ChatGPT from OpenAI has become the hype of AI, and has since made numerous companies to scramble for a way to monetize that hype.
Microsoft is one of the main backers of OpenAI, and with that, it has started promoting its Bing AI chatbot, after fixing the issues of it being "emotional" at times.
With the ability to understand and response in an uncanny way similar to humans, Microsoft researchers announced an experimental framework, where they embed the language abilities of ChatGPT, to control robots and drones.
What they did, was creating a way for AI to control mechanical movements.
Read: OpenAI Upgrades GPT-3, And Announces What It Calls The 'ChatGPT' AI
In a website post, the researchers at Microsoft detailed how they "extended the capabilities of ChatGPT to robotics, and controlled multiple platforms such as robot arms, drones, and home assistant robots intuitively with language."
To make this happen, ChatGPT takes its role as a medium, in which the researchers commands it using natural language, and have it write special codes to control the machines.
After that, the researchers view the results, and adjust it as necessary until a given task gets completed successfully.
In YouTube videos, shared by the channel Microsoft Autonomous Systems & Robotics Research, the researchers show how robots could be controlled by codes written by ChatGPT, while following human instructions.
In the experiment, the researchers showed how ChatGPT could control a robot arm to arrange blocks into a Microsoft logo, fly a drone to inspect the contents of a shelf, or finding objects using a robot with vision capabilities.
To get ChatGPT to control the mechanical movements of the hardware, the researchers taught ChatGPT a custom robotics API.
This way, when it's given instructions like "pick up the ball," ChatGPT can generate robotics control code just as it would write a poem or complete an essay.
The researchers who inspect the experiment, can make edits and tweaks to improve the accuracy and the safety of the overall experiment.
The researchers who execute the task can then evaluate its performance.
It's worth noting, that ChatGPT in controlling the mechanical movements isn't at all autonomous.
"We emphasize that the use of ChatGPT for robotics is not a fully automated process," reads the paper, "but rather acts as a tool to augment human capacity."
The research with the title "ChatGPT for Robotics: Design Principles and Model Abilities", is authored by Sai Vemprala, Rogerio Bonatti, Arthur Bucker, and Ashish Kapoor of the Microsoft Autonomous Systems and Robotics Group.
According to Microsoft, the challenges in robotics include how a human user can translate a task's requirement into code for a system to understand.
Traditionally, the engineer of the robot has to experience a loop, in which they have to write a new code and specifications to correct the robot’s behavior, and continue that process for unknown duration.
"Overall, this process is slow (user needs to write low-level code), expensive (requires highly skilled users with deep knowledge of robotics), and inefficient (requires multiple interactions to get things working properly)," said the researchers.
But using ChatGPT, the technology unlocks a new robotics paradigm, which allows a potentially non-technical user to sit on the loop, providing high-level feedback to the large language model (LLM) while monitoring the robot’s performance.
In the end, even without any fine-turning, those non-technical users could leverage LLM's knowledge to control the robot to do a variety of tasks.
Initially, results seem to be rudimentary. But regardless, they experiment represents early attempts at applying LLM into robotics.
This could open up robotics to a much wider audience in the future.
"Our goal with this research is to see if ChatGPT can think beyond text, and reason about the physical world to help with robotics tasks," reads a Microsoft Research blog post. "We want to help people interact with robots more easily, without needing to learn complex programming languages or details about robotic systems."