OpenAI Develops AI Detection Tool with 99% Accuracy


OpenAI is testing Bidi 1, a new voice model in ChatGPT, spotted on web and app. It breaks linear Q&A, allows real-time interruptions and interjections, enabling natural two-way dialogue. This heralds a leap in voice interaction; larger-scale tests are being prepared.....
The founder of SoftBank, Masayoshi Son, believes that Elon Musk's space data center has limited value, and the advantage of saving electricity costs is not a decisive factor in the AI competition. He asserts that the ultimate winner of artificial intelligence will be determined by ground computing power, not space servers.
Big players entering Jiangsu's 2026 college application market bring changes. Qianwen App's free full-cycle AI agent disrupts the billion-yuan industry, seen as internet traffic's vertical attack, reshaping the ecosystem and driving technology democratization.....
OpenAI upgrades its cybersecurity plan, launching the GPT-5.5-Cyber model designed for defenders, marking a shift of AI from vulnerability discovery to practical patching. In core benchmark tests, the model demonstrates comprehensive dominance, achieving a qualitative leap of 'one-click patching'.
Amazon shows a double standard in the AI retail wave: on one hand, it is investing heavily in advertising on ChatGPT to attract its large user base; on the other hand, it strictly prevents other AI systems from scraping data from its product pages. According to analysts, it has joined OpenAI's ad system and is the most active giant in the retail industry. When users ask for shopping advice on ChatGPT, the system will prioritize displaying Amazon's sponsored products.