ai news, how-tos and comparisions

OpenAI has announced a partnership with Japan's SoftBank, investing $3 billion annually to use OpenAI software. This comes after the rise of DeepSeek, a Chinese AI firm that uses less computing power than OpenAI’s ChatGPT and other products. OpenAI's CEO Sam Altman is considering making their products open-source, similar to DeepSeek's model.

This step-by-step guide explains how to set up and use DeepSeek with Ollama and Docker for local AI model deployment.

DeepSeek's hidden coding potentially allows user data to be sent directly to the Chinese government through CMPassport.com, a site owned by China Mobile. This revelation raises significant concerns about national security and privacy for U.S. users.

A recent report by Feroot Security has revealed potential data privacy concerns regarding DeepSeek, a Chinese AI company. The report claims that obfuscated code in the account creation and login process may be sending user information to China Mobile, a company banned from operating in the US due to national security concerns.

Feroot Security's discovery of hidden data pipelines from DeepSeek to China Mobile servers raises significant security concerns, particularly regarding compliance with PCI DSS 4.0.1 for payment providers who use AI tools that may transmit sensitive cardholder data (CHD). This article will delve into Requirement 6.4.3 of PCI DSS 4.0.1 and its implications on organizations that process card transactions.

China's top AI talents are choosing to stay at home rather than in Silicon Valley, indicating that more and more young engineers prefer to participate in China's local innovation. Zizheng Pan chose DeepSeek after his internship. The startup founded by Li Angfei and fully funded by his investment company High-Flyer has quickly become famous in less than two years for its two advanced and lower-cost AI models, giving Nvidia a $60 billion market value. This reflects the trend of China's top AI talents choosing to stay at home, preferring lower living expenses, opportunities to be close to family, and taking on more important roles early in their careers. At the same time, despite the US restrictions on China's access to chips, young and passionate technical workers such as DeepSeek are working hard to catch up with Silicon Valley tech giants.

The author, Lina M. Khan, argues against the notion that America's tech giants should be shielded from competition to maintain a technological edge over foreign competitors, particularly China. She posits that the recent emergence of strong AI models by Chinese firms, like DeepSeek, is a warning sign about the vulnerability of U.S. tech dominance due to lack of effective competition.

Australia bans the use of DeepSeek on all government devices and systems due to national security concerns. The ban applies specifically to government entities but does not extend to private citizens' devices. DeepSeek is an AI chatbot that recently caused a stir with its performance similar to US rivals at a lower cost. Other countries have also raised concerns about the app, particularly regarding data privacy.

The article discusses Chinese AI companies that could match DeepSeek's impact in the wake of its groundbreaking model release. Companies like Alibaba Cloud, Zhipu, Moonshot AI, ByteDance, and Tencent are highlighted as potential competitors to US tech firms. The article mentions specific models such as Qwen 2.5-Max, AutoGLM, Kimi k1.5, Doubao-1.5-pro, and Hunyuan, noting their performance metrics and pricing strategies.

Konstantin F. Pilz and Lennart Heim explore the interplay between artificial intelligence (AI) development and its deployment, emphasizing ethical considerations and potential impacts on society.

The article 'Is DeepSeek China’s Sputnik Moment?' by John Cassidy discusses the impact of Chinese company DeepSeek's advancements in AI models, which have surprised Silicon Valley and raised concerns that the West may be losing ground to China in various industries. The piece highlights a warning from a longtime China watcher about potential leapfrogging in other sectors as well.

Missouri Senator Josh Hawley has introduced a bill that would criminalize the downloading and use of AI models from China, potentially imposing harsh penalties on individuals and companies. The proposed 'Decoupling America’s Artificial Intelligence Capabilities from China Act' aims to restrict imports and exports of AI technology between the U.S. and China, prohibiting American companies from conducting AI research in China or collaborating with Chinese entities. Critics argue that such legislation is overly broad, could stifle innovation, and infringe on scientific freedom.

Anthropic is seeking applicants for a position and requires them to indicate agreement with the instruction that AI assistants should not be used during the application process. The company values authenticity in communication and non-AI-assisted skills evaluation.

DeepSeek has released a reasoning model called R1 that matches the performance of ChatGPT o1 in generating coherent and detailed responses. This is significant because it demonstrates the potential for smaller companies to compete with larger ones in developing advanced AI capabilities without relying on extensive human supervision or expensive hardware.