Generalist AI Inc., a startup that develops artificial intelligence software for robots, has reportedly raised $200 million in funding.
Axios today cited a source as saying that 8VC led the investment. It was reportedly joined by a number of unnamed existing investors. Generalist’s previous $400 million round in June included the participation of Nvidia Corp., Bezos Expeditions and more than a half-dozen others.
The company’s latest raise comes about a week after it debuted its newest AI model. Gen-1.5, as the algorithm is called, is designed to power robotic arms. Generalist says it can significantly reduce the amount of time needed to create factory automation workflows.
Historically, developers had to program robotic arms manually for each task they sought to automate. Moreover, robot code had to be updated in response to even minor operational changes. For example, a machine configured to place merchandise into boxes might need an update if its operator switches to larger boxes.
Some robots ship with AI models that reduce the need to write custom code. However, teaching an AI-powered robot a new task often requires developers to fine-tune its neural network, which can be time-consuming. Generalist says Gen-1.5 offers a more convenient alternative.
The model enables users to teach robots a new task by demonstrating how it’s performed with their hands. According to Generalist, Gen-1.5 can record the demonstration with the robots’ built-in cameras or sensors attached to the user’s hands. The model can also learn from clips of simulated robots.
The company tested Gen-1.5’s capabilities across 10 sample tasks, and said it achieved an average task completion rate of 59% when it was given a single example of how to perform them. That percentage rose to 83% when users provided a few additional examples.
Generalist says Gen-1.5 is the first AI model with the ability to learn a wide range of robotics tasks based on a single or a few examples. Furthermore, the model can autonomously refine its workflows. During Generalist’s testing, Gen-1.5 decided to complete some tasks with a different tool than the one it was instructed to use.
Today’s report didn’t specify how the company plans to use its new funding. However, it’s safe to assume that a sizable portion of the capital will go towards AI infrastructure: Generalist says Gen-1.5 took more than eight months to train.
The participation of Nvidia in the company’s June funding round hints that it will use the chipmaker’s graphics cards. Nvidia develops not only data center accelerators but also a line of AI chips called Jetson specifically optimized for robots.