Recent research indicates that social media algorithms, particularly on X, may be pushing content that conflicts with users' core beliefs. These algorithms prioritize online posts that users reply to, and people tend to comment more on content they disagree with. The study found that X's algorithm amplifies posts related to tradition and safety, while demoting those concerning care for others or environmental protection. This trend appears to be stronger for Democratic users than Republican users. The algorithms learn most strongly from replies, which are infrequent but carry significant weight. Because users often reply to content that clashes with their values, the algorithm then serves them more of the same. While users follow accounts that generally align with their values, the engagement patterns create the misalignment. The study suggests platforms could better align feeds by asking users about their values. This approach could help avoid echo chambers and potentially reduce polarization. Ultimately, users should have more agency in determining the content they see on social media platforms.