This week in AI news, Google's Project Genie debuted, allowing users to generate and control characters in AI-created 3D worlds. Gemini's integration into Chrome brought agentic AI features, enabling automated tasks like email drafting and form filling directly within the browser. Anthropic's Claude introduced connectors for various tools and an Excel plugin, though some integrations showed initial glitches. Decart's Lucy 2 offered real-time character animation, while OpenClaw (formerly Clawdbot), a self-modifying AI assistant, faced rebranding due to naming conflicts and raised security concerns. OpenAI announced the retirement of GPT-4o and related models, alongside a high-cost advertising model for ChatGPT. New open-source models like Kimmy K2.5 and Qwen3 Max Thinking showcased impressive benchmarks, while Microsoft and NVIDIA unveiled specialized AI chips and climate models. Apple's acquisition of Q.ai, a facial expression analysis firm, highlighted growing interest in biometric AI.
AI and Agentic Coding News
Topic
- Jan 30, 2026
- Jan 30, 2026
- Jan 30, 2026
- Jan 30, 2026
- Jan 30, 2026
Google DeepMind's Genie 3 is a groundbreaking AI world model that generates interactive virtual environments from text prompts and images. Wes Roth's demonstrations showcase its ability to create realistic character movements, dynamic object interactions, and intricate environmental details, from a cat exploring a tavern to a hippo moving through mud. The model excels at rendering atmospheric lighting and maintaining environmental coherence, such as distinct trails in a forest. While demonstrating impressive responsiveness and the capacity for complex scenarios like moving trains and multiple interacting characters, Genie 3 still faces challenges with certain complex prompts, occasional glitches like misrendered character features, and the appearance of anomalous artifacts. Beyond entertainment, Google DeepMind intends to leverage Genie 3 for critical applications like training robots in infinite simulated worlds, marking a significant step towards more sophisticated AI and virtual reality experiences.
Jan 30, 2026Explore Project Genie, Google DeepMind's AI platform for creating virtual worlds. Learn how to access it (Google AI Ultra subscription, US location), generate worlds using detailed prompts for characters and environments, and leverage image uploads. Discover navigation controls and essential tips for effective world exploration and creation, including the innovative sketch preview system and the ability to remix existing worlds.
Jan 30, 2026