May 2023 Summaries

Anthropic has raised $450 million in Series C funding led by Spark Capital to support the development of reliable AI products, including the conversational AI assistant Claude. The company aims to grow its product offerings and expand its research on AI safety techniques. With the new funding, Anthropic plans to further develop its technology, including a feature that allows users to submit hundreds of pages of materials for analysis. The company has also strengthened its governance structure to better achieve its Public Benefit orientation and is committed to building safe and capable AI systems. The investment is seen as a significant step towards unlocking the potential of AI in society.

We are announcing a new partnership with Zoom, a leader in enterprise collaboration and communication solutions, which will utilize Claude, our Constitutional AI assistant, to build customer-facing AI products focused on reliability, productivity, and safety. The partnership allows us to bring robust, steerable AI to more people in the workplace, prioritizing safety and helpfulness. Zoom's federated approach to AI will use its own technology plus other models, including Claude, for diverse customer needs. The first product integration of Claude will occur in the Zoom Contact Center portfolio, enhancing the end-user experience and agent performance. Additionally, we have received an investment from Zoom Ventures, which shares our vision of building customer-centric AI products with a foundation of trust and security, enabling beneficial applications that meet customer needs.

Claude's context window has been expanded from 9K tokens to 100K tokens, allowing businesses to submit and analyze hundreds of pages of materials in a matter of seconds. This enables Claude to perform complex tasks such as summarization, explanation, and synthesis of knowledge across multiple documents. The increased context window also facilitates the use of Claude for tasks like digesting dense documents, analyzing strategic risks, and identifying technical questions within large codebases. With this upgrade, businesses can now tap into Claude's capabilities more efficiently and effectively, making it an even more powerful tool for their operations.

Claude's Constitution is a framework for language models to follow explicit values determined by a constitution, rather than relying on implicit values from large-scale human feedback. This approach aims to make the AI system's values easier to understand and adjust as needed. The constitution in Claude's model draws from various sources, including the UN Declaration of Human Rights, trust and safety best practices, and principles proposed by other AI research labs. The principles prioritize common sense, helping users avoid harm, and being respectful and considerate of different perspectives. Constitutional AI is a promising approach to making language models more helpful, harmless, and honest, with benefits including scalable oversight and transparency. However, it's not a panacea, and there will still be questions about what the model can and cannot do.