SaMedia (Dec. 4)–Google has unveiled two new generative AI models, Veo and Imagen 3. The tools aim to transform video and image creation. Both models are now available on the Vertex AI platform.
Veo, in private preview, is Google’s most advanced video generation model. It can create high-definition videos from text or image prompts.
Using its understanding of visual semantics and language, Veo produces cinematic-quality footage with smooth motion and coherence.
The model is suited for industries like marketing and content production. It allows for rapid prototyping and lowers costs. Early adopters, such as Agoda, report major efficiency gains in video ad production with Veo.
Google also introduced GenCast, an AI-powered weather forecasting model. The tool delivers 15-day forecasts faster and more accurately than other systems, including the European Centre for Medium-Range Weather Forecasts (ECMWF).
In a study published in Nature, GenCast outperformed ECMWF’s model 97.2% of the time. For lead times longer than 36 hours, its accuracy increased to 99.8%.
A blog post by Google DeepMind explained how GenCast handles uncertainty. “An ensemble forecast expresses uncertainty by making multiple predictions for different scenarios,” the post said. “If most predictions show a cyclone hitting the same area, uncertainty is low. But if they predict different locations, uncertainty is higher. GenCast strikes the right balance, avoiding overstating or understating its confidence.”
