A Google robot that understands us when we speak naturally

The alphabet is the umbrella under which all divisions of Google are organized. The reason they decided to reorganize as a holding company is because it gives their operations some independence and also makes them more visible to investors.

Well, since 2019 this company very focused on robot development so that they can perform daily tasks.

Well, now they have set about achieving a rather simple goal a priori, but in fact a difficult one: help humans communicate better with robots through voice or text and allow them to perform complex tasks with a better understanding of the language.

This project we’re talking about today, which is owned by Everyday Robots (along with Google Research), is still in its infancy, but now the robots have received an update: Better language understanding thanks to the large Google PaLM Language Model (LLM).

Everyday Robots will fit one of its robotic assistants and Google into the language model, creating PaLm-SayCan.. “This is the first implementation to use a large-scale language model to plan a real robot.” The new project should help people communicate better with robots.” Google explains.

Google claims that thanks to this combination, the robots were able to generate correct answers to 101 instructions that were given in 84% of cases and successfully completed in 74%.

On the other hand, according to Android Central, the company influences the safety of its robots with PaLm-SayCan. The algorithm is limited to orders that take into account the safety of the robot and also save things “very interpretable”It is affirmed here.

Although this is a success and it is indisputable, one must proceed with caution, why? because real life is full of many orders and different formulas. We are facing a complete mess that these robots must face.

Google and Everyday Robots hope that the PaLm-SayCan algorithm will finally help robots achieve more natural interactions with humans.