Anthropic releases AI agents capable of performing complex tasks

Coinworld News Report:
On Tuesday, Anthropic AI launched a pair of updated AI models with new features that enable them to independently perform computer tasks like humans and save user keystrokes.
This artificial intelligence startup, backed by Amazon and Alphabet, as well as other tech giants like Meta and Microsoft, has formed a group of tech companies in the generative AI arms race as they seek to outdo each other and maintain a leading position.
Anthropic’s new models perform various tasks.
As a competitor to ChatGPT, the maker of OpenAI, Anthropic revealed that it has updated its model, Claude, which can perform various tasks on computer searches, such as web searches, opening applications, and inputting text using a mouse and keyboard.
“The new model with ‘computer usage’ capabilities can tell the AI where to use the mouse and click, what to input, and how to perform complex tasks,” said Jared Kaplan, Chief Scientist at Anthropic.
According to an article by Reuters, the ability to perform customized complex tasks lies in software developers acting as AI agents who can perform tailored complex tasks with minimal interference in executing multi-step operations.
Researchers have referred to these agents as the cutting edge of AI development to achieve better technology beyond just chatbots, as they go beyond simple prose or computer code and instead perform specific actions.
Anthropic demonstrated its capabilities in a use case scenario where it used Google search and Apple Maps to plan a sunrise trip and create a basic website.
Anthropic offers three versions of its AI model, called Claude, to software developers at different prices.
Sonnet is the second version and the cheapest, while Haiku will receive its first update this week.
Kaplan stated that the latest update, version 3.5 of Haiku, can generate computer code at a speed closer to the early versions of Sonnet launched in June.
Anthropic plans to offer the features to everyone in the first quarter of 2025.
Dario Amodei, CEO of Anthropic, told Reuters in an interview in June that their goal is to release the best version of Opus by the end of the year.
According to Anthropic, Amazon has been using the model for a long time, and the initial customers and other testers include Canva and Notion. Kaplan added that the company has been working on this development since the beginning of the year.
Despite the updates released this week, the company expressed its hope to make all features available to all customers either this quarter or in the first quarter of 2025.
According to Anthropic, booking flights, filling out forms, conducting online research, and submitting expense reports will become part of future consumer applications. Kaplan also stated that the goal of this AI startup is to enable Claude to help people solve different types of tasks.
“What would you do if you got rid of a few hours of copy-pasting or whatever it is you eventually have to do? I’d say play more guitar,” said Mike Krieger, Chief Product Officer at Anthropic.
Since its initial release in March 2023, Claude, which directly challenges OpenAI in the market, has made Anthropic one of the most popular AI startups.
These two AI companies have direct competition in both the enterprise and consumer markets.
Tuesday’s announcement is the latest move by Anthropic to build more virtual collaborators or agents, aligning with its long-term strategy.
To facilitate development, the startup launched Claude Enterprise in September, reportedly its best product since its chatbot debut, designed specifically for organizations that want to integrate Anthropic’s AI.
Early customers and testers for this particular technology are reported to include GitLab, Midjourney, and Menlo, according to the company.