On December 28, 2021, Alibaba Group's global research institute, Alibaba DAMO Academy, announced the "DAMO Academy Technology Trend Forecast 2022," which forecasts the changes that will be brought to the technology industry in the future.
DAMO Academy analyzes millions of papers and patent applications published in the last three years, and interviews about 100 scientists, mainly for the economy and society as a whole within the next five years. It introduces the top 10 technology trends that are expected to accelerate breakthroughs and impacts in various fields.
Jeff Zhang, head of Alibaba DAMO Academy, said:
"Over the past century, the evolution of digital science has pushed the boundaries of technological advancement and industrial development and has brought great benefits to society.
We are currently in the midst of a rapidly evolving digital environment. The physical world we live in is becoming more digital and networked, and more and more integrated with the intelligent online world.
We are about to move into a world where virtual reality technology (VR) reconstructs human-machine interactions and cutting-edge technologies such as AI drive industrial transformation such as medical research, weather forecasting, and manufacturing. is.
We are also actively working on the introduction of digital technology to promote a low-carbon economy. Alibaba DAMO Academy believes that good technology should be developed for a better future, and this belief will not change in the future. "
With the rapid development of new network technologies, cloud computing will evolve into a next-generation computing system that integrates clouds, networks, and devices.
The new system will make the division of roles between cloud, network and device clearer. The fusion of cloud, network, and devices is driving the emergence of new applications that enable more demanding tasks such as industrial precision simulation, real-time quality inspection, and mixed reality (MR). Over the next two years, the number of applications running on new computing systems is expected to surge.
In the last few hundred years, the scientific community has had two basic frameworks: experimental science and theoretical science. Currently, advances in AI technology are realizing a new scientific paradigm.
Since machine learning can process huge amounts of multidimensional data and solve complex scientific problems, scientific exploration in fields that were previously thought to be impossible will flourish.
AI technology not only accelerates the speed of scientific research, but also contributes to the discovery of new scientific laws. Over the next three years, AI is expected to be widely applied in the research process of applied science and to be used as a production tool in some basic sciences.
As the size of transistors approaches their physical limits, the development of electronic chips is becoming unable to meet the increasing demands for data throughput due to the rise of high performance computing.
Unlike electronic chips, silicon photonics chips use photons instead of electrons to transfer data. Photons do not interact directly with each other and can travel long distances, and silicon photonics chips that utilize photons can provide higher computational power and energy efficiency.
With the rise of cloud computing and AI, silicon photonics technology is rapidly developing. Silicon photonics chips are expected to become widespread in high-speed data transmission in large-scale data centers over the next three years.
In recent years, technologies such as wind power generation and solar power generation have been rapidly developed, and these renewable energies have become effective energy sources. However, problems such as difficulty in integrating power grids, low energy utilization, and storage of surplus energy are major obstacles to effective operation.
Since the amount of renewable energy generated is unpredictable, integrating renewable energy sources into the grid poses a challenge that impacts the safety and reliability of the grid.
The application of AI in the energy industry is crucial for improving the efficiency and automation of power systems, maximizing the use and stability of resources, and also contributing to the achievement of carbon neutrality.
Over the next three years, AI technology is expected to pave the way for the integration of renewable energy sources into the grid and contribute to the safe, efficient and reliable operation of the grid.
Medical care is an area that relies heavily on the expertise of healthcare professionals, and the effectiveness of treatment may vary from patient to patient. The fusion of AI technology and Precision Medicine is expected to support the integration of specialized knowledge and new assistive diagnostic techniques and provide a reliable guideline in clinical medicine.
Through high-precision medical care, doctors can diagnose illnesses as quickly and accurately as possible, make medical decisions, and quantify, calculate, predict, and prevent serious illnesses.
Over the next three years, human-centered precision medical technology is expected to become a major trend across multiple areas of health care, including disease prevention, diagnosis and treatment. AI will be a highly accurate compass that can pinpoint illnesses and their treatments.
For a long time, the application of privacy-protected computing has been limited to a narrow range of small-scale computations due to performance bottlenecks, unreliability in technology, and standardization issues.
However, with the emergence of integrated technologies such as dedicated chips, cryptographic algorithms, white box implementations, and data trusts, we go one step further from processing small amounts of data and data in private areas to process large amounts of data. Privacy protection computing technology will be adopted in situations such as integrating data in all areas.
As a result, it can be expected that productivity will be improved by utilizing data from all areas. Over the next three years, the performance of privacy-protected computing will improve dramatically, and companies will emerge to provide reliable services in data distribution.
With the development of technologies such as cloud edge computing, network communication, and digital twins, the XR field has blossomed in earnest.
Use XR glasses to deliver an immersive internet experience brought about by mixed reality. XR will be a technology rooted in a new industrial ecosystem that includes electronic components, devices, operating systems and applications.
Rebuild digital applications and revolutionize the way people interact with technology in areas such as entertainment, social networking, offices, shopping, education, and healthcare.
Over the next three years, new-generation XR eyeglass products that are indistinguishable from ordinary eyeglasses are expected to enter the market and become an important gateway to the next-generation Internet.
Perceptive soft robotics refers to a robot that has a physically flexible body and has enhanced perception of pressure, vision, and sound, unlike conventional robots.
These robots utilize state-of-the-art technologies such as flexible electronics, pressure-sensitive materials, and AI to perform highly specialized and complex tasks and transform to adapt to different physical environments. ..
The emergence of perceptual soft robotics is expected to shift the flow of the manufacturing industry from mass production of standardized products to specialized low-volume production.
Over the next five years, perceptive soft robotics is expected to replace traditional robots in the manufacturing industry and pave the way for the spread of service robots in everyday life.
Ground networks and computer systems provide digital services in densely populated areas, but not well in less densely populated areas such as deserts, seas, and space.
Satellite-terrestrial integrated computing technology connects artificial satellites with terrestrial mobile communication networks to provide seamless, multidimensional coverage, as well as satellite networks, terrestrial communication systems, and cloud computing. Build a system that integrates computing technology.
This makes digital services more accessible and comprehensive. Over the next five years, artificial satellites and terrestrial systems will act as computing nodes, creating an integrated network system that provides high connectivity.
Future AI technologies are expected to move to co-evolution of large and small models via the cloud, edge and devices.
A large-scale pre-learning model, also called a basic model, is a basic technology from weak AI to general AI, and relatively improves the performance of various applications using deep learning.
However, there is a limit to the development of large-scale models because the benefits of improved performance and power consumption are not balanced.
Future AI technologies are expected to be more practically useful as they move from competition for scalability of underlying models to co-evolution of large and small models via the cloud, edge and devices.
Related information: https://damo.alibaba.com/techtrends/2022
In the latest issue of "DIME", in addition to the introduction of this year's award-winning lineup, the secrets that led to the hit, behind the scenes, and everything that seems to be popular in 2022 are posted. The special appendix is an oversized 10-inch electronic memo pad that can be used at work or at home. Although it is an oversized 10-inch size that allows you to write plenty of pictures and letters, it is 6 mm thick and weighs 143 g, making it thin, lightweight, and easy to carry. Moreover, it can be rewritten about 100,000 times. There are various ways to use it, such as hooking it on a hook, hanging it with a string, or leaning it against a special pen.
* Major online bookstores have been sold out. Please purchase at a bookstore or convenience store near you.
Configuration / DIME editorial department