To explore the implementation of AI in thousands of industries, the AISummit 'AI Empowering Industrial Practice' sub-forum was successfully held

On August 6th and 7th, 2022, ​​AISummit Global Artificial Intelligence Technology Conference​​ will be held as scheduled. The theme of this conference is "Drive·Innovation·Digital Intelligence", and the content covers computer vision, natural language processing, speech recognition, algorithms and models, recommendation systems, machine learning, intelligent driving, smart finance, metaverse, MLOps and many other technologies. The segmented fields bring a gluttonous technology feast to mid-to-high-end technology managers and technology practitioners of technology companies, business managers who plan/are undergoing digital transformation, as well as people and entrepreneurs interested in the field of artificial intelligence. .

On the afternoon of the 7th, at the sub-forum with the theme of "AI Empowering Industry Practice", Jiang Wei, the head of the risk intelligent high-availability algorithm of the Ant Group Technology Risk Department, and Baidu Senior R&D Engineer/AI Drug Discovery Technology Five big names, Fang Xiaomin, the person in charge, Ma Guoning, deputy general manager of Taifan Technology, Zhang Bo, CTO of Cloud Wisdom, and Chen Guanling, technical partner of Fuyou Trucks, brought wonderful theme sharing around the application practice of AI in different fields, and provided insights into the application of AI in different fields. The implementation of thousands of industries provides a powerful reference.

Ant Green Intelligent Capacity Technology Practice

In cloud-native large-scale online microservice systems, failures mainly come from changes and capacity. Once a failure occurs, it may cause service interruptions and production accidents, resulting in huge economic losses and concentrated customer complaints. How to apply algorithm models to build change risk identification and automatic capacity assessment, improve system reliability, and ensure high availability? In the theme sharing of "Ant Green Intelligent Capacity Technology Practice" brought by Jiang Wei, head of the Risk Intelligent High Availability Algorithm of Ant Group's Technology Risk Department, he explained it in detail.

Jiang Wei said that a highly available system must have three main elements: less faults, fast recovery, and low costs. Ant Group's main faults come from change and capacity, accounting for more than 50% of the total. To this end, Ant Group has used its algorithm capabilities to build change risk identification capabilities in change scenarios and automated capacity assessment capabilities in capacity scenarios.

In the following time, Jiang Wei shared in detail the main technologies used by Ant Group in change scenarios, as well as some practices in green intelligent capacity technology. Jiang Wei said that through various technical optimizations, Ant Group’s risk identification and capacity assessment have achieved significant results in system reliability and guarantee.

Jiang Wei emphasized that data has data boundaries, and algorithms have algorithm boundaries. However, only by truly understanding business, data, and engineering can algorithms be implemented faster and better in business scenarios, and data Give full play to its value better and let technology truly create greater value for the enterprise.

Baidu Biocomputing Large Model's Drug Research and Development Approach

In recent years, "AI medical care" has developed rapidly. With its intelligent and automated characteristics, it is mainly used in Public health, medical imaging, medical robots, drug research and development, etc. Although "AI medical care" is still in its early stages, with relatively low commercial application and low overall market penetration, "AI medical care" has a very broad space for development.

Fang Xiaomin, senior R&D engineer at Baidu and head of AI drug discovery technology, pointed out in the sharing titled "Drug Research and Development of Baidu Biocomputing Large Model" that the current main focus of AI drug research and development is on the drug design and discovery stage. , focusing on using machine learning models to solve time-consuming and costly simulated chemical or biological experiments in drug design and discovery. Fang Xiaomin said that the main challenge faced in using AI for drug research and development is that there is very little annotated data in the biological field and the acquisition cost is very high.

In order to better apply AI technology in the field of biomedicine, Baidu launched PaddleHelix. PaddleHelix is ​​an AI-driven comprehensive biocomputing open source tool library. The bottom layer relies on PaddlePaddle's core framework, and includes two layers of open source tools and platform services. In the following time, Fang Xiaomin introduced in detail the main technical advantages of the propeller PaddleHelix.

Fang Xiaomin said that PaddleHelix hopes to use all kinds of data we can obtain as much as possible, such as unlabeled data. He emphasized that in the biological field, there is a lot of unlabeled data. Using PaddlePaddle, we can collect about 1B unlabeled data of compounds and 2B unlabeled data of proteins. According to reports, PaddlePaddle can complete compound modeling and protein modeling and folding, and has achieved remarkable results.

Picture everything, from Königsberg to empowering all industries

In the process of industrial empowerment, when AI encounters bottlenecks and its own capabilities are insufficient, who will come? Empowering AI? Using knowledge graphs is the best way.

Ma Guoning, deputy general manager of Taifan Technology, shared the theme of "Mapping Everything, from Königsberg to Empowering All Industries", starting from the knowledge graph, an important technology in the field of cognitive intelligence. The famous Seven Bridges of Königsberg problem proposed by Euler extends to how to use cutting-edge theoretical technologies in graph theory and other fields to solve the problems of complex entities, difficult retrieval, and excessive update overhead faced in the practical application of knowledge graphs.

Ma Guoning emphasized that it has become a consensus to solve different industry problems in a low-cost and high-efficiency way by creating an effective platform tool. In the following sharing, Ma Guoning combined a large number of practical cases in detail to demonstrate the technical practice of using the knowledge graph platform to empower different industries.

Ma Guoning said that we are committed to providing a tool for the industry to make cutting-edge and difficult-to-understand technology applications fool-proof, providing multiple possibilities for AI and industry empowerment, and allowing the artificial intelligence industry to have a hundred flowers bloom. s future.

From the laboratory to the user's desktop, the road to AI implementation

In recent years, AI has been widely used in various industries, promoting the intelligence of various industries , which has greatly improved the management level and decision-making level, including the IT industry. Applying AI to IT operations, also known as AIOps, is a hot spot for AI application in the IT industry. Therefore, how to operate and maintain efficiently has become a problem that IT departments and even CIOs must face.

Zhang Bo, CTO of Cloud Intelligence, pointed out in his sharing titled "From Laboratory to User Desktop, the Road to Implementation of AI" that adding Algorithm algorithms to data such as indicators, logs and call chains is the scene of AIOps . In the following sharing, Zhang Bo shared about intelligent operation and maintenance in the AI ​​2B industry, explaining how AI algorithms are adapted and implemented in the industry, and how AI engineering is adapted and implemented in the industry. He also shared Practical cases of enterprise development technology in the industry.

Zhang Bo said that AI to B is a particularly interesting industry that requires both Algorithm capabilities and implementation, and success or failure is judged by results. There is a sea of ​​​​stars with the entire Algorithm and the entire algorithm, and everyone needs to explore it together. Deep learning, machine learning and other technologies can truly empower industrial change.

Technical application of autonomous driving in trunk logistics

Autonomous driving is one of the most typical application scenarios of artificial intelligence. For logistics companies, in addition to safety, the core motivation for applying autonomous driving is to reduce costs.

Chen Guanling, technical partner of Fuyou Trucks, pointed out in his sharing titled "Technical Application of Autonomous Driving in Trunk Logistics" that road freight has long had many pain points. One of them is the long sensing distance. For heavy trucks, longer sensing distance means longer braking distance. Second, it is difficult to change lanes. It takes about 10 seconds for a truck to complete a lane change in a high-speed scenario. If the driver's advance observation is included, it may take longer, and the risk to the safe driving of surrounding vehicles will be greater. .

In the following time, Chen Guanling shared in detail the open source commercial operation scenarios of autonomous driving companies, and comprehensively analyzed the integrated development of AI and logistics from three perspectives: technology, implementation, and practice. In order to promote the advancement of autonomous driving technology, Fuyou Truck has launched the "Venus" plan to open source autonomous driving companies and open up Fuyou's commercial operation scenarios.

Chen Guanling said that our vision is to move from the current dispatching of human-driven trucks, to the future dispatching of intelligent vehicles that combine humans and machines, to the future dispatching of completely driverless trucks, to create a truly cross-city dispatch. Intelligent operation platform for trunk logistics.

Written at the end: With the optimization of computer vision, speech recognition, machine learning, algorithms, models and other technologies, as well as the continuous improvement of the industrial structure, artificial intelligence has richer application scenarios , such as applications in risk control assessment, engineering operation and maintenance, biopharmaceuticals, logistics and freight, etc., while accelerating the structural upgrade of the AI ​​industry. Through the convening of this event, senior experts in the field of artificial intelligence in five major industries shared their wonderful practices, which provided a powerful reference for the application of artificial intelligence in different fields and further drove the implementation of artificial intelligence technology in thousands of industries.

To watch the video replay, please go to the official website of the AISummit conference:​​aisummit.51cto.com​​

Follow [51CTO Technology Stack] official account, reply to [AI Conference] to receive the conference PPT

The above is the detailed content of To explore the implementation of AI in thousands of industries, the AISummit 'AI Empowering Industrial Practice' sub-forum was successfully held. For more information, please follow other related articles on the PHP Chinese website!