ChatGPT, released by OpenAI, has managed to elevate competition in the field of artificial intelligence to a higher level. Google announced that Robotics Transformer 2 (RT-2) was trained using a vision-language-action (VLA) model, which is the only one of its kind. RT-2, trained with a model similar to Google Bard, will be able to perceive what is happening around it.
Google introduces Google Robotics Transformer 2 based on VLA
In a blog post on Friday, the company stated, “By utilizing technologies such as artificial intelligence chatbots like Bard, ChatGPT, and Claude 2, we can enable robots to perform tasks quickly and more efficiently.”
Vincent Vanhoucke, President of Robotics at Google DeepMind, said, “At the core of RT-2, there is a technology similar to the large language models behind platforms such as Google Bard. We process text and image data available on the internet to obtain direct outputs of robot actions.”
Meta's Llama 2 AI, available on Poe, offers a ChatGPT-like experience with Llama-2-70b as the recommended model, trained on 70 billion data.
Vanhoucke stated that enabling robots to use artificial intelligence to understand the world around them is more challenging than for chatbots. While AI chat software needs to digest a set of textual data about a specific topic and organize this information in a way that can be understood by humans, robots need to literally understand the world around them.
Google claims that with RT-2, which utilizes online image data, robots can be quickly trained to understand what trash is and how to pick it up and dispose of it. According to Grand View Research, the industrial robotics industry is currently valued at $30 billion and is expected to reach $60 billion by 2030.
