Anthropic wants its AI assistant Claude to be your favorite coworker

Anthropic has upgraded its Claude generative AI assistant to be extra helpful within the workplace. Claude Professional and Claude Staff subscribers can now higher manage and monitor their work with the AI assistant because of the brand new Tasks and Artifacts options.

Tasks are each a spot for storing and interacting with knowledge for duties. Customers can add all their paperwork, code, and different related knowledge into one place. Every undertaking inside Claude.ai features a 200K context window, equal to a 500-page e-book. They’ll then ask Claude about it, and even arrange customized directions for how you can reply by way of tone or the context of who’s asking and what they could want. The concept is to keep away from what Anthropic calls “chilly begins,” the place customers should begin from scratch every time they interact with the AI assistant. By having a information base to attract from, Claude can reply to queries extra rapidly and precisely.

Workplace Artifacts

The Artifacts characteristic is a form of flip of the Tasks in that it could actually produce a variety of content material the way in which Tasks can retailer them. Customers can ask Claude to make textual content, code, and different ‘artifacts.’ Claude will share the output in a devoted window alongside the chat, like a preview window of what it is composing. This setup permits customers to see and work together with the generated content material in actual time, offering rapid suggestions and changes the place wanted. A further improve lets customers share one of the best bits of conversations with Claude with their workforce in a shared undertaking exercise feed.

Each Tasks and Artifacts are powered by Claude 3.5 Sonnet, Anthropic’s newest AI mannequin. In line with the corporate, Claude 3.5 Sonnet outperforms lately introduced fashions like GPT-4o and Google’s Gemini 1.5 on quite a lot of benchmarks.

“Our imaginative and prescient for Claude has all the time been to create AI programs that work alongside individuals and meaningfully improve their workflows,” Anthropic defined in a weblog publish. “With this new performance, Claude can allow thought technology, extra strategic decision-making, and distinctive outcomes.”