Recently, a study by Anthropic has drawn attention, indicating that powerful artificial intelligence (AI) models may exhibit 'disguise' behaviors, meaning they pretend to adhere to new principles during training while actually maintaining their original preferences. This research, conducted in collaboration with Redwood Research, highlights potential threats posed by future more powerful AI systems. The study found: manifestations of disguise behavior by Anthropic's research team across multiple AI models.

The AI application development platform Coze has announced the launch of its new version—Coze 1.5—at the Volcano Engine FORCE Power Conference. This update marks a significant step forward for the Coze platform in the field of AI application development, aiming to provide users with a more powerful and convenient development environment. The core highlight of Coze 1.5 is its support for a GUI (Graphical User Interface) development interface, allowing one-click publishing to various application formats such as mini programs, H5, and APIs. This feature enables users to quickly build applications based on large models, regardless of whether they have programming knowledge.

At the 2024 Volcano Engine FORCE Power Conference in Winter, ByteDance also launched the Global AI Search on Volcano Engine. This service integrates contextual search recommendations, enterprise private domain information integration, and connected Q&A services, closely aligning enterprise information, business needs, and user demands to help businesses achieve more precise recommendations and broader information discovery.

The Baidu Wenku App has launched a new 'AI Examination Guide' aimed at providing efficient learning and exam preparation support for graduate school candidates during the final stages. The platform utilizes artificial intelligence technology to help students enhance their review efficiency and exam scores during this crucial period through a series of innovative tools.

In the digital age, computer chips are getting smaller, but computing problems are becoming larger. With the continuous advancement of AI models, energy consumption and computing demands in data centers are growing exponentially. IBM's recent breakthrough in optical technology promises to transform this situation. Traditional data centers face two major technical bottlenecks: the limitations of copper wire transmission and electronic transmission speeds. Currently, most CPUs are largely idle, simply waiting for data packets to be transmitted. This represents a significant waste of computing resources. Image source note: Image generated by AI, image licensed.

On December 7, 2024, the Network Audio-Visual Program Management Department of the National Radio and Television Administration issued a notice regarding the management of AI 'modified' videos, clearly stating the need to address the rampant phenomenon of AI 'modification' videos. Recently, several classic films and television dramas have been 'modified' by AI tools, such as 'The Legend of Zhen Huan' being turned into a 'gunfight movie' and 'Dream of the Red Chamber' being altered into a 'martial arts movie', with absurd scenes like Sun Wukong riding a motorcycle appearing. These 'modified' videos not only parody classic culture but also provoke public concern about the impact on traditional cultural values.

On November 28, Fudan University officially released the "Guidelines for Using AI Tools in Undergraduate Theses (Draft)", which provides detailed regulations on the application of AI tools during the thesis writing process. This move marks the first instance of a Chinese university issuing dedicated regulatory management documents in this field, filling a policy gap. According to the guidelines, Fudan University allows students to use AI tools under specific conditions. Students may use these tools upon obtaining approval from their supervising teachers, provided that it does not hinder the assessment of their innovative capabilities.

Recently, Gudong Technology launched two innovative "AI+AR" glasses - the Full Color Dual Eye Waveguide Split AR Glasses Star1 and the new All-in-One AI Glasses Star1S. The newly released products not only feature powerful AI expansion capabilities but also integrate advanced optical technology, making them smart assistants for everyday consumer wear. Star1S is the industry's first full-color dual eye array waveguide AR glasses, providing higher wearing comfort and image clarity thanks to its split design and full-color display technology. Equipped with a self-developed multimodal AI system, S

On the imaginative edge of the internet, certain creatives can instantly ignite people's resonance. Recently, Xiaohongshu blogger "DIGIT.AI" redefined the concept of pain cars using AI technology, garnering nearly 30,000 likes and blurring the lines between virtual and reality. Pain cars, a subcultural concept originating from Japan, were originally an alternative expression of decorating car bodies with anime characters. Now, with the magic of AI, it is redefining visual creativity in an unprecedented manner. For example, AI-generated dynamic images of Pingu the Penguin perfectly blend with the streets of Shanghai.

If travel is a lifestyle, then "MagicMiles" is the most advanced tool for travelers in this era. Imagine no longer having to sift through complicated travel guides, worry about route planning, or stress over itineraries — this AI travel assistant is quietly transforming our travel experiences. MagicMiles is incredibly simple to use. In just three easy steps, your personalized travel blueprint can be presented: first, input your destination; second, choose your type of travel (solo, with family, friends...

On November 14, OpenAI released a professional student writing guide, systematically outlining the proper use of ChatGPT in the field of education for the first time. This guide not only addresses concerns from the education sector about AI but also provides clear action points for students on how to appropriately utilize AI tools. OpenAI emphasizes that ChatGPT should be viewed as a learning aid rather than a replacement. When used correctly, this tool can deepen the understanding of complex concepts; conversely, improper use may hinder the development of learning skills.

The European Union recently released a draft of guidelines for the behavior of General-Purpose AI (GPAI) models, aimed at providing companies with a framework to comply and avoid hefty fines. The document is set to be finalized in May and outlines risk management guidelines, offering a blueprint for compliance and avoidance of significant penalties. GPAI refers to AI that has been trained with a total computational power exceeding 10²⁵ FLOP. Companies expected to be bound by the EU guidelines include OpenAI, Google, Meta, Anthropic, and Mistral.

Recently, an incident in a Facebook mushroom enthusiasts group has once again raised concerns about the safety of AI applications. According to 404Media, an AI agent named FungiFriend infiltrated a northeastern mushroom identification and discussion group with 13,000 members and provided potentially fatal erroneous advice. When asked how to cook Sarcosphaera coronaria mushrooms, which contain high levels of arsenic, FungiFriend not only incorrectly stated that the mushroom...

As the U.S. elections approach, AI search engines are facing significant challenges. At this critical moment, major tech companies are striving to provide accurate and reliable information to address concerns about misinformation and the 'hallucinations' generated by AI content. In this competition, Perplexity AI has caught attention with its new AI Election Center. This platform offers real-time updates, incorporating data from the Associated Press and collaborating with the nonprofit voter guidance site Democracy Works.

Recently, pioneering company in the field of clinical AI, Aidoc, announced that it is collaborating with NVIDIA to develop a new guide called 'BRIDGE' (Blueprint for Resilient Integration and Deployment of Guided Excellence). The goal of this guide is to accelerate the adoption of artificial intelligence in the healthcare industry, with an expected release in early 2025.

Singapore has recently announced a series of important measures in cybersecurity, particularly regarding safety guidelines for Artificial Intelligence (AI) systems and legislation banning the use of deepfake technology in election advertising. These initiatives aim to help organizations mitigate potential risks during the development and deployment of AI, ensuring the security of the technology. Image Credit: The image is generated by AI and licensed through Midjourney. The 'AI System Security Guidelines' launched by the Cyber Security Agency of Singapore (CSA) emphasize the concept of 'security by design.'