Automattic is an important company in the history of the internet. Not only is it one of the first examples of a company that found success with a foundation of open-source software, inspiring many others to follow in its footsteps, but it has been long seen, at least externally, as a bastion of stability in an ecosystem when any potential tech service can conceivably disappear tomorrow.
That is in large part because of WordPress, the open-source content management system they helped bring to prominence that to this day roughly a third of all websites use. Sure, that puts a target on its back, but it also reflects its longstanding reputation as a sure thing.
Right now, however, I’m finding it tough to square its important role in maintaining one of the largest, most important open-source projects the world has ever seen with its recent actions.
If you find weird or unusual topics like this super-fascinating, the best way to tell us is to give us a nod on Ko-Fi. It helps ensure that we can keep this machine moving, support outside writers, and bring on the tools to support our writing. (Also it’s heartening when someone chips in.)
We accept advertising, too! Check out this page to learn more.
Honestly, it’s a bit of a double-whammy. The first problem emerged on Tumblr last week when Automattic CEO Matt Mullenweg, who brought Tumblr into the fold in 2019 and later became its CEO, tried to personally defend the company’s reasoning for banning a trans user, the result of which proved extremely controversial among the user base and, unusually, led to a rebuke of Mullenweg by the company’s staff account, in a post that specifically stated, in bold text, “Matt does not speak on behalf of the LGBTQ+ people who help run Tumblr or Automattic.”
The debate, which included a discussion about how serious a death threat actually was, eventually bled into the site I still call Twitter, where Mullenweg got into debates with the banned user and others about the reasons for doing so. Mullenweg eventually shared some details about the user’s private accounts, and later got into an extended debate about the concept of a “Lowtax speedrun.” (Some observers felt he was taking some slights, the kind of trash-talking common among public figures, way too personally.)
The Tumblr headquarters, circa 2012. (Scott Beale/Flickr)
To be clear: I’ve dealt with my share of Tumblr drama over the years. It’s a dangerous trap to fall into, and was admittedly one factor in my reason for ending ShortFormBlog back in 2014. But seeing the CEO of the company that owns Tumblr fail to stay above the fray is disheartening.
If that was all we were seeing, that would be one thing. But then, as my pals at 404 Media reported yesterday, the company has apparently been working on a plan to sell Tumblr and WordPress data to train AI tools, which the company then rushed an announcement for. It’s opt-out, yes, which at least offers an option to prevent this. But the reblog-heavy Tumblr is a bit of a sieve—if your content gets reblogged on 100 Tumblrs, if even one of those has the feature opted-in, odds are, they’re going to get your data.
The bigger concern, in my mind, is WordPress, which is very widely used by both individual users and large companies. While many of those sites are owned and managed in a self-hosted format, disconnecting Auttomatic from their daily management, some have been quick to note that the data collected by one of the most widely used WordPress plugins, Jetpack, is pushed directly back to WordPress.com. The company, if it doesn’t want to deal with controversy from Fortune 500 companies and large newspapers, needs to make clear that Jetpack will not allow for this data siphoning on self-hosted sites. It would be unacceptable to accept anything less for a project of this nature, that covers so much of the broader internet.
(Edit: Now we have that clarity! Please see update below.)
Yesterday, I somewhat clumsily tried to make a point about WordPress selling this data to AI companies. Which was, essentially, if WordPress—a supposedly neutral content management system—decides to sell such data, what stops, say, CloudFlare, or T-Mobile from doing it?
This is a company that not long ago pitched us on the idea of web hosting for a century, long after your passing—offering the ultimate support of end-user control—a decision that, at the time, I cheered. Yet, less than a year later, it has become mired in multiple controversies at the same time that, while they don’t necessarily suggest the firm is going away anytime soon, it does suggest that they may struggle to meet the long-haul goals that could keep them in a position to maintain content for a century.
Automattic is an important company in the history of digital publishing. They represent all the good things that online content can do, including the freedom for anyone to do it.
I am worried, via this double-whammy of controversies, that they’re in the middle of breaking end-user trust. Which could mean bad things for the open-source publishing ecosystem—an ecosystem that is ultimately bigger than Automattic.
Update (02/29/2024)
Since this was posted, I received direct confirmation from the general manager of WordPress’ Jetpack.com, James Grierson, that the company will not include Jetpack data in any content sold for large language models, an important point of confirmation. As the company states in its blog post: “We are not including content from sites hosted elsewhere even if they use Automattic plugins like Jetpack or WooCommerce.”
Automatic Links
I deeply enjoyed this feature in The Washington Post on Koji Kondo, the man who created the iconic soundtracks to many of Nintendo’s biggest hits.
Here’s a pretty great documentary on the history of the supercomputer, particularly the ones developed by Seymour Cray. The backstory is endlessly fascinating.
I support fellow former Motherboard freelancer Karl Bode’s efforts to attach the word “brunchlords” into our lexicon.
--
Find this one an interesting read? Share it with a pal! And back at this tomorrow.
