Realme GT5 Pro Equipped with 'New Little Assistant', Accessing 7 Billion AI Edge Model


It is irreversible that Google has decided to retire its long-standing voice assistant, Google Assistant. According to foreign media reports on September 29th, Google has sent out emails since last month, announcing that the Assistant on mobile devices will be discontinued; now many Android users have reported being unable to access it. Some users also found that the option to switch to Google Assistant in the Gemini account menu has disappeared. From options to services, the old assistant is gradually being removed.
CCTV: Yueqing, Zhejiang police busted an AI-voice scam gang; in two months over 1,000 victims, >10m yuan involved, 30 arrested. In July, a man borrowed 15,500 yuan, then was tricked into returning 5,500 yuan as a fake 'funder advance'.....
On September 28, AMD acquired startup company World Labs through a stock transaction worth approximately $8.2 billion (about 55.1 billion RMB). Co-founder "AI Goddess" Fei-Fei Li will join AMD as Executive Vice President and Chief Scientist, reporting directly to Dr. Lisa Su. World Labs was established in 2024, focusing on "world models," enabling AI to understand and build the world based on text, images, and visual information.
On September 28, xAI launched Team Bots, a type of Grok agent that works alongside teams and learns with them. After granting access to the necessary files, applications, and expertise and sharing them, all team members can collaborate based on the same context, while personal chats remain private. Each Team Bot is built around a shared role or workflow within the team, composed of four key capabilities, emphasizing context provision.
OpenAI announced that the 200 dollar Pro subscription will restart tomorrow. The API quota at the same price will be reduced to half of the original plan, and the controversial 5-hour usage limit will be removed, allowing developers and enterprises to freely manage their weekly quotas. This move marks a substantive shift in its large model commercial billing logic: while the quota is nominally halved, it is actually backed by a significant improvement in the underlying model's reasoning efficiency and a tiered reduction in API pricing.