Sam Ringer, a researcher at Anthropic, recently demonstrated the new capabilities of Claude, the company’s AI chatbot. In a video, Ringer showed how Claude can now use a computer like a human would, by looking at the screen, moving the cursor, clicking buttons, and typing text. In the demo, Ringer had Claude fill out a fictional vendor request form.
The AI searched for information on spreadsheets, completed the form, and sent it back to the requester. This showcased how Claude could handle the kind of routine, tedious work that many people currently do. Mike Krieger, Anthropic’s chief product officer, believes this technology could lead to more innovative products and streamlined back-office processes.
He said it could be particularly useful for tasks like conducting online research, testing new software, and automating complex multi-step processes. Anthropic isn’t the only company working on this kind of AI.
Anthropic’s chatbot demonstrates new abilities
Competitors like OpenAI, Google, and Microsoft have shown similar capabilities, such as AI tools that can see your computer screen. However, Anthropic is one of the first to push forward with a more interactive, task-performing AI. The U.S. market for AI software development is growing rapidly.
It’s expected to generate $214 billion in revenue by the end of 2024, which is $50 billion more than last year. By 2030, it could reach $1.3 trillion and contribute a 21% net increase to the U.S. gross domestic product. But consumers are still cautious.
A Forbes survey found that while 64% of businesses think AI will boost their productivity, over 75% of consumers are worried about the trustworthiness of information that comes from AI. As AI keeps evolving and becoming a bigger part of our daily tasks, finding a balance between innovation and trust will likely be a key focus for both developers and users.
