Threads launches "Dear Algo": a new way to steer your feed with AI

Threads launches "Dear Algo": a new way to steer your feed with AI

Threads, the social network owned by Meta, has introduced a feature called "Dear Algo" that promises to give users an unprecedented level of control over the algorithmic recommendation feed. Users can now send direct text instructions to the AI, telling it what content they would like to see more or less of in the near term. The feature lets them retune their feed to match momentary interests without the tedious process of mass follows or unfollows. The tool reflects Meta's effort to make its algorithms more transparent and responsive to what the audience asks for, addressing the rigid framework of automated recommendations that is typical of modern social networks.

Over recent years, social networks have shown a clear trend toward giving algorithms a bigger role in shaping the user experience. Feeds used to be a chronological list of posts from the accounts a user followed. Today, AI actively selects content based on presumed interests, viewing history and interactions. On one hand, this approach is meant to make a platform more engaging and personalized. On the other, it often leaves users trapped in "filter bubbles", seeing only what the algorithm considers interesting to them and missing a broad range of opinions and topics. The lack of flexibility and transparency in algorithmic systems has become one of the main complaints about modern social platforms.

"Dear Algo" is meant as an answer to these challenges. Instead of relying solely on passive observation of what a user does, the algorithm can now receive direct, if temporary, instructions. For example, a user can write: "Show me more posts about cooking" or "Reduce the amount of political news this week."

These commands are, in essence, direct messages to the algorithm's developers, but delivered through the platform's interface. This means the AI will temporarily adjust its content priorities, favoring the topics the user specified. It is worth noting that the setting is temporary.

That prevents radical, irreversible changes to the algorithm's picture of a user's preferences, while still letting users experiment and adapt their feed to their current needs or mood.

The consequences of "Dear Algo" could be significant. First, it may raise user satisfaction, because people get a sense of greater control over their digital space. Being able to react quickly to changes in their own interests, without fiddling with complicated follow settings, makes using the platform more intuitive and pleasant.

Second, it is a step toward greater algorithmic transparency. By giving users a direct channel to express their preferences, Meta shows a willingness to engage in dialogue and make its recommendation systems less opaque. Third, the feature may help users break out of information bubbles by actively asking for content they do not usually see but are interested in at that moment.

This could potentially support a more balanced consumption of information.

In conclusion, the launch of "Dear Algo" in Threads is a significant step forward in the evolution of social network interfaces. By letting users talk to the algorithm directly and influence how their feed is built, Meta not only makes its platform more appealing but also shows a new vision of how humans and artificial intelligence interact. It could potentially become a new standard for other social platforms trying to make their services more responsive, transparent and focused on users' real needs.

*Meta is designated an extremist organization and banned in Russia.