Contents Short cut Area
GlobalMenu Short cut
Contents Short cut

Current Activities

HOME > News > Research Activities

Research Activities 본문 내용
Title The 422nd STEPI Science and Technology Policy Forum - Discussion on the Future Outlook of Science and Technology: Deep Learning and Blockchain
Date 2018-10-29 Hit 57  

The Science and Technology Policy Institute held the 422nd STEPI Science and Technology Policy Forum on the theme of "Discussion on the Future Outlook of Science and Technology: Deep Learning and Blockchain" on October 19. This event was set as a venue to prospect future development potential and long-term strategies of the nation based on the understanding of deep learning and blockchain technology which is gaining attention as the core technology of the 4th industrial revolution.

   

This forum began with a brief presentation by Dr. Byeong-Sam Choi, Head of the Office of Future Industry Research, and discussion sessions followed in a form of question and answer with experts from the fields of technology, academy, policy, and law in order to form a social discourse on the deep learning and the development of the blockchain. In the first session, the technological and social aspects of “Technical limitations and possibility to solve social concerns on deep learning" were discussed. As for the 'Usability of deep learning', there was a discussion about whether deep learning can perform well on narrative type knowledge including common sense or structured language besides analyzing patterns or rules based on data. Prof. Seong-bae Cho (Yonsei University) said that deep learning has limit in processing traditional knowledge with no pattern at this point in time, but researches are being conducted actively, so this problem can be solved in near future. He also said that deep learning is an excellent technology for data patterning better than any other technology, thus efforts to apply it to the industry are also necessary. In addition, Senior Researcher Hyeong-seok Chu (Software Policy & Research Institute) said that this problem can be solved if two types of structurally different artificial intelligences are used by combining deep learning that is optimized for pattern recognition and common artificial intelligence that can utilize tacit knowledge of human.

   

Next, in terms of society, there was a discussion about what utility or problem could be caused to humans when artificial intelligence is introduced into the market. Dr. Choi commented that responsibility is the most important issue for the dissemination and diffusion of products, and the companies or the government should initiatively lead this project. In addition, it is impossible to predict the reduction of jobs due to artificial intelligence and deep learning, but artificial intelligence will affect all kinds of occupations in society, and it is necessary to consider the role and purpose of human being in a changed society, and take preemptive responses such as re-education.

   

The theme of the second discussion was "The technical limit of blockchain and the possibility of solving the social concern". Regarding the transaction processing speed of blockchain and the system overload issue, Dr. Jong-hyeon Kim (Institute for Information & Communication Technology Promotion) said the demand for transaction and accuracy differs depends on the nature of transaction, so it would be possible to apply blockchain to suitable fields. Moreover, Mr. Jung Yeon-taek, a lawyer from D’Light said that the key attribute of blockchain is securing confidence through transparency rather than speed, so if blockchain can be implemented with a high degree of reliability, it can consequently have higher efficiency even if the speed is lower than other systems. In addition, about coexistence of personal information protection law and blockchain, he said that a contract can be concluded by combining subscription and approval, and the disclosure of coding content can be regarded as subscription. Thus, blockchain can be classified as a special form of contract that implementation is conducted simultaneously with approval, so they can coexist if separate regulations are supplemented.

   

As wrapping up the discussions, the panels concluded that deep learning and blockchain technology are component technologies to move to the next level from the past centralized science and technology development to decentralization. Therefore, we should not undermine their potentials due to hasty measurements at the initial stage, and it is important for researchers to freely share and mutually develop information on the open platform.

   

This forum raised policy implications on the technology and supply side, society and demand side, and advantages and potentials as well as limitations on promising technologies such as deep learning and blockchain that are attracting nationwide attention. Based on the discussions in the forum, it is expected to help establishing policies for future technological development, dissemination and regulation.