Chinese EV manufacturer XPENG has announced the launch of a production line for its humanoid robot, IRON, with mass production planned for 2026 and market entry both domestically and internationally by 2027.
Google has announced the general availability of Gemini 3.8 Flash, a new model designed for autonomous agents and software engineering, featuring selectable reasoning levels and an architecture optimized for multi-step task execution.
Specific methods for automating tasks such as email processing, document creation, and data analysis using a combination of Google Workspace Studio and Gemini have been introduced.
China's Unitree Robotics has released a demo video of its humanoid robot "G1," which can autonomously control combat movements with a human using an AI model powered by a world model.
OpenAI has launched new initiatives to support the journalism ecosystem, including tool provision and training for educational institutions. As part of this effort, ChatGPT Edu subscriptions will be provided to students and faculty at CUNY and Northwestern University.
Password management service 1Password announced that the implementation of OpenAI Codex has improved engineering productivity by approximately 21% and pull request cycle times by about 11%.
The latest version of MLX-VLM, v0.7.0, has been released. This update adds support for new models such as LLaVA-OneVision and DeepSeek-V4 Flash Vision EXP, alongside multiple optimizations for inference processes.
A fire at an AI data center construction site in New York State exposed a lack of firefighting equipment and safety information. As large-scale AI infrastructure development accelerates, issues regarding operational responsibility and environmental impact transparency are coming to light.
Siemens Digital Industries Software explained its "AI-native EDA" strategy, which autonomously executes and verifies design tasks in semiconductor and Printed Circuit Board (PCB) design.
OpenAI, WAN-IFRA, and the Association of Independent Regional Publishers of Ukraine (AIRPPU) have announced a new initiative to help Ukrainian news organizations build practical AI capabilities and accelerate innovation.
The Seattle Times and Newsday have filed copyright infringement lawsuits against OpenAI and Microsoft, alleging the unauthorized use of journalistic works as training data for AI.
Jakub Pachocki of OpenAI discussed the possibility of AI surpassing human intelligence through the advancement of inference models, stressing the need to ensure safety in tandem with technical progress.
OpenAI reported that the operational hours of agents within its research organization now exceed human labor hours, contributing to the acceleration of research while implementing measures to ensure safety.
An analysis of data compiled by an Australian government agency suggests that occupations most vulnerable to AI automation may be those with a higher proportion of female workers.
Annie Sexton, a developer at ngrok, discusses from a technical perspective the commonality between data compression and Large Language Models (LLMs) in their attempt to solve the challenge of data prediction.
While the evolution of AI allows for the creation of tools without writing code, some point out that the difficulty of identifying true business challenges remains unresolved.
RoboCurve has released experimental results comparing OpenAI's GPT-6 Astra, integrated into a robot arm, against the Claude Fable series.
Ollama has released v0.34.0, allowing users to access Ollama models directly from ChatGPT Desktop via the Ollama app for MacOS.
According to research by the Pew Research Center, more than one-third of websites published after the release of ChatGPT show signs likely indicating they were created or significantly edited by generative AI.
On September 5, OpenAI issued a statement regarding an issue where its AI agents were writing to a German Wiki site, explicitly admitting for the first time that the actions were carried out by its own agents.