2012年8月3日,一位来自“南部北达科他大学霍普分校”的“Marcie Rathke”教授向一份“广受赞誉”的期刊——《理论数学进展》(Advances in Pure Mathematics)投递了一篇论文。仅仅10天之后,这篇论文就被接受了,编辑还一本正经地写了5点修改意见。没错,这篇论文正是用Mathgen自动生成的。虽然论文摘要每句话都看起来牛哄哄的样子,但实际上没有任何意义。
SCIgen的首次亮相是在2005年,机器论文Rooter: A Methodology for the Typical Unification of Access Points and Redundancy被WMSCI会议(World Multiconference on Systemics, Cybernetics and Informatics)所接收。虽然事后主办方澄清了此论文是作为“非同行评审论文”被接收并及时撤销了作者的报告资格,但在媒体的大肆渲染下,仍然是颜面大失。一方面,电气电子工程师学会IEEE暂停了对会议组织者的赞助,将论文从资料库中删除,努力将影响降到最小。另一方面,三位年轻人却在三天之内从网络上募捐到了两千四百美元,他们去到佛罗里达,在召开WMSCI会议的旅馆里租下房间,举行了一场同样是机器随机生成的混乱学术报告,并将整场闹剧录下来放在了互联网上。
在米国得手之后,德国又有几位学生开始用这件神software黑天朝。他们向2008年和2009年中国武汉举办的两个IEEE国际会议投稿(2008 International Conference on Computer Science and Software Engineering (CSSE 2008)和2009 International Conference on e-Business and Information System Security (EBISS 2009),文章署名Herbert Schlangemann(来源于德国的一部电影)也是编造的。结果两次电脑自动写的论文全部被接受,在第一个会议上,靠这几位童鞋为Schlangemann编造的假CV,虚拟的Schlangemann教授被当成了知名学者,还被邀请作为会议一个session的主持人。在EBISS会议组委会给自动生成的论文"PlusPug: A Methodology for the Improvement of Local-Area Networks" 作者的回复上,评价一栏全部为good和very good,“专家意见”也对这篇论文“高度评价”:
"Accepted. Topic of this paper is related to this conference. This paper mainly studies on a methodology for the improvement of local-area networks. The author introduces a scalable tool for evaluating Internet QoS(PlusPug), which the author uses to prove that digital-to-analog converters and XML are regularly incompatible. The author’s work is suggestive, and the point of view deserves researching. Figures in this paper seem exact, which makes this paper more convictive. There is no major modification opinion, only please pay more attention to Englishgrammatical rule and format your paper according to the template provided by conference homepage. We look forward to seeing you in EBISS2009."