Google sets ambitious $75 billion AI investment plan for 2025

In the chase to enhance their AI capabilities, Google has released the Gemini 2.5 Pro model, which is positioned as a more competitive AI alternative amidst ongoing challenges in the field. Despite a slow beginning compared to rivals like OpenAI, the Gemini team is now reporting significant advancements in benchmarks and user experiences. Tulsee Doshi, who is the director of product management for Gemini, expressed optimism about their current trajectory, indicating that the fundamental pieces for success are now aligning effectively. Google's ongoing commitment to the AI arena is underscored by their plan to invest $75 billion in AI infrastructure by 2025, a significant increase aimed at leveraging advanced hardware and improving overall efficiency. However, the Gemini model still faces hurdles. While previous iterations have struggled with accuracy and providing reliable outputs, the Gemini team believes these issues can be mitigated with the new model. Doshi commented on the necessity of improvements in handling simpler prompts, noting that the complexity of the large language models (LLMs) can result in their overthinking simpler queries, which is not ideal for user experience. The acknowledgment of learning from user feedback particularly focuses on enhancing what they refer to as the