Not really, Elon Musk is partially making good on his promise from about a week ago to open up the He previously promised that he would do so in 2022, and did so by publishing a snapshot of the code shortly thereafter, but that repository was not kept up to date enough to qualify the X platform as most people’s idea of an open source product.
So again, this release is a promising step towards X being a truly open source product. The next step will be to update this code repository in four weeks, as Musk promised he would do.
Nevertheless, this release will not mean that the open sourcing of X can be characterized as a “promise kept”. In his January 10th X post promising this release, Musk said he would release “all the code used to determine which organic and ad posts are recommended to users.” From where I’m sitting, that’s not even close to happening yet.
That’s because on November 26 last year, Musk and Grok’s accounts posted that Grok is used to sort posts on everyone’s following feed by default, although it can be toggled from “Popular” to “Recent” to make it chronological. That algorithm appears to be missing. There are also ads in the Following and For You feeds on X, which Musk has indicated are served through an algorithm that he has said will be made public. So in my opinion there should be at least two more releases, possibly more.
Gizmodo contacted Ax for information on whether the advertising and following feed code has already been released, or will be released at some point in the future. We’ll update if we hear back.
But still, here we are with a fresh dump of the code. The first thing you should know, according to Musk, is that it “sucks.”
yes, the algorithm sucks
– Elon Musk (@elonmusk) 20 January 2026
Earlier that day, after Musk said the algorithm was useless, who is right? Is it better than before, or worse?
The problem may be that Musk can’t seem to clear out all the stubborn awareness residue that filled X when it was called Twitter. His tweet that it sucks was in response to a complaint by former video game executive Mark Kern that the algorithm gives less importance to posts if they come from accounts that have been blocked too much. Kern says he suspects the algorithm biases posts from right-wing accounts like his. I believe this is laudable, although it almost certainly biases the algorithm towards accounts that post a lot of harassment and abuse, so do what you will with that.
Based on what’s in the plain text readme documents in the Github dump, my impression of what this latest one describes is that, not surprisingly, it prioritizes engagement, trying to figure out which post will make the user stop scrolling. This is taken from the accounts you follow, but also from the accounts you follow. This is attractive to your id, not your superego. It doesn’t matter what you think you’re there to see, it wants to show you everything it can to keep you watching.
Besides sucking, Elon Musk also calls it “dumb.” Responding to a complaint by blogger Robert Scoble that the algorithm favors posters hijacking news events, Musk says the algorithm will improve every month – seemingly referring to the expected cadence of four weeks for GitHub code dumps.
We’re trying to make the algorithm less stupid.
It will improve every month.
– Elon Musk (@elonmusk) 20 January 2026
And who knows, maybe users with amazing ideas will search not only in the Readme sections, but also directly in the code, find real problems, and make suggestions to Musk, and the algorithm will become more satisfying and profitable over time. Alternatively, perhaps the needs of a company that wants to attract users to view ads and generate revenue for itself, and the desires of humans who want to feel well-informed and happy, are two completely incompatible concepts, and it is completely pointless to open source a recommendation algorithm to meet both of those types of needs. I guess we’ll see which of these possibilities is actually true.
<a href