SAN FRANCISCO, CA – In a move signaling the continued exodus of top talent from Google’s DeepMind, Pete Florence, a former senior research scientist and key contributor to projects like RT-2 and PaLM-E, has launched his own startup, Generalist AI. The venture has already secured investment from Nvidia, underscoring the growing interest and financial backing for advancements in robotics.
Florence, who left DeepMind a year ago according to his LinkedIn profile, revealed his new role as co-founder and CEO of Generalist AI during a panel discussion at Nvidia’s recent GTC conference. While remaining tight-lipped about the specific details of his company’s operations, Florence hinted at a clear focus: robotics.
We’re basically still in stealth mode, Florence stated at the GTC discussion. However, when asked about the potential impact of his startup’s success, he articulated an ambitious vision: Our goal is to make robots that can do anything. Imagine a world where the marginal cost of physical labor approaches zero.
This bold statement encapsulates the core mission of Generalist AI: to make general-purpose robots a reality. The company aims to develop robots capable of performing a wide range of tasks, potentially revolutionizing industries reliant on manual labor.
The departure of Florence and the subsequent founding of Generalist AI highlight a significant trend within the AI landscape. As artificial intelligence continues to mature, leading researchers are increasingly venturing out to pursue their own visions and capitalize on the burgeoning opportunities in specialized fields like robotics.
Nvidia’s investment in Generalist AI further underscores the company’s strategic interest in the robotics sector. Riding the wave of the AI boom, Nvidia has not only achieved significant financial success but has also become a prominent venture capital player through NVentures. The exact amount of Nvidia’s investment remains undisclosed.
The development of general-purpose robots presents both immense potential and significant challenges. Creating robots capable of adapting to diverse environments and performing complex tasks requires advancements in areas such as computer vision, natural language processing, and reinforcement learning. Florence’s background at DeepMind, particularly his work on RT-2 (a vision-language-action model for robotic control) and PaLM-E (a multimodal model combining language and embodied perception), positions him well to tackle these challenges.
The success of Generalist AI could have far-reaching implications, potentially transforming industries ranging from manufacturing and logistics to healthcare and agriculture. As Florence envisions, a world with readily available, adaptable robots could dramatically reduce the cost of physical labor and unlock new possibilities for automation and efficiency.
While the specifics of Generalist AI’s technology remain under wraps, the company’s mission and the backing of Nvidia suggest a promising future for the development of general-purpose robots. The coming years will be crucial in determining whether Florence and his team can successfully navigate the complexities of robotics and bring their ambitious vision to fruition.
References:
- Machine Heart. (2024, March 20). 「让通用机器人成为现实」，谷歌RT-2、PaLM-E作者离职创业，已获英伟达投资 [“Making General-Purpose Robots a Reality”: Google RT-2, PaLM-E Authors Leave to Start a Business, Already Receiving Investment from Nvidia]. Retrieved from [Insert Original Article Link Here] (Note: Since the original article link was not provided, please replace this with the actual URL).
Views: 1