At Dreamforce 2025, Sundar Pichai, CEO of Google and Alphabet, announced significant developments for Gemini 3.0, expected to launch later this year. He described the advancements as extraordinary, suggesting that the upcoming model will be more advanced than its predecessors, emphasizing the evolution towards “really intelligent agents.” While details on Gemini 3.0 remain scarce, Pichai’s remarks indicate that Google is accelerating AI development and plans to integrate Gemini more thoroughly into its products. Gemini 1.0 introduced large-scale multimodal reasoning, while Gemini 2.0 enhanced agentic features. Recently, Google released the Gemini 2.5 Computer Use model, an advanced version that interacts with user interfaces, allowing AI agents to perform tasks like filling out forms and navigating logins through the Gemini API. Competitors, including OpenAI and Anthropic—who recently launched Claude Sonnet 4.5—are intensifying efforts in next-gen AI, prompting Google to innovate rapidly.
Source link
