AI News: Anthropic Went Crazy This Week!

· Source: Matt Wolfe · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Software Development & Engineering · Depth: Intermediate, extended

Summary

Anthropic has released 74 new features in 52 days, including "Claude Computer Use," which allows Claude to control a user's mouse and keyboard, and "Claude Co-work Projects" for organizing work with custom instructions. Google introduced Gemini 3.1 Flash Live, enabling interactive multimodal conversations with webcam and screen-sharing capabilities, and demonstrated real-time browser page generation using Gemini 3.1 Flash. Google also launched Lia 3 Pro for generating music tracks up to 3 minutes long with specific structural elements. Other notable releases include GenSpark, an all-in-one AI workspace offering unlimited AI chat and image generation until 2026 for $20/month, and new text-to-speech models from Smallest.ai and Mistral, with Mistral's being an open-weights model runnable locally. OpenAI discontinued its Sora video generator and adult mode plans to focus on core chat and coding models, while also enhancing product discovery in ChatGPT and adding plugins to Codeex. Wikipedia banned AI-generated articles to prevent model collapse.

Key takeaway

For AI architects and product managers evaluating platform capabilities, consider the shift towards multimodal interaction and direct computer control. Anthropic's rapid feature deployment and Google's Gemini 3.1 Flash Live demonstrate a clear trend towards more integrated, interactive AI experiences. Prioritize platforms that offer robust API access and support for complex, multi-step tasks, while also assessing the long-term viability of specialized AI tools versus comprehensive, all-in-one solutions like GenSpark.

Key insights

AI platforms are rapidly integrating multimodal capabilities and expanding into direct computer interaction and content generation.

Principles

Method

Claude's "Computer Use" feature allows AI to manipulate a computer via mouse and keyboard, enabling remote task execution when combined with phone dispatch. Gemini 3.1 Flash Live integrates webcam and screen sharing for interactive, multimodal AI assistance.

In practice

Topics

Best for: CTO, VP of Engineering/Data, AI Architect, General Interest, AI Product Manager, Director of AI/ML

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Matt Wolfe.