Anthropic’s CEO says AI boom is real, but some companies are YOLO-ing

Nvidia CEO Jensen Huang acknowledged uncertainty regarding the potential existential risks posed by artificial intelligence (AI) during a recent discussion, emphasizing that the impact of AI on society and global geopolitics remains largely unknown. He highlighted the importance of the U.S. maintaining a competitive edge in AI against China and other adversaries.

Google has been found to gather three times the amount of AI training data compared to OpenAI, leveraging its search monopoly to enhance its AI capabilities. This revelation comes amid ongoing antitrust scrutiny, where Google acknowledges using publisher content for AI training despite potential legal challenges.

Google Cloud has entered a multi-year partnership with the AI coding startup Replit, aiming to enhance its competitive edge against rivals such as Anthropic and Cursor. This collaboration is expected to leverage Replit's innovative coding platform alongside Google's cloud offerings and AI models.

Google Workspace has introduced new AI automation features, intensifying competition with OpenAI as the tech giant seeks to enhance its productivity tools. This update includes a one-page template for daily releases and insights from Anthropic on employee AI usage, reflecting a broader trend in AI integration within workplace environments.

A recent study has applied Singular Learning Theory (SLT), a framework inspired by physics, to analyze grokking and other phase transitions in neural networks. The research empirically investigates SLT's free energy and local learning coefficients, revealing insights into the behavior of neural networks under various conditions.

Moore Threads Technology Co., a prominent Chinese AI chipmaker, commenced trading on the Shanghai stock exchange, marking the second-largest initial public offering (IPO) in China this year. This significant event underscores the growing interest and investment in the AI sector within the country.

OpenAI researchers have developed a new method termed 'confessions' that encourages large language models (LLMs) to self-report errors and misbehavior, addressing concerns about AI honesty and transparency. This approach aims to enhance the reliability of AI systems by making them more accountable for their outputs.

GitHub Copilot users can now access Anthropic's Claude Opus 4.5 model in chat across Visual Studio Code and Visual Studio during a new public preview, enhancing the AI capabilities available for software development.