X, formerly known as Twitter, is reportedly planning to reintroduce a downvote feature for posts—a feature it previously experimented with in 2021, before Elon Musk's acquisition of the platform. During the initial testing phase, Twitter explored both upvoting and downvoting buttons, mirroring the functionality seen on Reddit. Recently discovered code snippets within the X iOS app suggest that the company may be considering implementing downvotes specifically for replies, potentially aimed at enhancing how replies are ranked.
The discovery of these code references came from reverse engineer Aaron Perris, known as @aaronp613 on X, who frequently uncovers upcoming app features ahead of their official release. According to Perris, these references were unearthed in the latest iOS app update on Tuesday.
Finding a reference in an app's code indicating a new feature or functionality doesn't automatically mean the company intends to publicly release it. Social media platforms, including X, frequently conduct internal testing of potential features before deciding whether to proceed with them.
However, in this instance, comments from at least one X engineer and X owner Elon Musk have shed light on the potential addition. According to Jay Baxter, a senior staff machine learning engineer at X involved with Community Notes, incorporating a negative signal could prove beneficial. This approach aligns with X's existing crowdsourced fact-checking feature, where negative signals can influence the ranking of posts.
Baxter suggested that aggregating negative signals could create a 'hivemind-like Reddit' effect. However, he proposed a refinement: only posts receiving negative ratings from those who typically hold differing views would be downranked.
This concept mirrors X's approach with Community Notes, where consensus among individuals with differing perspectives is required before fact-checks are validated and published.
While Baxter didn't explicitly confirm active development of downvotes, he emphasized that X's Community Notes relies on private ratings to prevent a 'hivemind' effect. He noted that private rating data is anonymized and released after 48 hours to maintain the integrity of the rating process.
Elon Musk also contributed to the discussion with a succinct comment: 'True'.
Implementing a consensus-building algorithm for ranking posts could potentially assist X in highlighting superior replies within lengthy threads. This approach might also serve as inspiration for other platforms to adopt similar tools. In an era marked by conflicting truths, such tools aim to establish what most people accept as factual while attributing sources. Crucially, this moderation process is driven by the platform's users rather than a centralized authority.
Beyond potential downvotes, additional changes may be on the horizon for X. Another discovery by Perris within X suggests a redesigned user interface that defaults to hiding engagement buttons like 'Like' and 'Repost' favouring a new gesture-based menu for accessing these actions and others.
/socialsamosa/media/agency_attachments/PrjL49L3c0mVA7YcMDHB.png)
/socialsamosa/media/media_files/2024/10/21/leaderboard-new-alt.png)
Follow Us/socialsamosa/media/media_files/4ZOh4VAqOW2ZzCAhv98F.jpg)