Information

On June 30, according to The Insider of Business, Andrew Ambrosino, head of OpenAI Codex, admittedly in his recent podcast, Lenny's Podcast, that AI's ability in the area of creative design remains significantly short — good or bad design is more difficult to quantify than the ability of the code to operate, making it difficult to establish a feedback loop for the training model “What is a good design”. Design is more difficult to rate than software, and it is much more cumbersome to build a cycle that can train models to distinguish between good and bad designs than to judge whether the code can run. Fig...

Jan. 3 (Bloomberg) -- NVIDIA has had a banner year in 2024, not only continuing to see its stock price climb, but even supporting the overall market to some degree. According to a new investment report from Bank of America analyst Vivek Arya, as reported by Wccftech on January 2, the GPU giant appears to be poised to build on its 2024 success. While Arya acknowledges that the media has gone ahead and disclosed the main elements of what NVIDIA will unveil at CES 2025, he ......

Jordi Ribas, Microsoft's vice president and head of engineering and products for Copilot and Bing, recently tweeted that the web version of the Copilot chatbot has gone live with a "notebook" feature that supports up to 18,000 characters. Visiting the page, in the upper left corner of the chat window, in addition to "search" and "Copilot", there is a new "Notebook" function, showing a left and right column design. Microsoft says the "Notebook" feature is ideal for complex queries and projects, making it easier for users to repeatedly enter prompts to get the best answer. ......

On July 9, OpenAI officially launched a new generation of voice models GPT-Live for global iOS, Android and ChatGPT.com users. The central change in GPT-Live is the architecture upgrade: from a "turn system" that used to rely on silent sound testing to a full-scale structure that allows both voice input and output. The model does not wait for the user to speak, but gives immediate feedback such as "mhmm" "got it" when the user speaks, and can wait when the user stops, without environmental noise...

Musk, along with other AI experts, agrees that real-world data for training AI models has almost run out, TechCrunch reports. In a live conversation with Mark Payne, chairman of Stagwell's board of directors, on Wednesday evening, Musk said, "We've now essentially consumed all of the accumulated human knowledge ...... of data used for AI training. This phenomenon basically happened last year." Musk's comments are in line with those made by former OpenAI chief scientist Ilya Sutskever last year ......

According to Reuters, on the 26th of local time, Google DeepMind CEO Demis Hassabis said in a panel discussion at MWC 2024, "We've taken Gemini's portrait-generating feature offline, and we're working on fixing the problem, and hope to bring it back online in the next few weeks." Demis said Gemini's portrait-generating feature "didn't work the way we expected it to." Google launched Gemini's portrait generation feature earlier this month, but some users noted that the feature generated images of historical figures that were sometimes inaccurate, after which Google suspended the feature ......

On 20 July, in a press conference at 10 a.m. on Monday, 20 July 2026, the Press Office of the State Council invited Wang Wei Ming, Chief Engineer of the Ministry of Industry and Information Technology, Ta-chung, Press Spokesman of the Ministry of Industry and Information Technology, Director General of the Directorate of Operational Monitoring and Coordination, Press Spokesperson of the Ministry of Industry and Information Technology, Director of the Information and Communication Management Department, to present the developments in industry and information in the first half of 2026 and to answer questions from journalists. Wang Wei Ming, the chief engineer of the Ministry of Industry and Information Technology, mentioned that the role of new dynamics in new and emerging industries has continued to gain prominence thanks to joint efforts. In the field of intelligent robots...

January 15, 2011 - Xunfei Starfire big model upgraded and released today, officially released Starfire Deep Reasoning Model X1, Xunfei Starfire 4.0 Turbo base ability ushered in a new upgrade, and at the same time debuted the Starfire Voice Simultaneous Transmission big model. KU Xunfei released the Xunfei Starfire Deep Reasoning Model X1, the first large model with deep thinking and reasoning ability trained on a national computing platform. Compared with general large models, Xunfei Starfire X1 is said to be closer to the "slow thinking" method of human beings in the problem solving process, and uses more time to solve problems in a "slow thinking" way. Compared with general-purpose big models, Cyberfire X1 claims to be closer to human "slow thinking" in the problem solving process, and with less arithmetic power, it "realizes the industry's first-class effect, with a number of indexes......