José Daniel Salinardi for Poder & Dinero and FinGurú
The upcoming adoption of humanoid robots
The robotics industry is celebrating a new "iPhone moment," with a powerful era underway for manufacturing and domestic assistance. "When, literally, humanoid robots are wandering around, it's not within five years, but in a few," said Jensen Huang, CEO of Nvidia, this week at the GTC conference. The company, along with others, already has a lot at stake, with its RTX GPUs powering robots and the upcoming Isaac GR00T N1 from the firm serving as a generalist base model.
Next generation: As a staple of any past science fiction movie, humanoid robots are now approaching reality thanks to advancements in general artificial intelligence. It is part of what is called "embodied AI," which leverages AI technology within hardware; in this case, a robot that can perform real-world functions. Until recently, most robots relied on code to perform their tasks, which limited their applications. The promise of AI will allow them to learn new tasks through reasoning and imitation, or eventually identify and resolve situations like a human by generalizing common tasks.
"We know for sure that the world has a severe shortage of human workers," Huang continued. Humanoid robots never need to take a day off, can work 24/7, and can be used in multiple environments. This includes the workforce, such as in factory settings or on assembly lines, as well as at home, for tasks like vacuuming, folding laundry, washing dishes, and even providing companionship.
In the race: One of the most known players is Tesla's Optimus, which Elon Musk has called the "biggest product in history," with plans to produce up to a million units by 2027 (could it help revive the stock?). Another American player is Figure AI from Brett Adcock, backed by Microsoft and OpenAI. The company’s bots are designed for manual tasks such as logistics and storage, and have already been implemented in BMW factories. Other available bots include Boston Dynamics' Atlas and Sanctuary AI's Phoenix, as well as Chinese bots like Unitree's G1, Xiaomi's CyberOne, and UBTECH's Walker S1.
But Gemini Robotics is the latest announcement made, the most advanced vision and language model (You can watch a demonstration by following this YouTube link: https://www.youtube.com/watch?v=4MvGnmmP3c0)
The new Gemini Robotics model brings Gemini 2.0 into the physical world. It is a more advanced model of vision, language, and action, enabling interactive, dexterous, and versatile robots.
José Daniel Salinardi is a Certified Public Accountant graduated from the Faculty of Economic Sciences of the University of Buenos Aires.
He works as a content producer for the program Panorama 360 aired by UCL Television for Latin America, the United States, and Europe.
Previously, he held the same role for "Poder & Dinero," a radio program that aired on Radio Libre AM790 in Miami, and was also viewable through its streaming edition on the Americano Media channel.
"Poder & Dinero" continues to publish content through its social media as well as the weekly broadcasts of Panorama 360.
Comments