Google has made significant strides in its generative AI model, Gemini 2.5 Pro, which promises to challenge ChatGPT’s dominance. The new release comes after the company refocused its efforts following a sudden increase in interest from users. According to Google’s Tulsee Doshi, director of product management for Gemini, the latest update is the result of long-term investments in the AI model.
The process of releasing a new model involves rigorous testing, starting with external academic benchmarks and internal evaluations. Doshi revealed that the team used a multilayered approach, including a game creation platform and complex 3D math problems to test Gemini 2.5 Pro. The results were impressive, demonstrating significant gains in capabilities.
Gemini 2.0 was released just three months prior to 2.5, showcasing the company’s accelerated development pace. Doshi attributed this success to the integration of fundamental building blocks, which have now come together in “really awesome ways.” The new model has been designed to assist users more effectively, saving time and effort.
The Gemini team is optimistic about the future of AI models at Google, with significant advancements expected in the coming months. With Gemini 2.5 Pro, the company may finally be able to make a dent in ChatGPT’s dominance, marking a new era for generative AI products.
Source: https://arstechnica.com/ai/2025/04/with-the-release-of-gemini-2-5-pro-google-looks-toward-a-dynamic-ai-future