The artificial intelligence landscape continues its rapid evolution, with several notable advancements emerging last week. xAI’s Grok Imagine now empowers users to create animated visuals simply by speaking their prompts, offering a more intuitive generative AI experience. OpenAI’s ChatGPT has introduced a branching feature, allowing users to explore diverse conversational paths without losing track of their original discussion. Google has entered the on-device AI arena with EmbeddingGemma, a compact 308 million parameter model. Moonshot AI’s Kimi K2-0905 received an upgrade, now featuring improved coding capabilities and an expanded 256k token context window for more complex tasks. Finally, Alibaba showcased Qwen3-Max-Preview, a massive 1 trillion parameter model promising enhanced reasoning and code generation capabilities. These developments signal continued growth and innovation across various facets of AI.
AI Weekly Roundup: Grok’s Voice-Powered Visuals, ChatGPT’s Conversational Branches, and More
