2020 has been a deeply difficult year for citizens, businesses and governments around the world. As covid-19 spread, requiring far-reaching health and safety restrictions, artificial intelligence (AI) applications have played a crucial role in saving lives and fostering economic resilience. Research and development (R&D) to improve basic capabilities of AI, from autonomous driving and natural language processing to quantum computing, has continued unabated.
Baidu was at the forefront of many important AI breakthroughs in 2020. This article describes five significant breakthroughs with implications for fighting covid-19 as well as transforming the future of our economies and society.
1. AI and vaccine development
The trend and why it matters. It usually takes years, if not decades, to develop a new vaccine. But in March 2020, vaccine candidates to fight covid-19 were already undergoing human testing, just three months after the first reported cases. The record speed of vaccine development was in part thanks to AI models that helped researchers analyze large amounts of data on the coronavirus.
There are tens of thousands of subcomponents to the external proteins of a virus. Machine learning models can sort through this stream of data and predict which subcomponents are the most immunogenic, that is, capable of producing an immune response, and thus guide researchers in designing targeted vaccines. The use of AI in vaccine development could revolutionize the way all vaccines will be created in the future.
Baidu innovations. In February, Baidu opened its LinearFold AI algorithm for scientific and medical teams working to fight the virus. LinearFold predicts the secondary structure of a virus’s ribonucleic acid (RNA) sequence – and does so much faster than traditional RNA folding algorithms. LinearFold was able to predict the secondary structure of the SARS-CoV-2 RNA sequence in just 27 seconds, 120 times faster than other methods. This is important because the key breakthrough for covid-19 vaccines has been the development of messenger RNA (mRNA) vaccines. Instead of conventional approaches, which insert a small part of a virus to trigger a human immune response, mRNA teaches cells how to make a protein that can trigger an immune response, which dramatically shortens the time frame involved in development and approval.
To support the development of mRNA vaccines, Baidu then developed and published an AI algorithm to optimize the design of mRNA sequences called LinearDesign, which aims to solve the problem of unstable and unproductive mRNA sequences in candidate vaccines. .
In addition to opening access to LinearFold and LinearDesign for researchers around the world, Baidu has also formed a strategic partnership with the National Institute for the Control and Prevention of Viral Diseases, which is part of the China Center for Control and disease prevention. Following an outbreak in Beijing’s Xinfadi market in June, Baidu’s artificial intelligence technology enabled authorities to complete genome sequencing of the coronavirus strain within 10 hours, helping to stem the epidemic. In December, Baidu unveiled PaddleHelix, a machine-learning-based bioinformatics framework aimed at facilitating the development of vaccine design, drug discovery, and precision medicine.
2. Fully automated driving and deployment of the robotaxis
The trend and why it matters. Autonomous driving technology continued to mature in 2020, with leading companies in the industry testing driverless cars and opening up robotaxi services to the public in various cities. Fully automated driving, which enables journeys without a human safety driver on board, will be necessary for the scalability and commercialization of autonomous driving.
Baidu innovations. In the past year, Baidu launched the Apollo Go Robotaxi service in the Chinese cities of Changsha, Cangzhou and Beijing, including high-traffic shopping areas, becoming the only company in China to start trial operations. of robotaxi in several cities.
These developments are the result of Baidu’s continuous innovation in the development of AI systems that can safely control a vehicle under complex road conditions and solve the majority of possible problems on the road, independent of a human driver. .
At Baidu World 2020, its annual technology conference, Baidu also demonstrated its capability for fully automated driving – where the AI system drives independently without an on-board safety driver. To support fully automated driving, Baidu has developed the 5G Remote Driving Service, a safety measure that allows remote human operators to take control of a vehicle in an exceptional emergency. Baidu’s success in fully automated driving and the deployment of its robotaxis suggests positive prospects for the technology’s commercialization in the near future.
3. Applied natural language processing
the trend – and why it matters. In 2020, natural language systems have become much more advanced in processing aspects of human language such as feeling and intention, generating language that aligns with human speech and writing patterns, and even visual comprehension, which means the ability to express understanding of an image through language. These natural language models generate more precise search results and more sophisticated chatbots and virtual assistants, leading to better user experiences and creating value for businesses.
Baidu innovations. Baidu has released a new multistream sequence framework for language generation called ERNIE-GEN. By training the model to predict semantically complete blocks of text, ERNIE-GEN performs at an elite level on a range of language generation tasks, including dialogue engagement, question generation, and abstractive synthesis.
Baidu’s ERNIE-ViL visual language model has also made significant strides in visual comprehension, ranking first in the VCR ranking, a 290,000-question data set built by the University of Washington and the Allen Institute. for AI, which aims to test the ability of visual comprehension. ERNIE-ViL also achieved peak performance on five downstream tasks of visual language. Visual comprehension lays the foundation for computer systems to interact physically in everyday scenes, as it involves both understanding visual content and its expression through language. It will be crucial to improve the quality of human-machine interaction.
4. Quantum computing
The trend and why it matters. Quantum computing made significant inroads in 2020, including the achievement of quantum supremacy by Jiuzhang’s computer. This is important for AI, as quantum computing has the potential to supercharge AI applications over traditional binary-based computers. For example, quantum computing could be used to run a generative machine learning model across a larger dataset than a conventional computer can process, thus making the model more accurate and useful in real contexts. Advanced technologies such as deep learning algorithms are also playing an increasingly critical role in the development of research in quantum computing.
Baidu innovations. Baidu achieved a number of technical breakthroughs in 2020 that promise to connect AI and quantum computing. In May, Baidu launched Paddle Quantum, a quantum machine learning development toolkit that can help scientists and developers quickly create and train models of quantum neural networks and deliver advanced quantum computing applications. The open source toolkit helps developers build quantum AI applications and helps deep learning enthusiasts develop quantum computing. In September, Baidu entered cloud-based quantum computing with the launch of Quantum Leaf, which provides quantum development kits such as QCompute, and can shorten the quantum programming lifecycle and help achieve a chain of quantum tools in “ closed loop ”.
5. AI chips
The trend and why it matters. AI hardware continued to expand in 2020, with the launch of several custom AI chips for specialized tasks. While an ordinary processor is capable of supporting AI tasks, AI-specific processors are modified with particular systems that can optimize performance for tasks such as deep learning. As AI applications become more widespread, any increase in performance or reduction in costs can generate more value for companies that operate a large network of data centers for commercial cloud services and can facilitate internal operations of the company. company.
Baidu innovations. At Baidu World 2020, the company offered a preview of its next-generation AI processor, the Kunlun 2, which it plans to put into mass production in early 2021. The chip uses 7-nanometer processing technology. (nm) and its maximum computing capacity is more than three times that of the previous generation Kunlun 1. Kunlun chips are characterized by high performance, low cost and high flexibility, which can support a wide range AI applications and scenarios, helping drive greater adoption of AI and lower user costs. More than 20,000 Kunlun 1 chips have now been deployed to support Baidu’s search engine and Baidu Cloud partners since their launch in 2018, strengthening industrial manufacturing, smart cities, intelligent transportation and other areas.
This content was produced by Baidu. It was not written by the editorial staff of MIT Technology Review.