is Chinese retaliation, limiting the United States’ access to key materials that it needs for high technology. More important, a decoupled global economy also creates a closed ecosystem in which U.S. technology companies will not face competition from the best. Is Tesla going to innovate at the highest level if it is not facing its strongest Chinese rival?
A technology decoupling means that AI will become the central part of a new global arms race, totally unregulated and unconstrained, with the world’s two largest economies hurtling toward superintelligence no-holds-barred, and incorporating it into all military applications — including nuclear weapons. If artificial intelligence is as revolutionary a technology as predicted, having it unleashed in every realm of human life with absolutely no guardrails points to a scary future — one far more dangerous than anything people imagined because of the Sputnik satellite.
https://www.wsj.com/health/pharma/the-drug-industry-is-having-its-own-deepseek-moment-68589d70
It isn’t just artificial intelligence—Chinese biotechs are now developing drugs faster and cheaper than their U.S. counterparts
The biotech industry’s DeepSeek moment came last fall.
That is when Summit Therapeutics SMMT 2.34%increase; green up pointing triangle, backed by billionaire Bob Duggan, announced that its drug had outperformed Merck’s blockbuster therapy Keytruda in a head-to-head lung-cancer trial. Keytruda, a $30 billion-a-year immunotherapy juggernaut, is the bestselling drug in the pharma industry and has long dominated the market. So the prospect of a superior competitor was seismic. Even more remarkable: Summit had licensed the drug just two years earlier from a little known Chinese biotech called Akeso 9926 2.36%increase; green up pointing triangle.
The news added billions of dollars to Summit’s market capitalization, catapulting it into biotech’s upper ranks despite having no approved drugs. While Summit’s drug still hasn’t received U.S. regulatory approval, the results were a watershed moment for the industry, underscoring the competitive threat emanating from China.
China’s rise in biotech has been years in the making, but it is now impossible to ignore. In 2020, less than 5% of large pharmaceutical transactions worth $50 million or more upfront involved China. By 2024, that number had surged to nearly 30%, according to DealForma. A decade from now, many drugs hitting the U.S. market will have originated in Chinese labs
https://asia.nikkei.com/Business/Tech/Semiconductors/Top-China-chipmaker-SMIC-says-tariff-war-sparking-rush-orders
TAIPEI -- China's top chipmaker says clients are stockpiling inventory and asking to move up delivery of orders scheduled for later in the year due to concerns over U.S. tariffs and broader geopolitical tensions.
Zhao Haijun, co-CEO of Semiconductor Manufacturing International Corp., told investors and reporters on Wednesday that his company has seen a rush in orders in the first two quarters of 2025.
"We see many [customers] want to ship the products to their export destinations as soon as possible, so they are building up inventories for the second half or even the full year. ... They are hoping to prepare as many parts for their products in advance as soon as possible," Zhao said. "Currently the ocean shipping is very crowded to fulfill all these rush orders as well. ... But we don't know how long this could last, given the dynamics of global uncertainties."
U.S. President Donald Trump has started a broad tariff war against China and other nations, including imposing an additional 10% blanket tariff on Chinese imports.
Zhao said a rapid acceleration of local chip production has raised SMIC's revenue to a record level, but warned of a price war and tariff uncertainties in the second half of 2025.
Efforts to boost local production in critical areas, including automotive applications, are bearing fruit, with some chip products now being produced in large volumes.
China has instructed domestic carmakers, including BYD, to use up to 25% locally made automotive chips to boost local self-sufficiency, Nikkei Asia reported earlier.
The industry reoriented toward the domestic supply chain at "a relatively rapid pace," Zhao said. "Our Chinese customers' market share expanded, and revenue from these customers grew 34%, year over year."
Beijing's economic stimulus measures have also encouraged clients to restock chips used in products like smartphones, notebooks, electric vehicles and bicycles, he said. "Our 8-inch wafer capacity is running at full utilization, beating the traditional seasonality for the first quarter."
"However, we are worried that the rush orders for the first half of this year would result in soft demand in the second half, when many peers will have new capacity go online at the same time," Zhao said. "That means there might be intensified price competition to fight for orders later in the year. ... The trend for the price is unclear for the second half, but it will not go up."
SMIC estimates its 2025 annual revenue will grow 6% to 8%, higher than the industry average in the same markets.
It plans capital investment in 2025 on a similar level as last year's $7.3 billion, much higher than its foreign contract chipmaking peers, such as United Microelectronics Corp.
SMIC's revenue for the last quarter of 2024 rose 31.5% to $2.2 billion, with operating profit nearly doubling from a year earlier. However, its net profit declined 38.4% to $108 million compared with the same period the previous year. The chipmaker attributed this to higher expenses from the startup of new plants.
Its full-year revenue for 2024 was $8.03 billion, up 27% from 2023, with net profit down 45%.
American rival Globalfoundries reported a net loss on Wednesday for the quarter ended Dec. 31. Its full-year revenue came to $6.75 billion, with a net loss of $262 million.
Nearly 90% of SMIC's revenue from the last quarter of 2024 came from China. The figure for the full year was 85%.
https://engineerine.com/china-504-qubit-xiaohong-chip/
China has just shattered records in quantum computing with the launch of Tianyan-504, a superconducting quantum computer powered by the Xiaohong chip. With an unprecedented 504 qubits, this marks a major milestone in the global quantum race, putting China in direct competition with tech giants like IBM and Google.
Quantum computing is one of the most revolutionary advancements in technology, with the potential to solve complex problems exponentially faster than traditional computers. From AI to cryptography, medical research, and national security, this breakthrough is a game-changer.
How does Tianyan-504 compare to global competitors, and what does it mean for the future of computing? Let’s break it down.
The Tianyan-504 is China’s most powerful quantum computer to date, developed using superconducting quantum technology. It is powered by the Xiaohong chip, an advanced 504-qubit processor, making it one of the most powerful quantum machines in the world.
Key Features of Tianyan-504:
Record-Breaking 504 Qubits – China’s highest-qubit quantum processor to date.
Superconducting Quantum Technology – Uses extremely cold temperatures to minimize errors and improve qubit stability.
Cloud-Accessible – Available for remote use in 50+ countries via China Telecom.
Rivaling IBM and Google – Competing directly with IBM’s Eagle (433 qubits) and Google’s Sycamore (72 qubits).
With this cutting-edge technology, China has positioned itself as a dominant force in the global quantum race.
Unlike traditional computers that use binary bits (0s and 1s), quantum computers use qubits, which can exist in multiple states simultaneously due to superposition and entanglement. This allows quantum computers to perform massively parallel calculations, solving problems that classical computers would take thousands of years to process.
How the Xiaohong Chip Works:
Qubits Operate in Superposition – This allows them to perform multiple calculations at once.
Quantum Entanglement – Qubits communicate instantaneously, boosting computing efficiency.
Error Correction Mechanisms – Advanced superconducting circuits reduce quantum errors, making calculations more accurate.
Why Is This a Big Deal?
Key Implications of Tianyan-504:
AI and Machine Learning – Quantum computing could train AI models thousands of times faster, accelerating breakthroughs in artificial intelligence.
Cybersecurity and Cryptography – Quantum computers have the power to break traditional encryption, making current security protocols obsolete.
Drug Discovery and Healthcare – Quantum simulations could develop new medicines and treatments at an unprecedented pace.
Climate and Energy Solutions – Quantum algorithms can optimize energy grids, simulate climate models, and enhance battery technology.
National Security and Defense – Quantum advancements could give China an edge in intelligence, surveillance, and secure communications.
By leading in quantum computing, China is gaining a strategic advantage across multiple industries.
https://www.verdict.co.uk/china-restricts-high-tech-manufacturing-exports-india-southeast-asia/
The restrictions are impacting companies such as Taiwan-based Foxconn, a key Apple partner.
Officials reportedly have “verbally encouraged” regulatory authorities and local governments to limit technology transfers and equipment exports to these regions, said the source.
The aim is to strengthen China's own production capabilities, cut down potential job losses, and deter foreign investors from leaving if the US imposes new trade barriers.
Foxconn, Apple's main assembly partner, has faced challenges in dispatching its Chinese staff to India and receiving additional specialised machinery from China. However, there has been no immediate impact on production, according to sources, reported the news publication.
In a statement, China's Ministry of Foreign Affairs has said that the nation "treats all countries equally and is open to enterprises" globally.
"China has never done anything that harms interests of other nations to the detriment of itself."
India is a crucial part of Apple's strategy to reduce its reliance on China amid ongoing trade tensions between US and China.
https://www.reuters.com/business/aerospace-defense/markets/commodities/trump-says-india-agreed-purchase-more-us-oil-gas-2025-02-13/
US will raise military sales to India by many billions of dollars, Trump says
Trump’s F-35 offer still a proposal, Indian official says
F-35 purchase process has not started yet, Indian official says
Any talks on F-35 sales will be between governments, Lockheed Martin says
WASHINGTON, Feb 13 (Reuters) - The United States will increase military sales to India starting in 2025 and will eventually provide F-35 fighter jets, U.S. President Donald Trump said on Thursday.
"We'll be increasing military sales to India by many billions of dollars. We're also paving the way to ultimately provide India with the F-35 stealth fighters," Trump told reporters.
Trump did not provide a timeline, but foreign military sales, especially for cutting-edge technology like the stealthy F-35 jet, typically take years to work through.
In January 2025, China's Experimental Advanced Superconducting Tokamak (EAST) nuclear fusion reactor set a new world record by sustaining a plasma at 100 million° C for 1,066 seconds. This breakthrough is a major step toward the goal of generating clean, limitless energy.
Explanation What is EAST.
EAST is also known as China's "artificial sun". It's a tokamak reactor that uses extreme heat and pressure to create nuclear fusion. What's the significance of this breakthrough.
This breakthrough is a major step toward practical fusion reactors that could power cities without pollution or radioactive waste.
What are the benefits of fusion energy?
Fusion energy has a zero emission footprint and produces far greater amounts of energy than any other source.
Other details
China invests nearly double the U.S. in fusion research.
China has ambitious plans for fusion power plants to provide clean, limitless energy.
The Comprehensive Research Facility for Fusion Technology (CRAFT) is under construction in Hefei, China and is scheduled for completion in 2025.
You can also watch this video to learn more about China's nuclear fusion breakthrough:
https://youtu.be/g-PLekYOv20
DeepSeek trained AI model using distillation, now a disruptive force
https://www.cnbc.com/2025/02/21/deepseek-trained-ai-model-using-distillation-now-a-disruptive-force.html
Distillation is a process of extracting knowledge from a larger AI model to create a smaller one. It can allow a small team with virtually no resources to make an advanced model.
A leading tech company invests years and millions of dollars developing a top-tier model from scratch. Then a smaller team such as DeepSeek swoops in and trains its own, more specialized model by asking the larger “teacher” model questions. The process creates a new model that’s nearly as capable as the big company’s model but trains more quickly and efficiently.
“This distillation technique is just so extremely powerful and so extremely cheap, and it’s just available to anyone,” said Databricks CEO Ali Ghodsi, adding that he expects to see innovation when it comes to how large language models, or LLMs, are built. “We’re going to see so much competition for LLMs. That’s what’s going to happen in this new era we’re entering.”
Distillation is now enabling less-capitalized startups and research labs to compete at the cutting edge faster than ever before.
Using this technique, researchers at Berkeley said, they recreated OpenAI’s reasoning model for $450 in 19 hours last month. Soon after, researchers at Stanford and the University of Washington created their own reasoning model in just 26 minutes, using less than $50 in compute credits, they said. The startup Hugging Face recreated OpenAI’s newest and flashiest feature, Deep Research, as a 24-hour coding challenge.
DeepSeek didn’t invent distillation, but it woke up the AI world to its disruptive potential. It also ushered in the rise of a new open-source order — a belief that transparency and accessibility drive innovation faster than closed-door research.
https://www.nature.com/nature-index/research-leaders/2024/institution/academic/all/global
https://www.tiktok.com/t/ZT2PGR4Yo/
Position Institution Share 2022 Share 2023 Count 2023 Change in Adjusted Share* 2022–2023
1 Harvard University, United States of America (USA) 1169.58 1143.43 3763 -3.1%
2 University of Chinese Academy of Sciences (UCAS), China 598.85 635.81 3227 5.2%
3 University of Science and Technology of China (USTC), China 586.57 631.20 1858 6.7%
4 Peking University (PKU), China 562.29 617.17 2349 8.8%
5 Nanjing University (NJU), China 565.57 609.45 1448 6.8%
6 Zhejiang University (ZJU), China 479.35 595.37 1540 23.1%
7 Tsinghua University, China 529.08 593.45 1946 11.2%
8 Sun Yat-sen University (SYSU), China 437.07 492.47 1266 11.7%
9 Shanghai Jiao Tong University (SJTU), China 439.68 488.94 1449 10.2%
10 Massachusetts Institute of Technology (MIT), United States of America (USA) 479.97 484.86 2032 0.1%
11 Stanford University, United States of America (USA) 592.62 474.13 1929 -20.7%
12 Fudan University, China 429.84 461.26 1333 6.4%
13 Sichuan University (SCU), China 327.62 413.63 768 25.2%
14 The University of Tokyo (UTokyo), Japan 398.90 389.36 1239 -3.2%
15 University of Oxford, United Kingdom (UK) 423.50 388.35 1625 -9.1%
16 University of Michigan (U-M), United States of America (USA) 372.55 380.50 1415 1.2%
17 University of Cambridge, United Kingdom (UK) 428.91 368.12 1436 -14.9%
18 Swiss Federal Institute of Technology Zurich (ETH Zurich), Switzerland 353.52 346.26 1042 -2.9%
19 Yale University, United States of America (USA) 380.20 342.16 1402 -10.8%
20 Nankai University (NKU), China 316.97 337.67 797 5.6%
21 Wuhan University (WHU), China 276.30 334.72 768 20.1%
22 University of Toronto (U of T), Canada 366.05 334.38 1192 -9.4%
23 University of California, San Diego (UC San Diego), United States of America (USA) 335.22 331.74 1062 -1.9%
24 Columbia University in the City of New York (CU), United States of America (USA) 342.91 330.31 1232 -4.5%
25 University of Pennsylvania (Penn), United States of America (USA) 374.65 323.65 1118 -14.4%
26 Cornell University, United States of America (USA) 302.76 320.47 1070 4.9%
27 University of California, Los Angeles (UCLA), United States of America (USA) 341.25 313.78 1214 -8.8%
28 Huazhong University of Science and Technology (HUST), China 248.29 312.65 725 24.8%
29 University of California, Berkeley (UC Berkeley), United States of America (USA) 350.71 312.19 1358 -11.8%
30 Shandong University (SDU), China 263.71 309.49 951 16.3%
31 University of Washington (UW), United States of America (USA) 305.50 295.93 1452 -4.0%
32 Northwestern University (NU), United States of America (USA) 313.12 295.49 904 -6.4%
33 Johns Hopkins University (JHU), United States of America (USA) 351.93 294.65 1358 -17.0%
34 Xiamen University (XMU), China 262.58 289.61 569 9.3%
35 Soochow University, China 246.72 289.15 614 16.2%
36 Southern University of Science and Technology (SUSTech), China 295.49 279.54 770 -6.2%
37 Jilin University (JLU), China 247.88 268.37 601 7.3%
38 National University of Singapore (NUS), Singapore 250.00 253.94 904 0.7%
39 Central South University (CSU), China 189.42 252.70 582 32.2%
40 Washington University in St. Louis (WUSTL), United States of America (USA) 261.57 251.01 804 -4.9%
https://www.scmp.com/news/china/science/article/3297914/china-leads-world-physics-research-us-distant-rival-nature-index-shows
The world’s top organisations for high-quality physics research are all in China or Europe, with the US’ highest-ranked institution coming in at No 13, according to the latest Nature Index.A China-based observer said this could have far-reaching implications for the United States’ technological edge, with physics underpinning much of modern technology and innovation. But another highlighted America’s overall lead in original research breakthroughs.The Nature Index – maintained by the prestigious journal of the same name – ranks research institutions based on their contributions to articles published in the world’s most influential scientific journals.The latest list was based on research output in the 12 months from November 2023.
China dominated the top 10, with the Chinese Academy of Sciences (CAS), the University of Science and Technology of China (USTC) and Tsinghua University taking the top three spots.Only two non-Chinese institutions broke into the top 10 – the Max Planck Society in Germany at No 4 and the French National Centre for Scientific Research at No 10.
The top three US institutions – the Massachusetts Institute of Technology, Harvard University and Stanford University – were ranked at 13th, 16th and 20th, respectively.
In the 2021 index, by contrast, the three American institutions were in the top 10, while only two from China – CAS and USTC – made it into the group.
The Nature Index measures two different scores – count and share. An institution receives one point for each article with at least one affiliated author, and these points add up to the institution’s count score.For the share score, each article is assigned a value of one, which is divided equally among all authors. For example, for an article with 10 authors, each receives a share value of 0.1. Each institution’s share score is calculated by adding the share values of the authors from that institution.
In the latest list, CAS leads by a wide margin in both metrics, while the Max Planck Society is the only non-Chinese research organisation in the top five.However, a CAS physicist warned that while the index was an informative measure of the strength of research institutions, it did have its limitations and the rankings should be taken with a pinch of salt.“The increasing number of high-quality scientific papers published by Chinese researchers in top journals in recent years reflects to some extent that China’s overall strength in physics research is on the rise,” said Chen Xiaolong, a physicist at the Institute of Physics under CAS.
But this metric was not everything, he said. “China still lags behind the United States overall in terms of original and truly groundbreaking research breakthroughs, except for a few branches,” Chen said.Analytical instrumentation engineer Jin Xianchi, who holds a doctorate in physics from the University of the Chinese Academy of Sciences (UCAS), was of the same view. He said many factors were behind the rankings.“Chinese and European labs tend to have larger teams, so it is easier to produce more influential papers,” Jin said. On the other hand, some notably pioneering research may not be accepted by academic journals at first, which could put an innovative organisation at a disadvantage.
https://www.scmp.com/news/china/science/article/3297914/china-leads-world-physics-research-us-distant-rival-nature-index-shows
“The increasing number of high-quality scientific papers published by Chinese researchers in top journals in recent years reflects to some extent that China’s overall strength in physics research is on the rise,” said Chen Xiaolong, a physicist at the Institute of Physics under CAS.
But this metric was not everything, he said. “China still lags behind the United States overall in terms of original and truly groundbreaking research breakthroughs, except for a few branches,” Chen said.Analytical instrumentation engineer Jin Xianchi, who holds a doctorate in physics from the University of the Chinese Academy of Sciences (UCAS), was of the same view. He said many factors were behind the rankings.“Chinese and European labs tend to have larger teams, so it is easier to produce more influential papers,” Jin said. On the other hand, some notably pioneering research may not be accepted by academic journals at first, which could put an innovative organisation at a disadvantage.
However, according to Jin, there has actually been a decline in America’s overall leadership in scientific research since World War II – including in physics. He said this had happened for a variety of reasons, including a brain drain.Physics is the foundation on which most modern technological developments and cutting-edge research are based. And throughout the 20th century and into the 21st, the US has been the leading destination for physics research, accounting for nearly 100 Nobel Prizes in the field.“A decline in the level of research over a period of time is bound to have implications for the future,” Jin said.
In 2018, an article in Forbes magazine warned that the US was at risk of “no longer attracting the best and brightest minds in physics”.An American Physical Society survey cited in the article showed that the country had seen an unprecedented drop in the number of international applications to physics PhD programmes over the previous year.Meanwhile, a number of Chinese physicists have chosen to leave the US in recent years.
In January last year, for example, award-winning Chinese-American physicist Gao Huajian joined Tsinghua University as a full-time professor. More recently, computational physicist Chen Hudong – an elected member of the US National Academy of Engineering – left America after more than four decades to take up a full-time position in China.An annual Chinese report assessing scientific activities and trends in major countries, released last November, showed that China had regained the world’s top spot in physics research in 2024, after losing it to the United States in 2023.
@NuryVittachi
There is no race between China and the US to be top player in AI development, says a new report from Nature, the world's top science journal.
The two countries are on different tracks with different endpoints. The US has been leading the pack in making big, headline-grabbing projects like ChatGPT, while China's focus has been, and remains, the production of practical AI programs that help industrialists, farmers, business people and factories.
AI programs in China are being used "to make trains run on time, monitor fish stocks and provide automated telehealth services," says the report by science writer Jacob Dreyer in the latest edition of the scientific publication.
Ultimately, China will do its usual thing, passing the systems that work to other nations, "especially to lower-income countries," the report says.
.
DIFFERENT SOURCES OF SUPPORT
In the west, companies need to have innovative ideas that make an impact in the media to attract venture capitalists to invest. In contrast, China has more of an engineering school approach—projects must have practical benefit and be shown to be working to get government investment.
"The divergence in priorities reflects the forces driving innovation in each economy: venture capital in the United States and large-scale manufacturing enterprises and organs of the state in China," Dreyer says.
That doesn't mean that Chinese AI projects are always smaller. One of them is to integrate AI into a system to control the national grid so that the best use can be made of energy.
China is keen to keep its reputation as a leader in making clean energy options affordable around the world. "Its emerging AI playbook mirrors its approach to other technologies, such as electric vehicles and clean energy: not the first to innovate, but the first to make them affordable for widespread use," Dreyer writes.
The huge attention paid to DeepSeek, a relatively cheap but high performance AI chatbot, has hidden the difference in strategies, inspiring journalists to conjure up the colorful but ultimately inaccurate scenario of a US-China AI "arms race".
.
STRATEGIC PLAN - FOR FARMERS
The appearance of the Nature report happily coincided with the February 20 publication of a Chinese government "strategic plan" for farmers to use scientific developments, including AI and genetic modification, to boost crop production and help the country move towards food self-sufficiency. You can't get much more practical than that.
Other recent data also bolsters the theory that Chinese AI projects may be lower profile, but have already made more inroads in business and industry.
A recent international study of usage of AI revealed that more respondents from China – 83 per cent – said their companies were using AI, than respondents from the United States – at only 65 per cent. In fact, the data in the SAS Generative AI Global Research Report shows that China was ahead of all other nations in the study in this regard (see pic).
[from fridayeveryday. com]
https://x.com/NuryVittachi/status/1894277527151284232
@wmhuo168
China claims to have built a jet fuel-powered engine for Mach 16 flight—faster than any known aircraft. If true, this rewrites the rules of war, rendering missile defenses obsolete. Game-changer or hype? Let’s break it down.
https://x.com/wmhuo168/status/1895319837569917216
---------------------
China’s Mach 16 Engine SpaceRacer6 on TikTok
https://www.tiktok.com/t/ZT2H5ckPw/
China's semiconductor foundry company SMIC achieved record quarterly performance in the third quarter of this year (July to September).
Despite challenges in producing advanced semiconductors due to U.S. sanctions, the demand for legacy chips used in local home appliances has increased, leading to higher sales.
SMIC announced on Nov. 8 that it recorded $2.17119 billion in sales in the third quarter, a 34% increase compared to the same period last year ($1.62058 billion).
This marks the highest quarterly sales ever, and it is the first time SMIC's sales have surpassed $2 billion. Operating profit was $169.89 million, and net profit was $148.80 million, representing increases of 94.4% and 58.3%, respectively, compared to the same period last year.
SMIC's performance continues to improve, supported by the Chinese government's backing for semiconductor self-sufficiency despite the ongoing U.S.-China conflict.
In the third quarter, domestic sales in China amounted to $1.875 billion, accounting for 86.4% of total sales, up from 84% ($1.361 billion) last year. Conversely, sales in the U.S., Asia, and Europe declined.
SMIC is also facing difficulties in acquiring advanced semiconductor equipment. Its capital expenditure in the third quarter was $1.179 billion, a 44.8% decrease from $2.135 billion last year.
However, production based on legacy processes continues to increase. SMIC's factory utilization rate in the third quarter was 90.4%, recovering from 77.1% in the same quarter last year.
SMIC expects its sales in the fourth quarter (October to December) to be similar to or slightly higher than the record high of the previous quarter.
Hua Hong Semiconductor, another leading foundry in China alongside SMIC, also showed significant improvement compared to last year.
The company's third-quarter sales were $526.30 million, a 7.4% decrease compared to the same period last year, but net profit turned positive at $22.91 million, compared to a loss of $25.90 million last year.
The strong performance of Chinese semiconductor companies could provoke further tightening of U.S. semiconductor regulations.
The South China Morning Post reported, "It is a common view that Trump will strengthen control over China's technology sector."
Arnaud Bertrand
@RnaudBertrand
This is actually an extraordinary admission to make for a US Vice President https://x.com/OopsGuess/status/1902396228404674853/video/1
Vance explains that "the idea of globalization was that rich countries would move further up the value chain while the poor countries made the simpler things."
But he laments that it didn't quite work out this way: as he explains it turns out that poor countries (mostly China) didn't want to just remain cheap labor forever and started moving up the value chain themselves. Which is why, according to him, globalization was a failure.
Meaning that the objective of globalization wasn't to reduce global inequalities but very much to maintain them, to institute a system of permanent economic hierarchy where rich countries would maintain their hold over the most profitable sectors while relegating poor countries to perpetual subordination in lower-value production.
This is basically all you need to know to explain 90% of U.S. foreign policy these past few years: colonial thinking is alive and well, and America's shift of strategy in recent years - away from the previous "Washington Consensus" of "free" markets towards a much more overt attempt to contain and restrict China's development - stems precisely from this mindset.
From semiconductor export controls to investment restrictions, these policies aren't about 'national security' in any genuine sense - they're about trying to preserve a global economic order where, simply put, poorer nations know their assigned place and stay there. At the very core, that's the "China threat": a China that stepped out of the economic lane assigned to it by the West.
It's deeply ironic when you think of it: a global game allegedly designed to "spread market principles" worldwide is being abandoned precisely because it worked too well. When China succeeded better than expected, the response wasn't to celebrate the validation of the game's effectiveness but to change its rules. Precisely because the real unspoken game - but now clearly stated by the U.S. Vice President - was to maintain global inequality, not eliminate it.
All in all, in case they hadn't yet gotten the memo, this sends a very clear message to the developing world: economic development will require challenging a U.S.-dominated economic order that views their advancement as a threat rather than a success. Which incidentally is why Vance's words might actually help accelerate the very redistribution of global economic power he laments, pushing more nations to recognize that genuine development requires strategic independence from a system intended to keep them in their place.
https://x.com/RnaudBertrand/status/1902528428643168632
https://propakistani.pk/2025/03/22/pakistan-launches-its-first-homegrown-ai-chatbot-tailored-for-local-needs/
Pakistan has achieved a notable milestone in the tech sector with the beta launch of Zahanat AI, the nation’s first locally developed artificial intelligence chatbot. Spearheaded by entrepreneur Mehwish Salman Ali, co-founder and CEO of Data Vault and Zahanat AI, the platform promises to address Pakistan’s unique challenges with culturally sensitive and locally relevant solutions.
Zahanat AI is the culmination of years of development, operating from a dedicated data center in Karachi since 2022. This means that the AI model’s data stays in Pakistan for processing and doesn’t go anywhere else. This data center is connected with high-speed internet and has robust DDoS protection.
The system utilizes a mixed GPU architecture, leveraging both Nvidia GPUs and chips, initially incorporating used gaming GPUs to build its computational power. It was a relatively low-cost development project. The owners said that DeepSeek cost $5 million to make, but Zahanat AI cost less, without specifying exactly how much.
Mehwish Salman Ali said:
Our goal was to create an AI that understands and responds to the specific needs of Pakistan. Zahanat AI is trained on a massive dataset of 2 billion parameters, all processed and stored within Pakistan, ensuring cultural awareness and relevance.
A key distinguishing feature of Zahanat AI is its focus on ethical and responsible AI development. The platform is specifically trained to censor sensitive topics and avoid discussions about particular individuals.
The beta launch of Zahanat AI is currently on an invitational basis, requiring interested users to submit their email and personal details, including their profession, work email, and social media accounts. This selective access aims to gather valuable feedback and refine the platform before a wider public release.
ALSO READ
Baidu’s New AI Model “Rivals Deepseek At Half the Price”
Ali further explained:
By training Zahanat AI on Pakistan’s data, we are building a tool that can provide Pakistan-focused solutions to the diverse problems faced by our communities. We believe this technology has the potential to transform various sectors, from education and healthcare to business and governance.
For now, the beta launch has introduced Zahanat AI’s initial Z1 model, with plans to improve and expand further in the future with Z2 and so on. The Z2 model will introduce multilingual support for all languages in Pakistan and will get voice input as well.
https://www.arabnews.com/node/2594353/pakistan
Zahanat AI is a text-based generative AI model that enables users to engage in human-like conversations, answer queries, and assist in various domains
Its key differentiator is its hosting and local training on Pakistani culture and localized issues, which makes it equipped to address regional challenges
---------------
Meet the woman who made Pakistan's first AI chatbot
https://theprint.in/go-to-pakistan/woman-pakistans-first-ai-chatbot-mehwish-salman-ali/2567308/
The platform is designed to empower Pakistani citizens, particularly in sectors like education and healthcare. One of Zahanat’s most anticipated developments is the upcoming Z2 model, which will support Urdu and multiple regional languages. This is a game-changer for more than half of Pakistan’s population, a large part of which struggleswith English or even Urdu.
Ali imagines the platform being used by a rural student to access world-class education in their native Sindhi or Pashto. She dreams of Zahanat helping an elderly womanreceive her healthcare diagnosis in Balochi.
“We’re not just enabling access to AI, we’re redefining who gets to be part of the ecosystem. We’re moving from a digital divide to digital empowerment. This isn’t just tech progress. It’s social progress,” Ali said.
Zahanat is a personal mission for Ali to break the gender barriers that persist in tech. She has faced bias in the male-dominated industry, both spoken and unspoken.
“When I lead a project like Zahanat, it’s not just innovation, it’s disruption. It’s proof that women can lead tech.”
------------
Kineto: K-Electric Becomes Pakistan’s First Power Utility to launch Generative AI Chatbot To Enhance Customer Experience
https://propakistani.pk/2025/03/26/kineto-k-electric-becomes-pakistans-first-power-utility-to-launch-generative-ai-chatbot-to-enhance-customer-experience/
“Users of the KE Live App have grown by 21% annually over the last 5 years and now stand at 1.3 million digitally connected customers. This is over one-third of KE’s total customer base and conveys our digital-savvy population. We then heralded another innovation when we launched the WhatsApp platform back in 2021, and now this platform caters to over 2.0 million people. Additionally, nearly another half a million subscribe to our e-billing feature, a step that helps save Pakistan paper and reduce its import bill.”
“Now, leading the way with digital transformation in customer engagement, Kineto was just the next logical step forward reflecting our investment in future-ready digital platforms, further transforming the way Karachi’s customers interact with its power utility.”
The chatbot has been developed in collaboration with Convex Interactive, KE’s technology partner for this initiative.
“This partnership with K-Electric aligns with our mission to revolutionize customer engagement through AI,” said Aamir Irfan Siddiqui, CEO & Founder of Convex Interactive. “By leveraging generative AI, we’re making customer interactions faster, smarter, and more intuitive.”