Steve Deng, associate professor of software at Tsinghua University, a well-known scientist of Artificial Intelligence and the chief Artificial Intelligence scientist of MATRIX blockchain team.
After getting his bachelor and master degrees in Tsinghua University in 1995 and 1998, Steve Deng went to Carnegie Mellon University in 2006 to pursue his Ph.D in Computer Science. Before he got his Doctor degree, Steve Deng had already been hired as a counseling engineer in American Incentia Design Automation company. In January 2006, Steve Deng joined American Magma Design Automation company and became a counseling researcher. In March 2008, he came back to China acting as the assistant professor in the Institute of Microelectronics of Tsinghua University.
Steve Deng’s major research direction is Al, the automation of electrical design, parallel algorithm and image processing machine structure. He used to design and develop an AI predicting safety plan for Chinese high-speed railway.
Steve Deng has published many well-known college textbooks such as Structured Integrated Circuit Design and High – level Synthesis and also published over 20 theses in leading journals. He used to lead a team for Pascal AI international competition and won the first prize.
We visited Prof. Steve Deng on October 18 to gain this interview. In the Bayesian laboratory of Software School of Tsinghua University, with his academic background, Steve Deng showed us how blockchain technology would blend with AI in the future.
Encountering the Blockchain
While Steve Deng’s worked in the United States, he was engaged in the computer-aided design of integrated circuits, which is a variety of algorithms for designing integrated circuits. At that time, Sun Jiaguang, academicians of Tsinghua University, was working on a research program to solve the problem about the control chip of the national railway train data. One of Sun’s students recommended prof. Steve Deng, thus, Steve started designing the train data communication chip.
After working on the design of the communication chip, Steve Deng found some new possibilities in his research. He found that the existence of the virtual currency is reasonable and the currency is created by trust; currency is the embodiment of the trust mechanism. However, Steve reckons that people should understand that there is no need for binding blockchain to encryption currency. Encryption currency will use blockchain technology, but encryption currency is only one part in blockchain applications. Blockchain is a kind of generic technology, it can be applied to a scene which requires credible data.
MATRIX Blockchain Team
Steve Deng has always been devoted to academic research, he hopes he can do something better than others. So he set up MATRIX team with a group of like-minded scholars, to “do something” by using their knowledge and imagination.
Plan One is to use the computational functions needed in academic research to replace the calculation function of bitcoin mining.
MATRIX team considers the algorithm as an alternative, the reason is that (1) the algorithm needs enough computation; (2) the end of each sampling charge for inspection verifier cannot ensure the calculation results, which conforms to the design of bitcoin mining function to some extent. Importantly, this study can be applied to the internal brain computation, image analysis, data processing and financial analysis in academic research.
Plan Two is to use artificial intelligence to solve two core problems in the practical application of blockchain, so as to realize blockchain’s self-iteration.
A: The hierarchy of blockchain
Blockchain technology would apply into peer-to-peer networks, each transaction is on broadcast and it would need 51% of the computing force if change the record. One of the best ways to manage is to make it hierarchical and change the existing consensus structure from flat network to hierarchical network. It is not necessary for some consensus to descend to all nodes, it could be in the middle; the middle node is a random selection, and it also represents physical time and a group of nodes around. To realize this kind of screening method requires the application of artificial intelligence’s cluster technology.
B: Intelligent design and verification of intelligent contract
The so-called “second blockchain technology generation” in the industry refers to the smart contracts of Ethereum. It requires each user to write a piece of code to generate a contract, but in reality it’s very difficult to implement. If using artificial intelligence, it can provide the following solutions : (1) script forms the intelligent contract code automatically; (2) analyzing the bug in the intelligent contract code that is already formed; (3) applying the depth neural network and antagonistic network technology to the in-operation code, attacking and repairing the code at the same time and make it became a safe contract gradually.
Embracing the Great Times
In October 2017, a research paper published in the international academic journal Nature showed that Deepmind, the affiliated company of Google, reported the new program “AlphaGo Zero”, the program can learn chess quickly by itself without any input from human, and finally with a 100:0 result defeat the master chess player. This exciting achievement inspires people to pursue the higher goals, the so-called strong AI, which builds machines that learn and think just like humans.
For Steve Deng, as the Turing award winner Alan Newell says, the most worthwhile thing to study is how human mind emerges from the universe. To find out the answers, an effective way is to realize the intelligence, which is the Artificial Intelligence. On the other hand, Steve considers the blockchain must be the trend in the future. The development of human society, or a better society, must be based on the credibility of consensus.
MATRIX AI Network is a new generation Blockchain that leverages the latest AI techniques to revolutionize the cryptocurrency landscape. With a fusion of Blockchain and AI technology, MATRIX has built a revolutionary cryptocurrency that supports significantly boosted transaction speed, superior accessibility to general users, enhanced security under malicious attacks, and highly flexible operations.
MATRIX will launch its ICO on Jan 16, 2018.
This is a press release. Trustnodes has not undertaken any verification of any of the above statements and any statement contained therein is not necessarily endorsed by Trustnodes. Readers are strongly advised to do your own research.