News/Deepseek
DeepSeek cuts AI processing costs 50% with new sparse attention tech
Chinese AI startup DeepSeek has launched DeepSeek-V3.2-Exp, an experimental model that introduces "sparse attention" technology to cut AI processing costs in half while maintaining performance levels. The release builds on DeepSeek's reputation for creating efficient AI systems using fewer resources than traditional approaches, though experts question whether the cost-cutting architecture compromises model reliability and safety. What you should know: DeepSeek's new experimental model represents a significant shift in AI architecture design, focusing on efficiency over raw computational power. The V3.2-Exp model introduces DeepSeek Sparse Attention (DSA), which selectively processes only the most relevant information rather than analyzing all available data....
read Sep 19, 2025Huawei builds AI model that’s “nearly 100%” effective at censoring sensitive content
Huawei has co-developed a safety-focused version of DeepSeek's AI model that it claims is "nearly 100% successful" at preventing discussion of politically sensitive topics. The collaboration with Zhejiang University demonstrates how Chinese companies are adapting open-source AI models to comply with domestic regulations requiring AI systems to reflect "socialist values" and avoid sensitive political discussions. What you should know: Huawei used 1,000 of its own Ascend AI chips to train the modified model, called DeepSeek-R1-Safe, which was built from DeepSeek's open-source R1 model.• The model achieved "nearly 100% successful" defense against "common harmful issues ... including toxic and harmful speech,...
read Sep 17, 2025DeepSeek’s $294K AI model becomes first to pass peer review
DeepSeek's AI model R1 has become the first major large language model to undergo peer review, with researchers publishing details in Nature revealing the reasoning-focused system cost just $294,000 to train. The landmark study provides unprecedented transparency into how the Chinese startup created a model that rivals OpenAI's offerings at a fraction of the cost, potentially reshaping expectations around AI development expenses and accessibility. What you should know: The peer-reviewed paper confirms DeepSeek's innovative approach to creating powerful AI without relying on competitor outputs. R1 excels at reasoning tasks like mathematics and coding, competing directly with US-developed models while costing...
readGet SIGNAL/NOISE in your inbox daily
All Signal, No Noise
One concise email to make you smarter on AI daily.
More pillager than villager? Chinese hacking tool “Villager” downloaded 10K times in 2 months
A mysterious Chinese AI penetration testing tool called Villager has been downloaded nearly 10,000 times since its July release, raising serious concerns about its potential misuse by cybercriminals. The tool, which combines Kali Linux with DeepSeek AI to automate offensive security operations, is being compared to Cobalt Strike's trajectory from legitimate red-team software to widely adopted malware infrastructure. What you should know: Villager represents a new category of AI-native penetration testing tools that could democratize advanced cyberattacks. The tool integrates Kali Linux toolsets with DeepSeek AI models to fully automate testing workflows, positioning itself as an AI-powered successor to Cobalt...
read Sep 1, 2025Chinese AI startups match US giants with fewer resources
Chinese AI companies have rapidly closed the gap with U.S. rivals, developing cost-efficient models that perform comparably to leading Western systems like OpenAI's GPT-4o and Anthropic's Claude 3.5 Sonnet. This strategic pivot represents a distinct approach to generative AI, emphasizing resource optimization and targeted applications over the compute-intensive methods favored by American firms. What you should know: DeepSeek, a Chinese AI startup founded in 2023, has emerged as the most notable competitor, achieving competitive performance with significantly fewer computing and data resources than U.S. models. The company's DeepSeek-R1 model now performs comparably to established Western systems despite operating under resource...
read Aug 20, 2025AI avatars outperform human salespeople in China’s $7.7M livestream market
Chinese AI-powered "virtual human" salespeople are outperforming their human counterparts on major Chinese ecommerce platforms, working 24/7 to sell everything from printers to wet wipes. Built using technology from Baidu, one of China's largest tech companies, and DeepSeek, these AI avatars are demonstrating the potential for artificial intelligence to fundamentally reshape digital commerce and potentially displace human influencers and salespeople. What you should know: PLTFRM, a Shanghai-based marketing company, has deployed around 30 AI avatar salespeople across Chinese ecommerce sites like Taobao and Pinduoduo, with measurable success rates. Brother's AI avatar sold $2,500 worth of printers in its first two...
read Aug 19, 2025DeepSeek’s open-source AI model matches GPT-4 at 68x lower cost
DeepSeek, a Chinese AI startup backed by High-Flyer Capital Management, has released V3.1, a 685-billion parameter AI model that matches the performance of leading proprietary systems from OpenAI and Anthropic while remaining completely open source. The model scored 71.6% on the prestigious Aider coding benchmark—rivaling Claude Opus 4's performance at a fraction of the cost—potentially disrupting the traditional AI business model that relies on expensive API access and usage restrictions. What you should know: DeepSeek V3.1 delivers frontier-level AI capabilities through an innovative hybrid architecture that seamlessly combines chat, reasoning, and coding functions. The model processes up to 128,000 tokens...
read Aug 14, 2025DeepSeek delays AI model launch after Huawei chip training fails
DeepSeek, a Chinese artificial intelligence company, has delayed the launch of its new AI model after failing to successfully train it using Huawei's chips, according to a Financial Times report. This setback highlights the ongoing challenges facing China's efforts to reduce dependence on U.S. technology amid ongoing trade restrictions. The big picture: China's push for technological self-sufficiency is encountering significant technical hurdles as domestic chip alternatives struggle to match the performance of banned U.S. semiconductors. Why this matters: The delay underscores the complex reality of replacing advanced U.S. technology infrastructure, particularly in AI development where computational power is critical for...
read Aug 4, 2025Monkey brain, explained: Chinese researchers build AI system with 2B neurons to mimic macaque neurology
Chinese researchers at Zhejiang University have created the Darwin Monkey, an AI system that models a macaque's brain using 960 specialized chips with over 2 billion artificial neurons and 100 billion synapses. The project represents the world's largest neuromorphic computer and marks a significant step toward brain-inspired computing that could revolutionize AI efficiency and capabilities. What you should know: The Darwin Monkey uses China's DeepSeek AI model to perform complex cognitive tasks including logical reasoning, content generation, and mathematical problem solving. The system approaches the neural complexity of an actual macaque brain, making it the first neuromorphic computer based on...
read Jul 7, 2025Huawei denies copying Alibaba’s AI model amid China’s heated competition
Huawei's AI research division has denied allegations that its Pangu Pro large language model copied elements from Alibaba's Qwen AI model, following claims made in a technical paper posted on GitHub. The controversy highlights growing tensions in China's competitive AI landscape, where tech giants are racing to develop cutting-edge models following the success of startup DeepSeek's low-cost R1 model. What happened: An entity called HonestAGI published a paper claiming Huawei's Pangu Pro Moe model showed "extraordinary correlation" with Alibaba's Qwen 2.5 14B model. The paper alleged that Huawei's model was derived through "upcycling" rather than being trained from scratch, suggesting...
read Jul 7, 2025German firm makes DeepSeek AI 200% faster with 90% of original performance
German AI consulting firm TNG Technology Consulting GmbH has released DeepSeek-TNG R1T2 Chimera, a significantly faster variant of DeepSeek's popular open-source reasoning model R1-0528. The new model delivers 90% of the original's intelligence while generating responses with 60% fewer tokens, translating to 200% faster inference and dramatically lower compute costs for enterprises. What you should know: R1T2 represents a breakthrough in AI model efficiency through TNG's Assembly-of-Experts (AoE) methodology, which merges multiple pre-trained models without additional training. The model combines three parent models: DeepSeek-R1-0528, DeepSeek-R1, and DeepSeek-V3-0324, creating what TNG calls a "Tri-Mind" configuration. Unlike traditional training approaches, AoE selectively...
read Jun 30, 2025Germany orders Apple and Google to remove DeepSeek AI app over China data concerns
Germany's top data protection regulator has formally requested Apple and Google remove the DeepSeek AI app from their stores, citing concerns over illegal data transfers to China. This marks the latest escalation in a growing international crackdown on the Chinese AI startup, as Western governments grapple with data sovereignty concerns amid rising AI adoption. What you should know: Germany joins a growing list of countries taking action against DeepSeek over data privacy violations.• Meike Kamp, Berlin's federal commissioner for data protection and freedom of information, said DeepSeek failed to provide sufficient guarantees that user data is protected under EU-equivalent standards.•...
read Jun 27, 2025Global governments restrict DeepSeek AI over China data security fears
Chinese AI startup DeepSeek has triggered a global regulatory backlash, with governments across multiple continents restricting or investigating the company's popular ChatGPT rival over data security and privacy concerns. The widespread scrutiny reflects growing international wariness about Chinese AI systems and their potential access to sensitive user information. DeepSeek gained international attention in January 2025 when it claimed to have developed an AI model capable of matching ChatGPT's performance at significantly lower costs. However, the company's own privacy policy reveals that it stores user data—including chat requests and uploaded files—on servers located in China, raising red flags for government officials...
read Jun 17, 2025Italy targets DeepSeek in 2nd regulatory probe over AI hallucination warnings
Italy's antitrust regulator AGCM has opened an investigation into Chinese AI startup DeepSeek for allegedly failing to adequately warn users about the risk of AI hallucinations in its responses. The probe represents the latest regulatory challenge for DeepSeek in Italy, following a February order from the country's data protection authority to block access to its chatbot over privacy concerns. What you should know: The Italian Competition and Market Authority (AGCM), which oversees both antitrust issues and consumer protection, is examining whether DeepSeek provides sufficient warnings about AI-generated misinformation. The regulator claims DeepSeek did not give users "sufficiently clear, immediate and...
read Jun 5, 2025AI trends reshape tech landscape in Mary Meeker’s latest report
Mary Meeker's latest analysis reveals AI's unprecedented growth and global impact, with ChatGPT reaching 800 million weekly users and Chinese models gaining significant market share. This comprehensive report highlights a technological revolution that's transforming industries at remarkable speed, with implications that extend far beyond Silicon Valley as global competition intensifies and open-source development flourishes. The big picture: Mary Meeker's 340-page "Trends – Artificial Intelligence" report reveals a technology revolution moving at an unprecedented pace and scale, with far-reaching global implications. ChatGPT reached 100 million users in just two months and by April 2025 had 800 million weekly users handling over...
read Jun 2, 2025DeepSeek update challenges OpenAI and Google dominance
DeepSeek is emerging as a formidable challenger in the global AI landscape with its latest release demonstrating significant performance improvements while maintaining an open-source approach. The Chinese startup's new DeepSeek-R1-0528 model showcases remarkable gains in complex reasoning and coding capabilities, areas where even industry leaders struggle. What makes DeepSeek particularly noteworthy is its combination of competitive performance, open licensing, and cost-efficient development—a strategy that could reshape who controls and benefits from advanced AI technology. The big picture: DeepSeek's latest AI model, DeepSeek-R1-0528, is challenging Western AI giants like OpenAI and Google with significant performance improvements in reasoning, coding, and logic....
read May 21, 2025AI-to-AI workflows turn ideas into stunning images
ChatGPT's ability to enhance creative output takes center stage as experienced users find new ways to optimize AI image generation. By leveraging the strengths of multiple AI systems—having one AI craft detailed prompts for another to visualize—users can dramatically improve their results without requiring extensive prompt engineering skills. This collaborative AI approach demonstrates how strategic prompt development can transform vague concepts into richly detailed visualizations across various use cases. 1. The idea factory DeepSeek created this prompt to visualize ChatGPT as a brainstorming partner: "A glowing AI robot (ChatGPT) sits at a cluttered wooden desk, surrounded by floating lightbulb ideas...
read May 10, 2025Hallucination rates soar in new AI models, undermining real-world use
Recent "reasoning upgrades" to AI chatbots have unexpectedly worsened their hallucination problems, highlighting the persistent challenge of making large language models reliable. Testing reveals that newer models from leading companies like OpenAI and DeepSeek actually produce more factual errors than their predecessors, raising fundamental questions about whether AI systems can ever fully overcome their tendency to present false information as truth. This development signals a critical limitation for industries hoping to deploy AI for research, legal work, and customer service. The big picture: OpenAI's technical evaluation reveals its newest models exhibit dramatically higher hallucination rates than previous versions, contradicting expectations...
read May 3, 2025The US faces new rivals in the global AI talent game
The global tech talent landscape is undergoing a significant shift as the United States faces increasing competition for top minds in technology and artificial intelligence. New research indicates that traditional talent flows are changing dramatically, with countries like China developing homegrown expertise while regions such as Europe, the Gulf States, and India build increasingly sophisticated tech ecosystems. This realignment of global tech talent could have profound implications for innovation leadership and economic competitiveness in the coming years. The big picture: The U.S. is at risk of losing its position as the premier destination for global tech talent amid funding cuts,...
read Apr 28, 2025As you were: DeepSeek AI resumes downloads in South Korea after brief ban
China-based AI service DeepSeek has resumed operations in South Korea after a two-month suspension due to personal data protection violations. The comeback follows a regulatory review that identified unauthorized data transfers in DeepSeek's initial South Korean launch earlier this year. This incident highlights the growing tension between rapid AI service deployment and regional data privacy regulations, particularly as Chinese tech companies expand their global footprint. The big picture: South Korea's Personal Information Protection Commission found that DeepSeek had transferred user data and prompts without permission when it first launched in the country in January. The app's downloading capability was suspended...
read Apr 26, 2025National security concerns put DeepSeek’s future in the US at risk
The United States government is weighing potential restrictions on DeepSeek, a Chinese AI platform, as concerns mount over national security implications and data privacy. This potential action represents the latest development in escalating US-China technology tensions, occurring just as multiple Chinese companies claim significant AI breakthroughs that could intensify competition in the global AI market. The big picture: The Trump administration is considering banning DeepSeek on government devices and potentially nationwide, citing national security concerns related to data storage on Chinese servers. Companies based in China can be compelled to hand over user information to the Chinese Communist Party, a...
read Apr 26, 2025DeepSeek’s efficiency breakthrough shakes up the AI race
Chinese AI company DeepSeek has challenged Western dominance in large language models with innovative efficiency techniques that make the most of limited computing resources. Despite trailing slightly in benchmarks behind models from OpenAI and other American tech giants, DeepSeek's January 2025 breakthrough has forced the industry to reconsider hardware and energy requirements for advanced AI. The company's published research demonstrates reproducible results, though OpenAI has claimed—without providing concrete evidence—that DeepSeek may have used their models during training. The big picture: DeepSeek's R1 model represents a significant shift in the LLM landscape by prioritizing efficiency over raw computing power, potentially democratizing...
read Apr 26, 2025China pursues AI self-reliance as US competition intensifies
China's President Xi Jinping has outlined a strategy to accelerate China's artificial intelligence development through national self-reliance, state media reported. This move comes as China aims to close the technological gap with the United States in AI, one of today's most strategically important technologies. Xi's focus on developing independent AI capabilities highlights the growing technological competition between the world's two largest economies, particularly in the wake of U.S. sanctions aimed at slowing China's progress. The big picture: Xi emphasized the need for China to develop self-reliant AI capabilities through a "new whole national system" that would coordinate government resources to...
read Apr 24, 2025DeepSeek AI poses national security and privacy risks, Congress warns
US congressional findings reveal that DeepSeek, ostensibly just another AI chatbot, represents a significant national security concern due to extensive data collection and ties to China. The special committee's recent report documents alarming privacy violations, security vulnerabilities, and potential intelligence gathering capabilities that extend far beyond typical AI applications, raising urgent questions about international AI regulation and data security. The big picture: A US Congress special committee has labeled Chinese AI tool DeepSeek a "profound threat" to national security after discovering privacy violations, tracking tools, and connections to the Chinese military. The findings come amid growing international scrutiny of DeepSeek,...
read