Anthropic gives its AI models limited ability to control your computer

Anthropic has upgraded its Claude 3.5 Sonnet model to control a user's computer and access the internet, marking a significant step in AI models' capabilities. This new feature allows Claude to perform tasks such as creating a personal website or planning a trip. The new capabilities are being rolled out to developers through an API as a public beta, with the company seeking feedback on its performance and usefulness. However, this increased power raises concerns about potential risks, and Anthropic advises users to limit the AI's access to sensitive data and websites, and to have a human confirm decisions with real-world consequences. As AI models become more autonomous, they could bring convenience but also potential harm. Other AI companies are expected to follow suit in the near future.