General Intuition, valued at $2.3 billion, has raised $320 million to explore using gaming data for training AI models. This approach aims to bridge the gap in understanding physical realities that current language models struggle with, potentially advancing artificial general intelligence (AGI).
General Intuition has recently completed a funding round that raised $320 million, elevating the company's valuation to $2.3 billion. Among the investors are Coatue, Eric Schmidt, and leading researchers from MIT and Google DeepMind, indicating strong interest in the startup's approach.
The CEO, Pim de Witte, articulated that traditional large language models like ChatGPT and Claude lack the necessary skills to understand physical dynamics, which is crucial for achieving artificial general intelligence (AGI). By leveraging gaming data, the company aims to train models that can better grasp how objects interact and move in space and time.
General Intuition spun out of the gaming platform Medal TV, which showcases a focus on merging entertainment with advanced AI technologies. This background provides the startup with unique insights into creating world models that could outperform conventional methods.
The conversation also delves into the ethical implications of using gaming data for AI, especially concerning potential defense applications. De Witte emphasizes the importance of defining ethical boundaries in the development and deployment of these advanced models.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
General Intuition, valued at $2.3 billion, has raised $320 million to explore using gaming data for training AI models. This approach aims to bridge the gap in understanding physical realities that current language models struggle with, potentially advancing artificial general intelligence (AGI).