On Tuesday, Anthropic, the AI research and safety company, announced an upgraded version of its Claude AI model, Claude 3.5 Sonnet, and a new model, Claude 3.5 Haiku, that could transform the way businesses automate complex workflows.
It also introduced a groundbreaking new capability called “Computer Use,” which allows Claude to use computers like a human, navigating screens, clicking buttons, and typing text. This feature could benefit businesses that rely on repetitive tasks, such as data entry, research, and customer service.
In an exclusive interview with VentureBeat, Mike Krieger, Chief Product Officer at Anthropic, said, “Computer use capabilities have the potential to change how tasks that require navigation across multiple applications are performed.”
He added, “This could lead to more innovative product experiences and streamlined back-office processes.” Kreiger implied that the technology is still in its early beta phase, but as it evolves, it could improve data analysis, visualization, and user interface interactions.
Anthropic’s early partners, GitLab, Replit, and Canva, are already said to be benefiting from the Claude 3.5 Sonnet’s new features. GitLab has been testing the model for automating tasks in the development processes, and it has improved reasoning capabilities by up to 10%
Replit and Canvas have been using it to fix bottlenecks in software development and speed up design and creativity.
‘Computer Use’ isn’t confined to any specific workflows or software programs; instead, it can “see” a screen using screenshots, interact with various applications, and adapt to different tasks as they come up. It implies that it’s better than RPA technologies.
However, this ability raises serious security and privacy concerns. Thus, for preventive measures, the company made it clear that Claude cannot access a computer without a developer providing the necessary tools.
‘Computer Use’ is available in limited public beta only through an API, allowing developers to test it in controlled environments.
