QIBENQI Preço

On October 23, The Wall Street Journal reported that OpenAI has fired three security researchers. Internal security personnel discovered that they had engaged in disappearing message chats via Signal with external AI security researchers, raising suspicions that sensitive information related to OpenAI's internal model design was involved. This investigation reportedly began following a September article about OpenAI's latest model, Astra. The article mentioned that Astra employs a new design that enhances model capabilities while reducing the observable reasoning processes for humans. OpenAI viewed the related report as a potential leak of internal model design and initiated a probe into the source of the information. The report also highlighted that this controversy touches on the issue of AI model 'monitorability.' Some security researchers are concerned that if future models can perform tasks without leaving complete traces of their reasoning, it may become more challenging for researchers to determine the reasons behind an AI agent's decisions, as they would only be able to observe the outcomes of its actions. OpenAI investigators have also identified Ryan Greenblatt, who assisted METR in analyzing the logs of the July Hugging Face hacking incident, as one of the suspected leakers, a claim Greenblatt denies.

On August 9, a support document titled 'Using Apple Smart with Qianwen on Mac' appeared on Apple's official website in the simplified Chinese Mac user manual on August 8, which explicitly mentioned that Apple Smart could work with Alibaba's Qianwen model. However, this Chinese guide introducing the Qianwen extension has now been removed from Apple's official website. The relevant page in the Apple Mac user manual has been reverted to 'Using Apple Smart with ChatGPT,' instead of the previously displayed 'Using Apple Smart with Qianwen.'

On July 17, at the opening ceremony of the 2026 World Artificial Intelligence Conference, Yin Qi, Chairman of Leap Star and Qianli Technology, attended as a special guest and delivered a keynote speech titled 'When Intelligent Agents Enter the Physical World' at the main forum (morning session). Yin stated that the next wave of growth in the AI industry will come from the integration of intelligent agent models and next-generation terminal products. Intelligence is moving beyond digital screens into the real physical world, which will drive three structural changes in the industry: new systems, new carriers, and new networks. Yin predicts that by 2026, model capabilities will have crossed a critical threshold, with AI evolving from executing simple tasks in seconds to being able to work independently for dozens of hours. Programming languages are becoming the core benchmark for AI capability advancement after natural language, and the industry is now at the foot of the AGI peak. With technological iterations, intelligent agents are becoming the smallest unit of productivity, and in the future, professions such as engineers, designers, and researchers will have dedicated intelligent agents, enabling individuals to possess team-level productivity. Yin pointed out that the wave of intelligent agents will bring about three fundamental industrial transformations: first, at the new system level, the capabilities of intelligent agents will be determined by both model capabilities and the Agentic OS, which connects models with data, tools, and devices, defining the practical action boundaries of intelligent agents and becoming the core infrastructure of the intelligent agent era; second, at the new carrier level, terminal design concepts will shift from 'human-centered' to 'human-machine symbiosis', with computers, smartphones, cars, and robots serving as multi-scenario carriers for the same intelligent agent, supporting continuous existence across terminals, free migration, and collaborative task completion; third, at the new network level, an A2A (Agent to Agent) network will establish a human-machine symbiotic value network composed of both humans and intelligent agents, where intelligent agents will have independent identities and credit systems, enabling them to autonomously engage in collaboration and transactions, thus becoming the foundational support of the intelligent agent economy. (Source: Science and Technology Innovation Board Daily)