It is worth noting that the three founders of Kuang Shi Technology, Qi Qi, Tang Wenbin and Yang Mu, both came from Yao Shizhi Computer Science Experimental Class of Tsinghua University (referred to as 'Yao Class'). The seminar was called 'Going Back to School' by three people.
As the world's founder of Modern Cryptography, Theoretical Computer Science, Quantum Computing and the only Chinese winner of the Turing Prize, 'Yao Class' founded by Yao Chi-Chi has produced a large number of top talents in computer science in China and its students have long been throughout the country Outside the AI industry and computer science research in all key areas.
Tang Wenbin, the founder and CTO of Kuang Shi Science and Technology, said: 'We have always been thinking about one thing that we can do in business, including technology, products and industry. How can we speed up the development of this industry? So far there is no particularly good answer, so taking advantage of today 's opportunity we returned to his alma mater and returned to Yao to ask Mr. Yao for the most essential question about artificial intelligence.
Yao Chi: AI is currently the best time for a new generation of talent to become the next opportunity for Newton and Einstein
For the essence of artificial intelligence and when is the theory of artificial intelligence born? Yao believes that "the development of artificial intelligence is not separated from the theoretical foundation of the present and the past, but as technology is further explored, we also know that deep learning may not be able to Become the ultimate theory to solve the problem. "In today's practice of artificial intelligence, the next innovation in basic theory where? In this regard, Academician Yao Chi gave his own view: 'deep learning is indeed to artificial intelligence to the entire human race, It opened the window to the question of what is meant by "smart." Although it was not yet possible for us to solve the problem altogether, it gave us some hints that this was the point that made people excited and made our research even bigger However, from the basic theory, how to find the mystery of artificial intelligence in an interdisciplinary way is the most challenging issue for us nowadays. "Yao Chih said," The exploration of artificial intelligence requires that you not only be a computer Scientist, should be a scientist.
According to Yao Chi, this is the best time of artificial intelligence, and the current artificial intelligence research is in the era where researchers were dreamed, because this period gave a new generation of talents the opportunity to be the next Newton and Einstein. At the same time, Academician Yao also Two elements of the talent needed in the AI industry are given, namely, cleverness and enough hands-on experience. "Cleverness is an important trait for AI talent because if you are using artificial intelligence to solve a problem, You can not learn a lot of things, if you are smart enough, after two years of training, you may become a very top expert in machine learning field and solve very sophisticated problems. "Academician Yao Chi also stressed that the blind pursuit of the accumulation of academic papers Without paying attention to practice, will be a problem .Yao Academician said when doing metaphor, 'After you practice Wudang Qigong, your sword law is good, but you do not have actual combat experience, then you must be a problem, It is very important to go out to master and may be solved at once. Therefore, this actual combat experience is very important.
Sun Jian: The essence of algorithmic innovation is practical
In view of the algorithmic innovation, Sun Jian, chief scientist and dean of Academia Science and Technology, pointed out: 'Artificial Intelligence is an interdisciplinary research.We must clarify three important aspects of the algorithm before we research how the algorithm innovates: To solve the algorithm to solve what kind of problems have a clear and profound understanding, followed by your algorithm in a particular period of time for solving the problem is generic, that is more versatile, the greater the impact of the algorithm; Finally, there is no peer Or whether the industry is using it, whether it solves the pain points and touches on the nature of the issue. "Sun Jian added: 'What we want to do is to be able to publish it to the industry and take the initiative to use the algorithm that we even have to use so that we can Known as the nature of the algorithm. '
As ResNet was developed by Dr. Sun Jian's team in 2015, he not only scooped up the top five world championships like ImageNet and MS COCO, won the best paper award at CVPR 2016, but also was cited by AlphaGo Zero, which greatly enhanced the efficiency and accuracy of decision-making , The real value of innovation is that universal ResNet becomes one of the foundations of the vision and of the entire AI community because it is one of the most popular architectures in computer vision that addresses core issues.
India odd: Kuang vision three years focused on financial security, urban brain, mobile smart
Since its establishment in 2011, Indochic leadership has gone through six years as a technology leader, and for industry innovations, it is not surprising that tech CEO Indochina has his own experience. 'Defining a problem is often more than solving a problem. For example, in the field of financial security, the product of open-mindedness has a very good application scenario, and now it basically covers all the head users of Internet finance .From the beginning, we think that for users, Face recognition itself is the most important technology, but with the deepening of the industry we found that in vivo technology and anti-fraud technology is the essence of the pain point of the pokes the user point, and our definition of the essence of this industry is based on the deep bar To the constant practice in this industry and scene.
For the next three years of planning, India Qi said that in the next three years, we think the most important thing is to 'empower the eye of the machine to build the urban brain.' On specific business, Kuangshiyu will focus on financial security, urban brain, mobile smart Three areas.
India odd that the future of security must be partial networked, centralized system, rather than hardware-based. 'Empowering the eye of the machine, is to create a combination of cloud and the end, combined with the financial, real estate, security, different industries , Kuang vision to become the city's brain builder.
In the case of high financing, Indigo claims mainly to prove leadership in the industry, "the AI industry is not at the stage of burning money."
Is artificial intelligence and human intelligence finally a solution?
Finally, the academician Yao Zhizhi gave the ultimate answer when answering the question of whether or not artificial intelligence and human intelligence ultimately are a set of solutions: Quantum computation may be the theoretical basis of artificial intelligence, and he proposed: 'We humans want to imitate nature This is a last-ditch point, and once we have a quantum computer, we can mimic the operation of everything in the universe, including design materials, so we can count quantum computing as the first thing we are competing against God '
According to Yao Chi, artificial intelligence is a new beginning for us, which requires us to understand the underlying things before doing the algorithm, which means 99% of the future should be done in a new way. "We use physics Home-based approach to algorithmic design. "Yao Chi-chih emphasized: 'As a computer scientist, open up your mind to solving problems and create a theory that suits us.'
Artificial intelligence is getting hotter and hotter and more and more people are pouring into the torrent.Perhaps it is the key to the next level to calm down in a noisy environment and find the solution to the intrinsic innovation of artificial intelligence.