IBM and NASA Collaborate to Develop Multimodal Foundation Models for Weather Prediction


In the continuous development of the AI field, people are beginning to worry about the impact of AI on Earth, particularly its high energy consumption and carbon emissions. The latest research shows that generative AI queries consume much more energy than traditional search engine requests, up to four to five times the energy of traditional operations. Especially, the energy required for image generation tasks is much higher than that of text-based operations. According to research by Morgan Sta
Weather forecasting and energy conservation are the roles of artificial intelligence in climate change. Companies like Google, NVIDIA, IBM, and Huawei are noteworthy in the field of climate change. Artificial intelligence can accelerate innovative solutions to address the impacts of climate change.
European AI company Mistral AI launches the full-stack production platform Mistral AI Studio, providing enterprises with secure, transparent, and scalable AI solutions. The platform integrates model deployment, monitoring, and optimization features, and is built on EU-based infrastructure to address data sovereignty and compliance challenges for multinational companies, enabling full-stack observability of AI decision-making.
NVIDIA released the OmniVinci all-modal understanding model, leading top models by 19.05 points in multiple benchmark tests. The model uses only 0.2 trillion training tokens, achieving six times the data efficiency of competitors. It aims to achieve unified understanding of vision, audio, and text, advancing machine multimodal cognitive capabilities.
OpenAI's newly launched Sora video generation platform is based on the Sora 2 model, capable of creating highly realistic videos, including fake images of celebrities such as Martin Luther King Jr., and using copyright characters like SpongeBob, generating shocking or harmful content. Users have some awareness of the generated videos.