Anthropic announced the release of two new AI models: **Claude 3.5 Sonnet** and **Claude 3.5 Haiku**. The upgraded Claude 3.5 Sonnet features major improvements, especially in coding, with higher performance in software engineering tasks compared to its predecessor. The new Claude 3.5 Haiku offers state-of-the-art capabilities at the same cost and speed as previous models, outperforming older versions like Claude 3 Opus. A groundbreaking new feature called **computer use** was also introduced in public beta. This feature enables Claude to interact with computers like a human, using the screen, cursor, and keyboard. While still experimental, it holds significant potential for automating complex tasks. Companies like Replit and Canva are already exploring its capabilities. The computer use feature, though promising, is still imperfect, with challenges in performing certain actions like scrolling or dragging. However, the public beta aims to gather feedback for improvement. Safety measures have been developed to prevent misuse, such as spam or fraud. There is also a "chatgpt canvas" like feature that is working even better than ChatGPT The new models and features are available via Anthropic's API, Amazon Bedrock, and Google Cloud’s Vertex AI, with Claude 3.5 Haiku expected to release later in the month.