Gemini 4 Release Nears as Google Accelerates AI Development
Google's Gemini AI model is nearing release after months of development.
The company's DeepMind chief, Koray Kavukcuoglu, confirmed that Gemini 4 has entered 'post-training' phase, where the model's behavior is refined for public use. This includes safety testing and implementing guardrails.
Kavukcuoglu stated that Google aims to release an early post-training output as soon as possible, with results showing promising progress. He expressed enthusiasm for continuing fast-paced iterations in AI development.
The news comes after rival companies Anthropic and OpenAI released major updates to their flagship models. Kavukcuoglu reassured the public that Google is committed to remaining at the forefront of AI innovation, citing the team's expertise and dedication.