An autonomous debating system

被引：93

作者：

Slonim, Noam ^{[1
]}

Bilu, Yonatan ^{[1
]}

Alzate, Carlos ^{[2
]}

Bar-Haim, Roy ^{[1
]}

Bogin, Ben ^{[1
]}

Bonin, Francesca ^{[2
]}

Choshen, Leshem ^{[1
]}

Cohen-Karlik, Edo ^{[1
]}

Dankin, Lena ^{[1
]}

Edelstein, Lilach ^{[1
]}

Ein-Dor, Liat ^{[1
]}

Friedman-Melamed, Roni ^{[1
]}

Gavron, Assaf ^{[1
]}

Gera, Ariel ^{[1
]}

Gleize, Martin ^{[2
]}

Gretz, Shai ^{[1
]}

Gutfreund, Dan ^{[1
]}

Halfon, Alon ^{[1
]}

Hershcovich, Daniel ^{[1
]}

Hoory, Ron ^{[1
]}

Hou, Yufang ^{[2
]}

Hummel, Shay ^{[1
]}

Jacovi, Michal ^{[1
]}

Jochim, Charles ^{[2
]}

Kantor, Yoav ^{[1
]}

Katz, Yoav ^{[1
]}

Konopnicki, David ^{[1
]}

Kons, Zvi ^{[1
]}

Kotlerman, Lili ^{[1
]}

Krieger, Dalia ^{[1
]}

Lahav, Dan ^{[1
]}

Lavee, Tamar ^{[1
]}

Levy, Ran ^{[1
]}

Liberman, Naftali ^{[1
]}

Mass, Yosi ^{[1
]}

Menczel, Amir ^{[1
]}

Mirkin, Shachar ^{[1
]}

Moshkowich, Guy ^{[1
]}

Ofek-Koifman, Shila ^{[1
]}

Orbach, Matan ^{[1
]}

Rabinovich, Ella ^{[1
]}

Rinott, Ruty ^{[1
]}

Shechtman, Slava ^{[1
]}

Sheinwald, Dafna ^{[1
]}

Shnarch, Eyal ^{[1
]}

Shnayderman, Ilya ^{[1
]}

Soffer, Aya ^{[1
]}

Spector, Artem ^{[1
]}

Sznajder, Benjamin ^{[1
]}

Toledo, Assaf ^{[1
]}

机构：

[1] IBM Res AI, Haifa, Israel

[2] IBM Res AI, Dublin, Ireland

来源：

NATURE | 2021年 / 591卷 / 7850期

关键词：

LEVEL;

D O I：

10.1038/s41586-021-03215-w

中图分类号：

O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];

学科分类号：

07 ; 0710 ; 09 ;

摘要：

Artificial intelligence (AI) is defined as the ability of machines to perform tasks that are usually associated with intelligent beings. Argument and debate are fundamental capabilities of human intelligence, essential for a wide range of human activities, and common to all human societies. The development of computational argumentation technologies is therefore an important emerging discipline in AI research(1). Here we present Project Debater, an autonomous debating system that can engage in a competitive debate with humans. We provide a complete description of the system's architecture, a thorough and systematic evaluation of its operation across a wide range of debate topics, and a detailed account of the system's performance in its public debut against three expert human debaters. We also highlight the fundamental differences between debating with humans as opposed to challenging humans in game competitions, the latter being the focus of classical 'grand challenges' pursued by the AI research community over the past few decades. We suggest that such challenges lie in the 'comfort zone' of AI, whereas debating with humans lies in a different territory, in which humans still prevail, and for which novel paradigms are required to make substantial progress.

引用

页码：379 / +

页数：7

共 44 条

[1]

Aharoni E., 2014, 25 INT C COMP LING T

[2]

[Anonymous], 2019, OPENAI BLOG

[3]

[Anonymous], 1999, MAXIMUM ENTROPY APPR

[4]

Bar-Haim R, 2017, 15TH CONFERENCE OF THE EUROPEAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (EACL 2017), VOL 1: LONG PAPERS, P251

[5]

Bar-Haim Roy, 2017, ARGMINING, P32, DOI DOI 10.18653/V1/W17-5104

[6]

Bilu Y, 2019, 57TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2019), P1013

[7] Deep blue [J].

Campbell, M ;

Hoane, AJ ;

Hsu, FH .

ARTIFICIAL INTELLIGENCE, 2002, 134 (1-2) :57-83

[8]

Cho K., 2014, P 8 WORKSH SYNT SEM

[9]

Coulom R., 2006, 5 INT C COMP GAM INR

[10]

Daxenberger J., 2020, DATENBANK SPEKTRUM, V20, P115

← 1 2 3 4 5 →