China's DeepSeek that 'shocked' America and American technology companies has an update that it says .... to OpenAI's o3 and Google's Gemini 2.5 Pro
Chinese AI startup DeepSeek has announced a significant upgrade to its R1 model, named DeepSeek-R1-0528, which claims to boast enhanced capabilities in mathematics, programming, and general logic while reducing AI "hallucinations." The announcement, made via a post on the AI model platform Hugging Face, positions the new model as a formidable competitor to industry leaders like OpenAI's o3 and Google's Gemini 2.5 Pro.
When Deepseek stunned America's technology industry
DeepSeek stunned the global tech community in January with the original R1, a model that rivaled the performance of systems developed by much larger U.S. companies, reportedly at a fraction of the cost. The debut disrupted the AI industry, raising questions about the massive global spending on computational resources and sparking a wave of AI releases from Chinese tech giants like Alibaba Group Holding Ltd. and Zhipu AI.
Built for under $6 million using less-advanced Nvidia H800 chips, R1's efficiency challenged assumptions about the necessity of massive computational investments, triggering a $1 trillion stock market drop, including Nvidia's $589 billion loss. DeepSeek's open-source model, praised for its cost-effectiveness, raised concerns about U.S. AI dominance, prompting calls for stricter export controls and urgent policy reassessments
American lawmakers, citing national security, swiftly reacted. Representatives Gottheimer and LaHood introduced the No DeepSeek on Government Devices Act, alleging Chinese Communist Party data theft risks. NASA, the Navy, and Congress banned DeepSeek on government devices, fearing espionage due to its China-based servers and lax data policies. Companies like Oracle blocked employee use, wary of intellectual property leaks.
New DeepSeek model works on same foundation
The upgraded DeepSeek-R1-0528 builds on this foundation, offering what the company claims is "greater depth of reasoning" and performance approaching that of the world's top AI models. 'Its overall performance is now approaching that of leading models,' DeepSeek stated, signaling its ambition to compete at the forefront of the global AI race.
The rise of DeepSeek has also catapulted its founder,
Liang Wenfeng
, into the spotlight. Once a relatively unknown figure, Liang became a symbol of China's growing prowess in AI innovation following R1's debut. In February, he was invited by President Xi Jinping to a high-profile meeting alongside prominent entrepreneurs like Alibaba co-founder Jack Ma and Tencent's Pony Ma, underscoring DeepSeek's significance in China's tech landscape.
What makes timing of DeepSeek announcement important
The timing of the DeepSeek-R1-0528 announcement was notable, coming just hours before Nvidia Corp., the Santa Clara-based leader in AI chips, released its latest financial report. Nvidia, whose shares took a hit following R1's initial launch due to concerns about reduced AI infrastructure costs, has since recovered as global investment in AI data centers remains robust. Nvidia's latest forecast reaffirmed strong demand for its chips, suggesting that DeepSeek's cost-efficient approach has not slowed the broader AI investment boom.
DeepSeek's rapid progress highlights China's accelerating role in the global AI arms race. Industry analysts note that the startup's ability to deliver high-performing models at lower costs could reshape the economics of AI development, challenging the dominance of U.S. tech giants.
AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hindustan Times
32 minutes ago
- Hindustan Times
A new framework to govern Ladakh
The regulations regarding Ladakh notified by the Centre on Tuesday are significant for multiple reasons. One, it suggests a direction for addressing governance concerns in the Union territory (UT), a first since 2019, when Article 370 was abrogated and Jammu and Kashmir was bifurcated into two UTs. Two, it provides partial redress to concerns that local communities raised, and highlighted in their 2023 protests, regarding safeguards for their unique regional identity, culture, and material resources. Three, a stable Ladakh with a robust governance structure and infrastructure is essential since the UT shares a restive border with China: Indian and Chinese troops clashed in the Galwan Valley in 2020 and the stand-off continued till October last year. The new rules address matters related to domicile status and job quotas (85% reservation for Ladakhis excluding economically weaker section quota), which, hopefully, will assuage the fears about large-scale settlement of 'outsiders' in this ecologically sensitive terrain. The rules concerning domicile status are similar to those in Jammu and Kashmir, though Ladakh has no elected legislature. This means regulations governing Ladakh are executive decisions made by the Centre — in this case in consultation with local representatives — and not the outcome of deliberations in an elected House. It would also mean that these regulations can be changed, withdrawn or amended through executive orders and without the approval of local stakeholders. Though such fears seem unwarranted, the Centre must have a conversation with the broader segment of Ladakhi voices if necessary, and ensure that the regulations, say for the protection of local languages, become enabling provisions. After all, the call in Ladakh has been for statehood and inclusion under the Sixth Schedule of the Constitution, which provides legislative, executive, judicial and financial powers, and the current initiative falls short of that demand.
Economic Times
41 minutes ago
- Economic Times
Chinese 'kill switches' found in equipment at US solar firms trigger national security fears. What are they?
Engineers have discovered 'kill switches' embedded in Chinese-manufactured parts on American solar farms, raising fears Beijing could manipulate supplies or 'physically destroy' grids across the US, UK and Europe. Rogue communication devices found in Chinese solar inverters have triggered fears of national security concerns. But what exactly are kill switches? Tired of too many ads? Remove Ads Tired of too many ads? Remove Ads What are kill switches? Tired of too many ads? Remove Ads Chinese "kill switches" that could allow Beijing to cripple power grids and trigger blackouts across the West were found in equipment at US solar farms earlier this month, raising fears that China could manipulate supplies or 'physically destroy' grids across the US, UK and Europe. US energy officials re-assessed the risk posed by small communication devices in power inverters - an integral component of renewable energy systems that connects them to the power to Times, the rogue devices, including cellular radios, were discovered in Chinese-made power inverters that are used to connect solar panels and wind turbines to electricity grids across the world, including the UK. The hidden communications equipment could be deployed remotely to switch off inverters with potentially catastrophic US experts found rogue communication devices in some solar power inverters not listed in product documents, Reuters reported citing these devices to skirt firewalls and switch off inverters remotely, or change their settings, could destabilise power grids, damage energy infrastructure and trigger widespread blackouts. The discovery has sparked concerns that Beijing could potentially disrupt power grids in Western nations, given the heavy reliance of renewable energy systems on Chinese-manufactured solar panels incorporate parts from various countries, including China. However, it remains unclear whether any power converters installed at UK wind or solar farms contain the suspected Chinese "killswitches." kill switch is a mechanism used to shut down or disable a device or program, according to Tech Target. A kill switch is typically used to prevent the theft of equipment or data, or to shut down machinery in emergency situations. The extent to which it restricts, modifies, or halts an action or process depends on the specific system, operation, or software it is designed to switch is a broad term when it comes to the types of technology, software and tools used to create and facilitate it. In manufacturing, for example, a factory might use a kill switch -- also called a big red button -- to shut down machinery if a worker is in danger. Kill switch software, on the other hand, can sometimes include software-encoded kill switches such as anti-piracy radios, acting as kill switches, were discovered embedded in power inverters produced by Chinese manufacturers and supplied to US power inclusion of these remotely controllable devices introduces potential cybersecurity vulnerabilities within the nation's power infrastructure. Power inverters are essential components in solar and wind farms, converting renewable energy into a form compatible with the national grid. This makes them a critical point of vulnerability. If compromised, these devices could be manipulated—potentially by China—to cause widespread blackouts across the West or even disrupt the entire power grid."We know that China believes there is value in placing at least some elements of our core infrastructure at risk of destruction or disruption," said Mike Rogers, a former director of the US National Security Agency."I think that the Chinese are, in part, hoping that the widespread use of inverters limits the options that the West has to deal with the security issue."Over the last nine months, US experts discovered suspicious devices hidden inside inverters and batteries from several Chinese manufacturers. These specialists carefully dismantle the equipment to identify potential security threats before the products are integrated into power typically have remote access capabilities through cellular networks, and utility companies usually install firewalls to block unauthorized access and prevent communication with China. However, these unauthorized components were not disclosed in the product documentation when the equipment was shipped to the United States.(With inputs from Reuters)
Time of India
an hour ago
- Time of India
Freedom of speech does not include freedom to make defamatory statements against Army: HC to Rahul Gandhi
Live Events (You can now subscribe to our (You can now subscribe to our Economic Times WhatsApp channel Freedom of speech is subject to reasonable restrictions and it does not include the freedom to make defamatory statements against any person or the Indian Army, the Allahabad High Court has said, rejecting Congress leader Rahul Gandhi's plea to quash summons against him for alleged derogatory remarks during his Bharat Jodo Yatra in Subhash Vidyathi of the Lucknow bench, in his May 29 verdict, observed that there was a prima facie case against Gandhi for making defamatory statements against the Indian Army and upheld the lower court order issued in February summoning the Leader of Opposition in Lok Sabha to face trial."No doubt, Article 19(1)(a) of the Constitution of India guarantees freedom of speech and expression, this freedom is subject to reasonable restrictions and it does not include the freedom to make statements which are defamatory to any person or defamatory to the Indian Army," the judge said in the detailed order released case stems from a complaint filed by Udai Shanker Srivastava, a retired Director from the Border Roads Organisation (a position equivalent to a Colonel in the Indian Army).Srivastava alleged that on December 16, 2022, during his 'Bharat Jodo Yatra' in Lucknow, Gandhi made disparaging comments about a face-off between the Indian and Chinese armies in Arunachal Pradesh on December 9, contended that Gandhi's statement was "false and baseless" and made with "evil intention of demoralising the Indian Army and to damage the faith of the Indian population in the Indian Army."He highlighted that the official statement from the Indian Army on December 12, 2022, confirmed that "PLA troops contacted the LAC in Tawang Sector which was contested by our troops in a firm and resolute manner. This faceoff led to minor injuries to a few personnel from both sides."The complainant said that Gandhi's "baseless and derogatory statement" deeply hurt him and other Additional Chief Judicial Magistrate, Court No. 27, Lucknow, had on February 11, 2025, summoned Gandhi to face trial for the offence under Section 500 of the Penal Code (defamation).The lower court observed that prima facie, Gandhi's statement appeared to demoralise the Indian Army and its personnel, and was not made in the performance of his official the High Court, Gandhi's counsel Pranshu Agarwal argued that the complaint was politically motivated and lacked contended that the complainant was not an "aggrieved person" under Section 199 of the Criminal Procedure Code, as the alleged defamation was against the Indian Army as an institution, not Srivastava High Court, however, clarified that the complainant as an "aggrieved person" is capable of filing the complaint under Section 199 Vidyarthi also noted that the trial court's decision to summon Gandhi was based on a "judicious application of mind" after considering the complaint and witness statements, and was not mechanical, while dismissing Gandhi's plea challenging the summons. PTI
