The Twitter Algorithm is now open source available for download

Twitter has released to open source the source code for many parts of the platform, including the recommendations algorithm which controls the tweets users see on their timeline. The move comes as the company declares “a new era of transparency for Twitter” and commits to releasing future source code “that does not present a significant risk to Twitter or people on our platform.” The source code is available now on GitHub.

Twitter announces the release of its source code components and algorithm to open source on GitHub. (Source: Twitter)

On GitHub, there are two repositories available for developers interested in dissecting the code: one for the main source code repository, and another for machine learning. However, Twitter notes that they have decided “not to release training data or model weights associated with the Twitter algorithm at this point.”

Musk tweeted alongside the company’s blog post that the recommendation algorithm, claiming that the “acid test” will be if “independent third parties” can “determine, with reasonable accuracy, what will probably be shown to users.”

Twitter calls for any discovered security vulnerabilities to be routed through their official bug bounty program through HackerOne.

Twitter Algorithm Source Code Analysis

Twitter released the source code to open source repos on GitHub on March 31, 2023, and as expected, interested developers and security analysts wasted no time.

Key findings of the algorithm include:

  • Any topics deemed to be misinformation will be highly downranked
  • Discussions of Ukraine will be highly downranked. Musk remains highly ambiguous on his stance on the Ukraine and Russia war conflict. Ukrainian President Volodymyr Zelenskyy has asked Twitter if Musk is “one who supports Ukraine…or one who supports Russia?”
  • Tweets by Elon Musk are boosted
  • Authors of tweets are categorized as “power user,” “democrat,” or “republican.” Twitter defends this categorization as “used purely for metrics collection…this helps us validate in our A/B experimentation platform”
  • Twitter Likes boost 30x; Retweets 20x; Twitter Blue 2-4x; Trusted circle 3x; Images/Videos 2x; Replies 1x
  • Twitter Negatives: URL only, no text, mute, block, unfollow, report

What is not known at this time is if the release of Twitter’s algorithm or other open-sourced code components will contribute to an escalation of automated bots on the platform.

The release of the algorithm comes after news spread quickly that Twitter utilizes a secret list of 35 top Twitter users, including United States President Joe Biden, LeBron James, Ben Shapiro, and Musk.

Twitter bots have long remained a controversial topic of the platform, even as a scapegoat for Musk to initially try and back out of the purchase of the platform for $44 billion back in October 2022.

You May Also Like

Sam Altman Returns to OpenAI as CEO

In a whirlwind of events since last week, OpenAI’s board of directors…

COVID-bit: Exfiltrating data from air-gapped computers

A new research paper by Dr. Mordechai Guri, head of R&D in…

Google Announces Artificial Intelligence Bug Bounty Program

Tech giant Google has announced that it is expanding its existing bug…