X Is Transitioning to a Customizable AI-Powered Algorithm

X is transitioning to a customizable AI-powered algorithm that gives users greater control over feeds as well as more personalized content discovery.

X is moving towards implementing an entirely artificial intelligence-driven algorithm that will provide the most relevant content to its users. The aim is to boost engagement and increase daily visits to the platform by tailoring each user’s feed to their specific interests.

The Challenge of Low Engagement Rates

On Friday, the head of X’s department of products, Nikita Bier, explains the idea that drove this shift:

“The goal for your X timeline is to get out of the mainstream algo and the political crusades and find your niche. You should be able to post about your interests and have friendly, relevant people chime in. If you’re seeing gas station fight videos, your account is not ramped up yet. We are working everyday to fix this.”

Ultimately, X aims to assist users in discovering communities that align with their interests, relying on signals from content rather than generalized, broad trends.

Given that around 80% of X users do not engage with or comment on the site, this target faces significant obstacles. Promoting active participation and generating connections driven by interests remains a substantial issue for Bier and the team of developers.

Transitioning to a Customizable AI-Powered Algorithm

Elon Musk followed up Bier’s explainer with this note:

“The algorithm will be purely AI by November, with significant progress along the way. We will open source the algorithm every two weeks or so. By November or certainly December, you will be able to adjust your feed dynamically just by asking Grok.”

Users can modify their content feeds by instructing an AI called Grok to boost or reduce specific topics or posts.

Threads appears to be employing a similar strategy, allowing users to customize their feeds by tagging a @threads.algo profile in their posts.

The concept shared by Threads enables users to control the algorithm’s interest graph with specific commands, creating an experience that is more personal than mere liking or passively viewing.

Practical Limitations of Explicit Customization

While this technology may be interesting, widespread acceptance of manual adjustments to feeds might be minimal. 

Platforms such as TikTok already offer highly personalized streams that are automatically generated based on the user’s behavior and viewing patterns, requiring no effort from the users.

Some users may prefer the simplicity of automated curation over actively directing their feeds, so explicit algorithm tuning could remain a niche activity, regardless of platform support.

Historical Perspective on User Engagement and Privacy

In the aftermath of the year 2018 and the Cambridge Analytica scandal and the subsequent outrage, Facebook saw little tangible change in the users’ behavior or sharing of data:

Facebook’s vice president, Carolyn Everson, stated:

“We’ve not seen wild changes in behavior with people saying I’m not going to share any data with Facebook anymore.”

Mark Zuckerberg also testified to the same effect, telling Congress that the scandal had “no meaningful impact” on Facebook usage, illustrating how convenience can outweigh privacy concerns.

This preference for simplicity over manual control is the reason decentralized social networks struggle to gain momentum. People prefer platforms that understand their preferences automatically and make the process of finding content easy.

This is also a factor in the increasing difficulty for creators and users to naturally establish followers, as algorithm-driven discovery is replacing explicit subscriptions.

The Role of AI-Powered User Bots

Another trend gaining momentum involves the use of artificial intelligence-powered user bots to create the illusion of interaction. Meta is reportedly testing AI characters that communicate with users about their most beloved subjects, thereby giving the appearance of a receptive audience.

Although this kind of borderline behavior “cheats” the system, it can increase engagement and satisfy users’ psychological need to interact, even if some users are automated.

Platforms are likely to continue exploring AI-driven engagement and customization strategies to improve user experience, while maintaining user comfort and individual control. The effectiveness of these innovations will depend on how users accept and adapt to these new tools.

Bottom Line

The AI-powered, custom algorithm promises to be an exciting chapter; however, the actual adoption and its impact are yet to be determined, particularly in the context of users’ long-standing desire for a simple way to discover content.