Late last night at a machine learning conference in Chicago IL, a civilized argument grew into a violent riot over two software libraries. Due to inexperienced organizers scheduling lectures and round panels, the TensorFlow and Scikit-learn library development panels were not only scheduled at the same time but in adjacent lecture rooms. According to surveys from prospective attendants, scheduling these to round table panels at the same time in the two adjacent largest rooms made much sense on paper. However, to a conference organizer who is not familiar with these two machine learning library communities it would be a mistake the Hyatt Regency conference center would regret.
TensorFlow and Scikit-learn are two popular machine learning libraries. While TensorFlow provides the lego like building blocks for many machine learning algorithms, Scikit-learn provides the ability to use many useful algorithms in one line of code. For years many data scientists and algorithm developers have built loyalties with each library. There have always been differences in between each tool set, such as TensorFlow’s ability to optimize neural networks or Scikit-Learn’s classification and regression visualization libraries, but this year there couldn’t be more differences in its’ user base.
In a normal year these two communities would work together to solve many problems facing the machine learning community but of course this isn’t a normal year. Facebook algorithm developers becoming increasingly more jealous of Google’s TensorFlow library began spreading hateful and misleading information trashing the codebase as ‘overly complicated‘ ‘too verbose to be useful‘ and ‘a waste of time‘ compared to Scikit-learn. In defense of the library many TensorFlow fans began flooding the message boards to strike back. There they accused Scikit-learn as ‘only being good for demos‘ and ‘Designed by an engineer and not a statistician‘. The online tensions were at a peak before the big Chicago conference.
6pm CST: When the simultaneous round table panels began, the attendants of the conference were mostly civil. Of course much of the divisive online language would bubble up here and there but it was nothing that couldn’t be handled over a couple of complimentary drinks by the Hyatt’s catering service. At the start of the panels, the drink tickets ran out and each room became more and more agitated while the developers presented more competing functions and additions to each library. The competitive panels turned into trash talk and on the bathroom break, vitriolic arguments broke out in the hallways of the conference center. Many drinks were thrown into the faces of colleagues who spent years working together.
8pm CST: At the conclusion of each of the feuding panels and that days lectures, the hotel room was a powder keg. Even more advances had been made in each competing library and all the engineers, coders and statisticians thought ‘any one who would use that library over my library would be a complete idiot’. Many long time working friendships were broken at that moment as the division had now become clear at the conclusion of the day’s scheduled events.
9pm CST: In an elevator on the way up to get changed for the bar. A suggestion was made that Scikit-Learn’s classification functions ‘weren’t real machine learning’ was when the fighting broke out. First with glass bottles thrown against each rival gang of developers rising to fist fighting. The video was uploaded to youtube and the conference attendants went nuts.
10pm CST: As more members of the conference viewed the fight, TensorFlow users blamed Scikit-learn users and vice versa. Many of the attendants who were members of a now growing radical rival facebook groups gathered at the hotel lobby and bar. It didn’t take long for the management to kick everyone outside the hotel room and into the streets of Chicago before there could be damage to the hotel. There, a brawl would begin that changed the world of machine learning.
10:30pm CST: Sporting masks, improvised weapons and shields the two sides began clashing in the streets. Temperatures were boiling when piles of bricks were left out on the streets. Tossed brick after dumpster fire, the crowd grew as more local work from home data scientists heard of the clash and joined in after learning from the online groups. Upon the first store looting and fire, the police became involved to force these rabid scientists and engineers to heel. It took hours before the action died down late in the night.
“This violent protest, or whatever that was, is an embarrassment to this city and the scientific community,” Lori Lightfoot, the mayor of Chicago. “Supporters of the TensorFlow and Scikit-Learn libraries are now banned from organizing and attending conferences in this city from now on.” The cities leadership took a hard stance to the violent uproar last night. Many in the ACLU have already began defending the right to gather and support a favorite machine learning library “…Most of the software developers last night were peaceful. There is evidence that the developers of the PyTorch library placed the bricks and infiltrated the communities to stoke violence.” One ACLU lawyer defended.
After this night of violent looting and rioting, many more cities with heavy engineer and data scientist populations are concerned they may be the next target of the violent actions. “I am concerned for Atlanta,” Mayor Keisha Lancebottoms commented “Yes we have a heavy population of young professionals who use both these libraries. I do not think we will see such violent protests but I believe this city is mature enough to handle such a divisive issue.” Across the country conferences are getting canceled worrying many hotel owners who were already struggling from the Pandemic. Just when we thought this year was over we just get one more surprise.
If you enjoyed this article please like, share, and subscribe with your email, our twitter handle (@JABDE6), our facebook group here, or the Journal of Immaterial Science Subreddit for weekly content.
If you REEEEALY love the content, for the equivalent price of a Chipotle Burrito, chips and Queso, you could buy our new book Et Al with over 20 hand picked Jabde articles for your reading pleasure, it’s the perfect Christmas/Birthday gift for confusing your unsuspecting family members! Order on amazon here: https://packt.link/at4bw Please rate and review so that you can brag to your friends about having opinions or showcase your excellent taste in reading material!
One thought on “Clash at Machine Learning Conference Leaves Dozens Wounded”