Luma AI is expanding internationally by opening a new office in London to access top European talent and strengthen its presence in Europe and the Middle East, focusing on building multimodal artificial general intelligence (AGI) through video, audio, and language integration. With significant funding and plans for massive compute infrastructure, Luma aims to develop advanced AI models capable of understanding and interacting with the physical world, driving innovations in robotics and AI.
Luma AI is expanding internationally by opening a new office in London, which serves as a strategic gateway to Europe and the Middle East. The company has a strong pipeline of researchers and engineers from Europe, including talent from DeepMind, who are eager to join Luma. London was chosen as the location for their second office outside of Palo Alto due to its access to top talent and its business significance in the region. The launch of the London office marks a significant step in Luma’s growth and international presence.
Luma AI attracts exceptional talent from prestigious institutions like DeepMind, MIT, Stanford, and Berkeley by offering a unique environment where researchers can focus intensely on a single mission: building multimodal artificial general intelligence (AGI). Unlike larger labs with multiple projects, Luma’s relatively small size of about 150 people allows for abundant resources per individual and a highly aligned team dedicated solely to this ambitious goal. This focused mission appeals to those who want to contribute meaningfully to the development of advanced AI models.
The core of Luma’s technology revolves around video as a critical path to AGI. While language provides reasoning and abstraction, video captures the physical universe, showing how objects and natural phenomena behave. By combining video, audio, and language, Luma aims to create a universal simulator capable of understanding and predicting real-world scenarios. This approach not only has applications in entertainment and content creation but also opens the door to general-purpose robotics, enabling machines to simulate and reason about physical interactions in their environment.
One of the biggest challenges Luma faces is securing the right talent and computational resources to advance their research. Although the company has raised $900 million and secured promises of compute power from Saudi Arabia, the need for brilliant, aligned researchers remains critical. Luma is focused on hiring a relatively small but highly skilled team of around 200 to 300 people to tackle the complex problems involved in building multimodal AI systems. Alongside talent acquisition, the company is investing heavily in compute infrastructure, including a collaboration to build a two-gigawatt compute cluster, which is among the largest in the AI field.
Ultimately, Luma AI’s mission is to develop multimodal AGI that can generate, understand, and operate within the physical world. This requires overcoming significant research challenges and scaling compute resources economically. By combining cutting-edge research, top-tier talent, and massive computational power, Luma aims to create AI models that not only excel in video understanding but also pave the way for physical intelligence and robotics, marking a transformative step in the evolution of artificial intelligence.