by E. A. Barbour
For much of the 20th century, IBM dominated the data-processing field, both by control of the patents on Hollerith punch-card processing equipment and via considerable later innovation in electronic digital computers. Plus ruthless and aggressive marketing practices. By the 1960s the American computer industry was known as “IBM and the seven dwarfs”. IBM was larger than all of its competitors combined.
And yet, it’s difficult to get a feeling for this from the Wikipedia coverage of the company. Despite a 100-year history, long dominance of the data-processing field, and uncountable thousands of past and current products, the main article is only 65k bytes with 92 references. The entire Category:IBM contains more than a thousand articles, most buried in the “IBM People”, “Power Architecture” and “IBM Products” subcategories, thus difficult to find. The interest in Power Computing products is mainly due to Apple’s use of PowerPC processors in its products in the 1990s, not due to the Wikipedians’ interest in IBM. There is considerable Wikipedia content about IBM overall, but it is not obvious from the main articles, and much of it is highly specialized and aimed at current or recent products. IBM’s massive and complex history is not well documented, although there is a substantial “History of IBM” article , currently at 171k bytes. It seems rather poorly written and leaves out a lot of detail; as a summary it is just passable. Anyone looking for the “sum of all human knowledge” with respect to IBM’s history might be better off reading IBM’s website, supplemented by books like the 1986 text IBM’s Early Computers: A Technical History by Bashe et al., 735 pages that you won’t find on Wikipedia.
It’s not unusual to see IBM-related articles that were written by obvious company sockpuppets, and sometimes by their subjects. A perfect example is “IBM Distinguished Engineer”, written almost entirely by a user called “IBMDE”, who did little else on Wikipedia. Barry Leiba was heavily rewritten by user “Tagthestar14″, who just happens to be Barry Leiba. Nate Edwards was substantially written by user “Jpaulm”, who just happens to be longtime fellow IBM programmer John Paul Morrison, and who also worked heavily on his own Wikipedia biography. Abe Peled was worked over by “Abepeled”. And so it goes.
Even stranger is the List of IBM products, which is a gigantic, hot mess, almost totally dependent on IBM’s own history website, and still containing hundreds of red links, despite having existed since September 2002. Difficult to navigate; why bother looking at this, if IBM has its own, much larger history site? For the last few years the list has been substantially controlled by two IT workers, “Peter Flass” (a long-experienced programmer of IBM systems) and “Guy Harris”.
For that matter, I could point out the case of David Mertz. Better known as longtime Wikipedia noticeboard troll “Lulu of the Lotus-Eaters”, Mertz is a writer and columnist at IBM developerWorks. He created his own Wikipedia biography, which was AFDed three times, and is currently deleted. [Editor's note: "AFD" stands for "articles for deletion," a Wikipedia bureaucratic function whose initials have progressed from acronym to jargon to verb.] Mertz has also spammed his own writings on software development into a number of Wikipedia articles.
The book from hell
Or consider the case of Edwin Black and his 2001 smash-hit book, IBM And The Holocaust. He went to great lengths to research and prove that IBM was selling punch card processing equipment to the Third Reich, who then cheerfully used it to maintain databases about the concentration-camp inmates. Needless to say, IBM and its (few remaining) apologists were not happy about this, especially after lawsuits over it were filed.
And so the battle ended up on Wikipedia. An article about the book was started in May 2007, and soon attracted a whole lot of editing. An odd and seemingly-random series of editors, some identified by their IP addresses and some with Wikipedia accounts, all with no obvious connection to IBM, squabbled over it in slow motion.
Black himself, signing his comments with an IP address rather than adopting a Wikipedia account name, showed up on the article talkpage and started arguing with people he thought were IBM fans or employees. He also appeared on Wikipedia Review to make his case; Jeff “Ottava” Peters and Kurt “Fred the Oyster” Adkins showed up to attack him, for reasons that are unknown. Other denizens of WR were sympathetic but showed little interest in assisting him. Adkins was later blocked from Wikipedia over this issue.
Two of the people fighting over the article just happened to be Paul C. Lasewicz, IBM’s chief archivist, who blocked Black’s access to certain documents when Black was researching the book; and Brian “Blaxthos” Jacobs, an IT professional and self-assigned “1337 h4xor” who had started his career in the 1990s — as an IBM intern. (This fact is not noted on Jacobs’ LinkedIn account, strangely enough. Nor was the federal prosecution against him in 2000 for breaking into ISP servers.) Both Jacobs and Lasewicz were members of Wikipedia’s “IBM Task Force”, a project which apparently failed. Yes, they both worked on a wide range of IBM articles, including the massive “History of IBM”.
The battle heated up in 2009 and led to a December 2010 AFD, nominated by a politically-conservative Wikipedia editor who had no clear connection to IBM or this squabble and who popped out of nowhere. The AFD, which claimed the book was “not notable”, flopped horribly; since the book had been on the Times bestseller list, this attempt made no sense.
Just as an aside, Blaxthos wasn’t exactly the greatest Wikipedian we’ve ever seen. A Wikipedia article about his hacker-quotes website, bash.org, was created sometime in 2004, probably by him. After surviving a deletion nomination in 2004, the article was eventually deleted in 2007. Blaxthos, and/or possible sockpuppets thereof, showed up to fight for preservation in both cases. Even after the November 2007 AFD was closed as “delete”, others kept recreating it, and administrators kept deleting it, for two more years.
And now, we just have to mention one of the silliest “coincidences” in Wikipedia’s history, the appearance and rocket to the top of the Wiki-Heap of Karen “Fluffernutter” Ingraffea. Because it just so happens that Ingraffea is an employee of IBM, working as an “annotations team member” on their Watson AI project. She worked on Wikipedia during office hours, openly. As the Watson article on Wikipedia says, “Watson had access to 200 million pages of structured and unstructured content consuming four terabytes of disk storage including the full text of Wikipedia”. Yes, she has edited and patrolled this article, although not very much or very well. As with the other IBM content noted above, the Watson article is currently poorly written and organized.
Ingraffea showed up on Wikipedia with an account whose edit histories were mysteriously erased; her Fluffernutter account begins in April 2008. She frequented Wikipedia IRC channels, becoming friends with Wikipedia administrators like Oliver “Ironholds” Keyes and Brad “Courcelles” Brown. The latter, a powerful patroller, nominated her for adminship in August 2011. And married her in September 2013. She gained oversight powers in October 2011, and he became an arbitrator in December 2011 … almost as if it were planned.
Watson was used to play the “Jeopardy!” game show as a dry run. In January 2014, IBM announced that Watson would become a major business unit, with $1 billion of funding. And it incorporates Wikipedia, where one of the Watson employees just happens to be a Wikipedia administrator and oversighter, who is married to a (now former) Wikipedia arbitrator. Isn’t that an amazing convergence?
And the cherry on top: IBM’s Watson engineering team won the 2013 Feigenbaum Prize from the Association for the Advancement of Artificial Intelligence. And what did they do with the $10,000 prize money? They gave it to the Wikimedia Foundation. They loved it that much?
In conclusion, some questions
Since IBM’s employees are so willing and able to break Jimbo’s self-imposed bright line, why are they doing such a poor job of representing their company at Wikipedia? Perhaps IBM should make a decision, either to put more effort into improving its articles on Wikipedia to a more professional level, or else to leave them completely alone and stop risking the wrath of Wikipedia’s God-King Jimbo. Either that, or get more of its employees elected to administrator, ArbCom, and oversight roles, and proceed to take over WP completely.