From creating images, generating text, and enabling self-driving cars, the potential uses of artificial intelligence (AI) are vast and transformative. However, all this capability comes at a very high energy cost. For instance, estimates indicate that training OPEN AI’s popular GPT-3 model consumed over 1,287 MWh, enough to supply an average U.S. household for 120 years.
Covariant this week announced the launch of RFM-1 (Robotics Foundation Model 1). Peter Chen, the co-founder and CEO of the UC Berkeley artificial intelligence spinout tells TechCrunch the platform, “is basically a large language model (LLM), but for robot language.”
RFM-1 is the result of, among other things, a massive trove of data collected from the deployment of Covariant’s Brain AI platform. With customer consent, the startup has been building the robot equivalent of an LLM database.
“The vision of RFM-1 is to power the billions of robots to come,” Chen says. “We at Covariant have already deployed lots of robots at warehouses with success. But that is not the limit of where we want to get to. We really want to power robots in manufacturing, food processing, recycling, agriculture, the service industry and even into people’s homes.”
I have been engaged in the telecommunication industry for over 25 years, and it is so refreshing to see the positive energy moving into this sector as a result of applying generative AI technologies.
The telecom industry can accelerate growth from generative AI. Here are use cases for the telco industry.
The new nonprofit Fairly Trained certifies that artificial intelligence models license copyrighted data—which often isn’t the case.
By Ben Guarino
A new computer model mimics Moon dust so well that it could lead to smoother and safer Lunar robot teleoperations. The tool, developed by researchers at the University of Bristol and based at the Bristol Robotics Laboratory, could be used to train astronauts ahead of Lunar missions. Working with their industry partner, Thales Alenia Space in the UK, who has specific interest in creating working robotic systems for space applications, the team investigated a virtual version of regolith, another name for Moon dust.
Lunar regolith is of particular interest for the upcoming Lunar exploration missions planned over the next decade. From it, scientists can potentially extract valuable resources such as oxygen, rocket fuel or construction materials, to support a long-term presence on the Moon. To collect regolith, remotely operated robots emerge as a practical choice due to their lower risks and costs compared to human spaceflight.
However, operating robots over these large distances introduces large delays into the system, which make them more difficult to control. Now that the team know this simulation behaves similarly to reality, they can use it to mirror operating a robot on the Moon. This approach allows operators to control the robot without delays, providing a smoother and more efficient experience.