Source: Visual China
By Lin Zhijia and Shaw Wan
BEIJING, May 11 (TiPost) — “Please write a 2023 company plan on behalf of the CEO.” “Why is it so hard to book a train ticket in China for the May Day holiday in China?” “Please write a program to crawl web content.” These were some of the questions that were frequently raised to ChatGPT, an artificial intelligence (AI) chatbot developed by U.S.-base OpenAI. Since its birth in November 2022, the generative AI technology has swept the world, starting an “arms race” among global tech giants such as Google, Amazon and Microsoft.
As a language model that can create humanlike conversations with users to answer their questions, ChatGPT is more than just a product. It will usher in a new era of artificial intelligence-generated content (AIGC), where AI is used to automate the information creation process while fulfilling the personalized requirements of users. On March 2, OpenAI released ChatGPT API, a tool that gives applications, products, or services access to ChatGPT’s ability to generate human-like responses to questions and engage in casual conversations, at the price of 0.002 dollars per 1k tokens. Since 1k tokens equal nearly 750 words, it means that generating 1 million words costs 2.7 dollars (18 yuan), ten times cheaper than GPT-3.5 models.
Since ChatGPT and AIGC can process different types of data, such as audio and video clips, at the application level, they will benefit the robot industry by bringing in humanoid robots that function as a nanny, a secretary and a best friend.
China’s RobotGPT
Dataa Robotics, a cloud intelligent robot operator, will launch a multi-modal cognitive model called RobotGPT, the first in the field of robot manufacturing.
With RobotGPT, HARIX operating system (OS) and Robot Development Kits (RDK), the company said it would be able to accelerate the development of robot skills and applications by applying large AI models to cloud robots. Since generative AI tools like ChatGPT have proven AI and conversational systems can play a bigger role in other areas, the company believed more authentic humanoid robots would no longer be a dream if they were trained on a massive corpus of data.
“The emergence of ChatGPT proves that robots need a cloud brain, because it is impossible to operate ChatGPT on a small computer. However, to have the unimodal technology useful for robots, we need to develop an AI model that accepts input in various forms, such as texts, visuals and motions, which is why we are working on RobotGPT,” said Bill Huang, the founder and Chief Executive Officer of Dataa Robotics.
To be more specific, RobotGPT has the main features and capabilities similar to ChatGPT in the field of intelligent dialogue of robots. It can also equip robots with multi-modal perception capabilities including vision, hearing, movement, and touch. Other capabilities include more powerful autonomous cognition, more complex logical reasoning and intelligent decision-making abilities, and multi-modal behavior expressions that support languages, vision and movement.
“Actually, it’s not that we started working on RobotGPT after the launch of ChatGPT. We started laying out the multi-modal large model of robots from the beginning of 2022. In products like robot navigation based on natural language instructions and intelligent decision-making, hand-eye coordination for robot grasping of disinfection robot Ginger and Cloudia, a human-computer interactive digital human,” Huang added.
Robots are about to enter the era of a “cloud brain.” Benefiting from digitalization, cloud computing, 5G network, AI technology and service robots, which assist human workers in completing tasks in a variety of service-focused industries, are being upgraded to be more intelligent and applicable. Starting from pilot programs, they will be popularized to reach a period of explosive growth. The cloud-backed intelligent robots based on the “cloud-network-end” architecture may become the future of the robot industry.
“If we persist in technological innovation and fully integrate 5G+cloud+AI, China will take the lead in the trillion-dollar service robot industry in the future,” Huang told TMTPost.
About eight years ago, Huang believed that the optimization of human intervention, which is one of the three core technologies of GPT, was very critical. That’s why the company puts efforts into developing the HARIX OS — the cloud brain and operating system for cloud robots.
Booming Robot Market in China
China’s robot market has surpassed that of Japan and the U.S., thanks to the rapid growth. In 2021, China’s robot shipments reached 268,200 units, ranking first in Asia with a year-on-year increase of 51 percent, according to the World Robotics Report 2022 by the International Federation of Robotics (IFR). The operating inventory exceeded 1 million units in 2021, up 27 percent compared from the previous year.
Due to massive investments in recent years, the density of industrial robots in the Chinese mainland has surpassed that of the U.S. for the first time, ranking among the top countries in the world. The report indicated that there were 322 industrial robots for every 10,000 workers in China in 2021, which was 2.28 times the world average. As the robot density reflects the level of manufacturing automation and the flexibility of production lines, it is a very important indicator of national strength.
Intelligent robots have been facilitating the high-quality development of China’s economy, especially the digital one. According to the data released by the National Development and Reform Commission (NDRC) in October 2022, the expanding scale of the core industries of the digital economy accelerated. The national software business revenue increased from 2.5 trillion yuan in 2012 to 9.6 trillion yuan in 2021, with an average annual growth rate of 16.1 percent. As of 2021, the scale of the core industry of China’s industrial Internet exceeded 1 trillion yuan. The scale of the big data industry reached 1.3 trillion yuan, making China one of the fastest growing cloud computing markets globally.
“It is estimated that by 2025, the integration of China’s digital technology and the real economy will achieve remarkable results, where an internationally competitive digital industry cluster will take shape, the digital economy governance system will be more complete, and China’s digital economy’s competitiveness and influence will steadily improve. As of 2035, China’s digital economy will enter a period of prosperity and maturity, where the development foundation and industrial system development level of the digital economy will rank among the top globally,” said NDRC chairperson He Lifeng.
The smart robots can be helpful in places like restaurants, airports, nursing homes and handicraft manufacturers. Take the nursing homes as an example. As China faces an aging society, more and more elderly people will need care services. 2025 is expected to witness more than 210 million persons aged 65 years or over, and that number will reach 310 million in a decade later. In the next 30 years, the global community will need 10 billion humanoid intelligent robots to provide services. And the global market for intelligent robots will reach over 30 trillion dollars.
Benefiting from digitalization, cloud computing, 5G network and AI technology innovation, cloud intelligent robots have become the first choice of customers in the field of robotics. They are an important digital economic infrastructure. As of the end of July 2022, there were more than 3,100 “5G and Industrial Internet” construction projects, forming a series of new scenarios, new models, and new formats.
“A robot is a third computer for humans. Humans have PCs and smartphones, and the next most important computing device is a robot,” said Huang.
Turning Point for Cloud Robots
In the future, the development of “a cloud brain plus a local robot” or “cloud service” type of robots for relevant application scenarios is expected to become one of the important models for the large-scale promotion and application of robots, according to the Top 10 Forefront Technologies of Robotics (2022-2023) Report released at the World Robot Conference 2022.
To design the development path of intelligent robots, the upgrading of technology, products and models should be taken into consideration. The hardware should be of high quality. The software should have enough experience with data accumulation and processing. And the intelligence level should be elevated to achieve intelligent data-driven robots.
That is the path that Dataa Robotics has been following, like Tesla. Back in 2015, the company took the lead in proposing the concept of “a cloud robot.” Through the upgrade of cloud intelligence and high-end manufacturing software and hardware, it invested in cloud services, robot design, manufacturing of key components, building smart factories, and research and development of technologies such as flexible intelligent joints. The company has now started commercialization in various fields, such as commercial retail, public health, communication, education, public safety, electric energy, agriculture and sanitation, with huge room for growth.
The company’s potential came at a fitting time, when China is stepping up efforts to develop domestic brands. In September 2022, the Ministry of Science and Technology approved and supported Dataa Robotics’ plan to construct “a national new generation artificial intelligence open innovation platform based on cloud robots.” In January, 17 national departments, including the Ministry of Industry and Information Technology, issued the “Robot +” Application Action Implementation Plan, proposing that by 2025, the density of manufacturing robots will be doubled compared with 2020. The plan also mentioned that the depth and breadth of the application of service robots and special robots should be significantly improved. Robots’ abilities to promote high-quality economic and social development should be significantly enhanced. “[It is important to] lay out future industries such as humanoid robots and metaverses, and to comprehensively promote the research and development of 6G technology,” said Jin Zhuanglong, the Minister of Industry and Information Technology.
更多精彩内容,关注钛媒体微信号(ID:taimeiti),或者下载钛媒体App