vivoAI Officially Releases the Open Source BlueLM-7B Model Series, with Plans to Release 13B and Multi-modal 7B-vl Models in the Future

Recently, vivo AI has officially open-sourced the BlueLM-7B base model and the 7B chat model, along with the long-text base and chat models that support 32K tokens. This model boasts a scale of 2.6 trillion tokens and includes data in Chinese, English, and a small amount of Japanese and Korean. Moving forward, vivo plans to open-source the 13B model and the multi-modal 7B-vl model. vivo's open-source initiative is instrumental in promoting the widespread adoption and development of AI technology, encouraging more businesses and developers to utilize and enhance vivo's AI technology.

AI and Organizational Efficiency Drive Changes as Canva Launches Large-Scale Restructuring

Australian design software company Canva has launched a large-scale organizational restructuring, with 1,271 employees facing layoffs, role changes, or relocation, accounting for 22% of its total workforce. CEO Perkins said that the company is experiencing its first growth slowdown in nearly a decade, and its existing structure is difficult to adapt to the next stage, requiring restructuring to improve efficiency. The product, growth, and platform teams will be merged into a unified product organization, simplifying management layers, integrating overlapping responsibilities, and forming smaller teams.

Anthropic Updates Claude to Restrict Remaking Existing Video Games

Anthropic recently updated its AI assistant Claude, formally introducing coding restrictions that prohibit users from using the model to recreate existing commercial video games. System instructions exposed on X show that Claude may only create original games, games of specific genres, and games using users' own designs and assets. It is strictly prohibited from generating remakes of any existing games, even if they are renamed or disguised as original projects. Technologist Addy Osmani has publicly confirmed this.

Has AI Taken Away Programmers' Jobs? Entry-Level Coders' Employment in the US Dropped 20%, Experts Worry About a Talent Shortage in Ten Years

CNBC reports that since February 2025, when Anthropic's Claude Code was launched, the number of software development job postings in the United States did not decrease but instead increased. Data from Indeed Hiring Lab shows that software job hiring has risen by nearly 15% against the trend, while the overall job market has shrunk by 7%. Kuch, chairman of McKinsey North America, was not surprised: the total volume of code written by companies has reached a new high, and the demand for developers remains strong in the AI era. However, the growth is highly concentrated in senior and AI-related positions, with senior roles absorbing most of the increase.

Wang Xingxing of HuaShu Technology Speaks Out: The Biggest Mistake in the Past Was Not Enough Belief in AI Embodied Intelligence, the Inflection Point of Which May Erupt Within One or Two Years

At the opening ceremony of the 26th China International Industrial Exhibition, Wang Xingxing, founder of HuaShu Technology, gave a speech stating that looking back on his entrepreneurial journey, his biggest mistake was not believing enough in the development of artificial intelligence. Regarding when embodied intelligence will take off, he provided specific criteria: when robots can complete tasks based solely on human voice commands in 80% of unfamiliar and unknown scenarios, it means the inflection point has arrived. This standard emphasizes the generalization of open-scene environments and natural language interaction capabilities, reflecting his renewed assessment of AI-driven robotic breakthroughs.

Altman Faces IPO Rumors: No Clear Timeline for OpenAI's IPO, Safety Commitments Must Come Before Capital

In response to months of speculation about an IPO, OpenAI CEO Altman said after the developer conference that the company will only formally consider going public if it can make more comprehensive commitments to model safety, and there is currently no set timeline. He emphasized that with the leapfrog development of large models, the stronger the system, the more solid the safety statements must be; he also admitted that if the wait for an IPO becomes too long, it could have an impact.