Technical Countermeasures for Security Risks of Artificial General Intelligence
Human beings might face significant security risks after entering into the artificial general intelligence (AGI) era. By summarizing the difference between AGI and traditional artificial intelligence, we analyze the sources of the security risks of AGI from the aspects of model uninterpretability, u...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
《中国工程科学》杂志社
2021-06-01
|
| Series: | 中国工程科学 |
| Subjects: | |
| Online Access: | http://www.engineering.org.cn/en/10.15302/J-SSCAE-2021.03.005 |
| _version_ | 1797658386045599744 |
|---|---|
| author | Liu Yuqing Zhang Yuhuai Duan Peiqi Shi Boxin Yu Zhaofei Huang Tiejun Gao Wen |
| author_facet | Liu Yuqing Zhang Yuhuai Duan Peiqi Shi Boxin Yu Zhaofei Huang Tiejun Gao Wen |
| author_sort | Liu Yuqing |
| collection | DOAJ |
| description | Human beings might face significant security risks after entering into the artificial general intelligence (AGI) era. By summarizing the difference between AGI and traditional artificial intelligence, we analyze the sources of the security risks of AGI from the aspects of model uninterpretability, unreliability of algorithms and hardware, and uncontrollability over autonomous consciousness. Moreover, we propose a security risk assessment system for AGI from the aspects of ability, motivation, and behavior. Subsequently, we discuss the defense countermeasures in the research and application stages. In the research stage, theoretical verification should be improved to develop interpretable models, the basic values of AGI should be rigorously constrained, and technologies should be standardized. In the application stage, man-made risks should be prevented, motivations should be selected for AGI, and human values should be given to AGI. Furthermore, it is necessary to strengthen international cooperation and the education of AGI professionals, to well prepare for the unknown coming era of AGI.
|
| first_indexed | 2024-03-11T17:58:17Z |
| format | Article |
| id | doaj.art-b98eff623d98424a998a4533f7e27613 |
| institution | Directory Open Access Journal |
| issn | 2096-0034 |
| language | zho |
| last_indexed | 2024-03-11T17:58:17Z |
| publishDate | 2021-06-01 |
| publisher | 《中国工程科学》杂志社 |
| record_format | Article |
| series | 中国工程科学 |
| spelling | doaj.art-b98eff623d98424a998a4533f7e276132023-10-17T10:53:33Zzho《中国工程科学》杂志社中国工程科学2096-00342021-06-01233758110.15302/J-SSCAE-2021.03.005Technical Countermeasures for Security Risks of Artificial General IntelligenceLiu Yuqing0Zhang Yuhuai1Duan Peiqi2Shi Boxin3Yu Zhaofei4Huang Tiejun5Gao Wen6Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Department of Computer Science and Technology, Peking University, Beijing 100871, China Human beings might face significant security risks after entering into the artificial general intelligence (AGI) era. By summarizing the difference between AGI and traditional artificial intelligence, we analyze the sources of the security risks of AGI from the aspects of model uninterpretability, unreliability of algorithms and hardware, and uncontrollability over autonomous consciousness. Moreover, we propose a security risk assessment system for AGI from the aspects of ability, motivation, and behavior. Subsequently, we discuss the defense countermeasures in the research and application stages. In the research stage, theoretical verification should be improved to develop interpretable models, the basic values of AGI should be rigorously constrained, and technologies should be standardized. In the application stage, man-made risks should be prevented, motivations should be selected for AGI, and human values should be given to AGI. Furthermore, it is necessary to strengthen international cooperation and the education of AGI professionals, to well prepare for the unknown coming era of AGI. http://www.engineering.org.cn/en/10.15302/J-SSCAE-2021.03.005artificial general intelligence (AGI), security risk, risk assessment, coping strategy |
| spellingShingle | Liu Yuqing Zhang Yuhuai Duan Peiqi Shi Boxin Yu Zhaofei Huang Tiejun Gao Wen Technical Countermeasures for Security Risks of Artificial General Intelligence 中国工程科学 artificial general intelligence (AGI), security risk, risk assessment, coping strategy |
| title | Technical Countermeasures for Security Risks of Artificial General Intelligence |
| title_full | Technical Countermeasures for Security Risks of Artificial General Intelligence |
| title_fullStr | Technical Countermeasures for Security Risks of Artificial General Intelligence |
| title_full_unstemmed | Technical Countermeasures for Security Risks of Artificial General Intelligence |
| title_short | Technical Countermeasures for Security Risks of Artificial General Intelligence |
| title_sort | technical countermeasures for security risks of artificial general intelligence |
| topic | artificial general intelligence (AGI), security risk, risk assessment, coping strategy |
| url | http://www.engineering.org.cn/en/10.15302/J-SSCAE-2021.03.005 |
| work_keys_str_mv | AT liuyuqing technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT zhangyuhuai technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT duanpeiqi technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT shiboxin technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT yuzhaofei technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT huangtiejun technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence AT gaowen technicalcountermeasuresforsecurityrisksofartificialgeneralintelligence |