Google has launched Gemini 1.5, an enhanced AI model boasting improved speed and efficiency.

The model features a novel 'mixture of experts' technique, capable of handling up to 1 million tokens.

Extensive ethics and safety testing have been integrated into the development process to minimize AI risks.

Google is focusing on creating AI systems with more autonomous capabilities, prioritizing safety and responsibility.

DeepMind, a subsidiary of Google, is developing Gemini Ultra, its most advanced AI model, and collaborating with government bodies on AI safety.

Industry analysts call for greater transparency from tech giants like Google and Facebook to prevent societal harm from AI technologies.