Claude AI, one of the main competitors of ChatGPT, has recently introduced a groundbreaking feature called „computer use“ that is set to revolutionize the world of artificial intelligence. This new functionality allows the model to interact with a computer just like a human would, performing tasks such as observing a screen, moving the cursor, clicking buttons, and typing text. Currently in public beta, this feature enables Claude to carry out real operations, as announced by the company.
The „computer use“ feature in Claude aims to automate repetitive tasks that often consume time, making the AI even more useful for users. The goal is to simplify computer interaction by allowing the virtual assistant to perform tasks on behalf of the human user. It is important to note that this feature is currently an API for developers interested in integrating such functionality into their applications.
Anthropic, the company behind Claude, has released a couple of videos demonstrating the capabilities of „computer use.“ The AI can fill out forms, browse the web, open downloaded files, execute terminal commands, and more. This feature essentially allows Claude to take control of the mouse and keyboard, as shown in the demos.
While this new feature is groundbreaking, it is still experimental and may have some errors. Anthropic has released it early to gather feedback from developers. The API enables Claude to perceive and interact with computer interfaces, translating prompts into computer commands. Developers can use this API to automate tasks, conduct testing and quality assurance, and engage in open-ended research.
As acknowledged by Anthropic in a blog post, this technology is still in its early stages of development and is said to be „full of bugs.“ During internal testing, Claude, for example, interrupted a programming task halfway through and started opening images of Yellowstone National Park. However, it is common to see immature artificial intelligences, and with time, this technology is expected to mature.
In addition to the introduction of „computer use,“ Anthropic has launched an updated version of the Claude 3.5 Sonnet model, along with a new model called Claude 3.5 Haiku, expected to be released by the end of October. These developments follow the announcement made in August when Anthropic, along with OpenAI, joined the US AI Safety Institute to share their developments, a significant step towards greater safety and transparency in the field of artificial intelligence.
Overall, the introduction of „computer use“ in Claude AI represents a significant advancement in the capabilities of artificial intelligence models. As this technology continues to evolve and improve, it has the potential to streamline tasks, enhance productivity, and provide valuable assistance to users in their daily routines.