Companies have been racking their brains for years to automate repetitive processes. Anthropic AI wants to be your solution

Putting nuts on is not fun, but for years it was assembly line employees who did it. Robots ended up replacing human beings in many of these processes, and for years they have sought to do the same in tasks that are performed in front of a computer and are equally repetitive.

Robotic Process Automation (RPA). This discipline combines the use of APIs (Application Programming Interfaces) with interactions with the user interface of a computer. The objective, to perform repetitive tasks in productivity and, above all, business applications.

Emulating human interaction. These processes attempt to complete the autonomous execution of various activities and transactions in all types of software systems. The idea is to emulate the interaction of a human user and reproduce it, thus avoiding investing time in that to spend it on more complex tasks and thus generate a greater return on investment by employees.

RPA was not AI (but now it is). Until now this type of automation could not be considered an artificial intelligence system because RPA is process-based, and AI is data-centric. Anthropic and its new ‘Computer Use’ feature They just showed that both things can be combined, and it is possible to automate interactions with the user interface to execute all kinds of repetitive tasks.

This image shows the ‘Computer Use’ AI function completing a form by taking information from a CRM system. The dream of many companies comes true.

At Anthropic they are clear about it. Mike Krieger, director of Anthropic, raised the possibilities of this function in an interview with VentureBeat. He explained that “we anticipate it will be especially useful for tasks such as online research, performing repetitive processes such as testing new computer programs, and automating complex multi-step tasks.” And who loves automating complex tasks? Exactly .To companies.

A business candy. The ‘Computer use’ function is capable of recognizing what is on the screen to act accordingly. The level of automation is promising for sectors such as finance, but also for legal services and customer support. Krieger explained that “Claude can open spreadsheets, do analysis and create visualizations. For customer services, he will be able to navigate CRM systems to find and update information on any customer.”

And huge potential. Features like this present a promising solution for data entry, or customer service and support services, but that automation can go much further by being able to recognize and analyze the information you see on the screen and then act on it. This is precisely what AI agents promise, and this can be a great early example of their capability.

But. At the moment ‘Computer Use’ is available on a limited basis, and as Anthropic clarifies “it cannot use your PC as such.” The tool is available as part of an isolated environment in which we can execute all types of tasks with mouse movements and autonomous and AI-generated keystrokes. Even so, at Anthropic they have developed mechanisms to detect if the feature is being used for dangerous activities, such as possible malicious interactions with government portals, for example. The power of the tool can certainly also be put to malevolent uses, as we are seeing with generative AI and deepfakes.

In Xataka | Microsoft begins offering autonomous AI agents. At the moment they are hypervitaminized IFTTT recipes