NVIDIA Releases Open AI Models, Google Invests $75M in A24

NVIDIA released a suite of open models and datasets on January 5, contributing tools for AI systems across industries from language to biomedical research. The company said the new offerings include models for speech, multimodal retrieval-augmented generation, and safety, part of its broader Nemotron family. Alongside these models, NVIDIA is providing one of the world's largest collections of open multimodal data, including 10 trillion language training tokens, 500,000 robotics trajectories, 500,000 protein structures, and 100 terabytes of vehicle sensor data. Leading firms like Bosch, ServiceNow, and Palantir are already integrating these tools into their workflows. ServiceNow is training its Apriel model family on open datasets, while CrowdStrike and Fortinet are using Nemotron Safety models to improve AI trustworthiness. The scale of this release is unprecedented, offering a unified resource for advancing AI in diverse domains. NVIDIA's move signals a growing trend toward open-source AI development, challenging proprietary models with community-driven alternatives. The company's recent Nemotron 3 family release laid the groundwork for this expansion, now covering agentic AI, physical AI, and autonomous vehicle development. Meanwhile, Google DeepMind announced a separate $75 million investment in independent film studio A24, aimed at shaping AI tools for creative industries. The partnership will let A24 filmmakers test and refine new workflows powered by DeepMind research, with Google gaining feedback from working artists. This is not a production or data-sharing deal, but a research collaboration focused on integrating AI into storytelling processes. A24, known for films like 'Backrooms' and 'Marty Supreme,' will work alongside DeepMind researchers to explore experimental techniques. The studio's input will help ensure the tools are designed for creators rather than imposed on them. This follows a broader trend of studios exploring AI integration, from Netflix's INKubator to OpenAI's backing of the animated film 'Critterz.' Google's investment values A24 at approximately $75 million, matching previous funding rounds and keeping the studio independent. The deal reflects a shift in how tech companies approach creative industries, prioritizing collaboration over acquisition. For practitioners, these developments highlight two paths: open models for broad accessibility and AI partnerships for specialized applications. Both approaches aim to reduce reliance on closed systems while maintaining performance. However, open models face challenges in data privacy and model alignment, while creative AI raises questions about authorship and artistic integrity. The success of these initiatives will depend on how well they balance innovation with ethical considerations. As AI becomes more embedded in workflows, the line between tool and collaborator blurs. Companies must navigate this terrain carefully, ensuring transparency and user control. For now, the focus remains on building tools that enhance rather than replace human creativity. The coming months will reveal whether these open models and creative partnerships can deliver on their promises without compromising the very qualities that make AI valuable.