⬤ Elon Musk recently explained how a new Grok-powered algorithm will change the way content gets distributed on 𝕏. The system uses AI to analyze posts at scale and match them with users based on what they're actually likely to engage with. Instead of relying just on basic metrics like likes or retweets, Grok digs deeper into content relevance and user preferences to decide what shows up in feeds.
⬤ The numbers are massive—Grok will scan every post published on 𝕏, which currently means processing over 100 million posts each day. After filtering through all that content, the algorithm matches posts to somewhere between 300 million and 400 million users daily. The idea is simple: show people content they'll actually care about, not just what's trending or what comes from big accounts.
⬤ One major feature is automatic spam and scam filtering. Grok is designed to catch low-quality and malicious posts before they spread, keeping feeds cleaner without manual reporting. But what's really interesting is how this helps smaller accounts. Right now, great posts from new or lesser-known users often get buried because they don't have the follower count to generate initial traction. The new system is supposed to fix that by evaluating content quality first, giving strong posts a chance to surface regardless of who posted them.
⬤ Musk also mentioned that users will be able to talk directly to Grok to tweak their feeds. You'll be able to ask for temporary or permanent changes to what you see, giving you more hands-on control over personalization. It's a shift toward letting AI handle the heavy lifting of content discovery while still keeping user preferences front and center. As platforms lean harder into AI-driven curation, systems like this will likely define how millions of people discover and consume information online.
Peter Smith
Peter Smith