cool hit counter Experts talk about the new path of artificial intelligence from research to industry_Intefrankly

Experts talk about the new path of artificial intelligence from research to industry


On the morning of October 23, the 2018 Artificial Intelligence Summit was held in Suzhou Industrial Park. Hosted by Soochow University, organized by the Institute of Artificial Intelligence of Soochow University and supported by Jiangsu Artificial Intelligence Industry Technology Innovation Strategy Alliance, the conference is themed "From Research to Industry", gathering influential scientists and entrepreneurs in the field of AI worldwide to deliver speeches and conduct high-end dialogues on the frontiers of technology, industrial trends and related hot issues in AI. Wu Qingwen, member of the Standing Committee of the Municipal Party Committee and Secretary of the Party Working Committee of the Park, Lu Chunyun, Vice Mayor, Lin Xiaoming, member of the Party Working Committee of the Park and Deputy Director of the Management Committee, attended the event.

Established in November 2017, the Institute of Artificial Intelligence at Soochow University has brought in more than a dozen domestic and foreign experts and enrolled more than a hundred graduate students over the past year, focusing on research in natural language processing, health care, big data and other fields, and actively cooperating with local enterprises in industrialization. During the event, the Academic Committee of the Institute of Artificial Intelligence of Soochow University was inaugurated. "We hope that through our efforts, we can accelerate the process of AI from research to industry, so that the results of the institutions can really help local economic development." So said Ling Xiaofeng, director of the Institute of Artificial Intelligence at Soochow University.

Let the "city brain" really like the human eye and brain division of labor cooperation

The application of ultra-large scale artificial vision systems, referred to as "city brains", is already being used in many places. Gao Wen, a professor at Peking University and a member of the Chinese Academy of Engineering, believes that its development is still in its infancy and should rely on "brain-like visual computing" for the next step of deepening. "If you compare the camera to the human eye, the cloud platform can be seen as the human brain, and in the division of labor between the human eye and the human brain, not everything the eye sees is sent to the brain." In Gowan's view, the mechanism of human eye-brain cooperation is worth learning from AI systems.

"Most current cameras only do information recording and recognition; in fact it should also have enough information processing power to filter out what the cloud platform really needs and transmit it." Gowen calls it a digital retina, "The digital retina should include cameras, cloud platforms and connections between cameras and cloud platforms to truly enable real-time aggregation of features and on-demand video retrieval, optimizing feedback between cameras and the city brain. "

Talking about the "landing" of artificial intelligence from research to industry, Gowan said it mainly depends on soil, water and fertilizer. Soil is the environment for industry development, and water and fertilizer are the support of local governments. "I've been to the park many times and it has a good industrial base and government services, which is very suitable for the growth of AI innovation seedlings."

Artificial intelligence security is as important as social security

"Now that the internet, as a virtual society, has completely overlapped with the physical society, and almost everything in the physical society has been 'moved' online, from the economy to education to healthcare, covering all aspects of life, cyber security is as important as social security." On site, Fang Binxing, President of the China Cyberspace Security Association and Academician of the Chinese Academy of Engineering, discussed the three spaces of AI security from a risk perspective. He argued that there are two types of relationship between any new technology and security: enabling security, which refers to the use of technology for security and is equivalent to adding security to enhance security; and companion security, which refers to the presence of technology before security is considered. "There are safety risks if the technology is not mature, as if a child, if he is weak then surely he is often sick, but there is another kind of naughty child who is not weak himself but often goes to harass and influence other children and make them sick, again there are safety issues."

In Fang Binxing's view, AI security includes four levels of ICT chips, code, data, and applications, and should deal with both attacks on AI technology systems and the data they involve themselves, as well as security issues that endanger the external world, such as humans, by the application of AI systems. "If a self-driving car, for example, puts two small ads on a 'stop' traffic sign, it will still look like a stop to the human eye, but the machine could misidentify it, and the future of AI safety has a lot of way to go."

Editor Tang Xiaowen Photography Zhang Peiwen

Graphic selected from the 100


Recommended>>
1、LruCache源码解析
2、Encrypting information about database connections in spring and springboot Preface i Encrypting information about database connections in spring ii Encrypting information about database connections in springboot projects summary
3、手游客户端性能分析工具Cube全新改版为你的手游深度体检
4、This college girl designed a plugin board that puts all the big name manufacturers to shame
5、Twitter的野心我们终将弃用密码

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送

    已发送

    朋友将在看一看看到

    确定
    分享你的想法...
    取消

    分享想法到看一看

    确定
    最多200字,当前共

    发送中

    网络异常,请稍后重试

    微信扫一扫
    关注该公众号