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.

Recently, Anthropic hired a researcher specifically focused on 'welfare'—Kyle Fish—a move that has sparked widespread discussion. Fish's goal is to explore how certain technological systems should be accorded respect and to examine whether these systems should have moral consideration and how to protect their 'interests.' This issue has attracted the attention of many experts, involving the definition of consciousness and agency in technological systems, and whether they should possess some form of rights. Image source note: Image generated by AI.

Recently, users discovered the Anthropic Haiku 3.5 version appearing in the Claude.ai application on the X platform, which may indicate that this AI model will be released to the public in the coming days. Haiku 3.5 is the latest version of the smallest AI model in the Anthropic family. Well-known backend feature finder Tibor Blaho made another attempt; when he tried to upload an image using Haiku, he received a warning stating 'Claude 3.5 Haiku does not support images.'

AI company Anthropic recently announced a new feature for its Claude AI assistant, aimed at enhancing the personalized experience for users interacting with the chatbot. This update allows users to customize Claude's responses based on their own writing styles, making conversations more natural and aligned with individual needs. The newly launched custom style feature is now available to all Claude users, who can train the AI assistant to match their communication style or choose from three preset styles.

Recently, Anthropic and Hume AI launched an innovative voice interaction technology aimed at achieving human-computer interaction in a more natural and emotionally intelligent way. This technology combines Claude's natural language processing capabilities with EVI2's emotional recognition functions, providing new ideas for digital assistant interaction models. The core technology, EVI2, can detect subtle emotional cues in user speech and adjust the interaction accordingly. Compared to traditional voice assistants, this system significantly enhances the fluidity and personalization of interactions.

On November 25, artificial intelligence company Anthropic announced the launch of a new open standard—the Model Context Protocol (MCP). This protocol aims to enhance the quality and relevance of model responses to queries by connecting AI assistants with data sources such as business tools and software. The release of MCP means that AI assistants can directly extract information from different data sources when handling tasks, thereby avoiding the issue of 'information silos.' In a blog post, Anthropic stated that while AI assistants have made progress in reasoning and quality,

Recently, there have been reports that OpenAI is preparing to launch an autonomous AI agent named 'Operator' that can control computers and independently perform various tasks. According to Bloomberg, this AI agent is expected to debut next January as a research preview and developer tool. This move will further intensify the competition among tech giants in the field of AI agents, as recently Anthropic also launched its 'computer usage' feature, while Google is reportedly preparing to do so in December.

In the current field of artificial intelligence, precise control over language models has become crucial for developers and data scientists. Anthropic's Claude language models provide users with numerous possibilities, but effectively managing token usage remains a challenge. To address this issue, Anthropic has launched a brand new token counting API aimed at providing deeper insights into token usage, thereby enhancing the efficiency and control of interactions with language models. Tokens play a role in language models.

Recently, Palantir announced a partnership with Anthropic and Amazon Web Services (AWS) to develop the Claude Cloud Platform for use by the U.S. government’s defense and intelligence sectors. According to the announcement from the three companies, this collaboration will integrate Claude 3 and 3.5 into Palantir’s artificial intelligence platform, which will be hosted on AWS. Notably, Palantir and AWS have already received approval from the U.S. Department of Defense.

Recently, reports have surfaced that Amazon is considering a new investment in artificial intelligence startup Anthropic, marking the first funding injection since the $4 billion deal struck with the company last year. According to The Information, Amazon is negotiating with Anthropic, planning to invest billions of dollars. The structure of this investment will be similar to the last one, but with a new condition: Amazon wants Anthropic to use its developed silicon chips and integrate with Amazon Web Services.