Blog post #4: Assessing current platforms’ attempts to curb misinformation
I think something I’ve noticed about misinformation on social media and from my own job is how important it is to get someone to stop scrolling and the techniques people use to do that. The first few seconds of a video, music, editing, text on screen, pacing, etc are all so important, especially on TikTok and Instagram. I use these platforms a lot in my personal and professional life, and credibility is mainly established on these platforms by how something looks and where that information is from isn’t a thought that crosses most people’s minds. If someone is confidently addressing the camera with clean captions and a popular audio, it is going to be convincing to the passive scroller, even if they are saying information that isn’t true, and most people wouldn’t bother to look further into it; if they do, they are “researching” it on the same platform by searching similar posts.
Recently, with Instagram, the way misinformation is approached on that platform has changed a lot, and Meta went from using a third-party program for fact-checking to using their Community Notes system, which lets certain users add additional content to posts that are misleading, and these notes have to have sources that support them and meet additional criteria before the notes are seen by the public.
I think this approach is cool because it aligns with a lot of the ideas and strategies we’ve learned in this class so far, like the SIFT method and lateral reading. Community Notes gives people additional context and sources to look at right there with the post, letting users view the information in front of them in a new, more accurate way. One thing that might hinder Meta’s approach is that social media content, especially on platforms like Instagram, is designed to move fast. It is all about creating content that catches someone’s attention, and people can swipe away from a post in seconds; if someone can quickly slip a misleading claim into that post, by the time a user sees the community notes, they have already scrolled past. Also, getting these notes on a post takes time, and the video could already have been seen by thousands of people who won’t take the time to look more closely at the claim. Also, under the new Meta policy, posts with Community Notes won’t be limited in reach, whereas before, posts faced distribution penalties. I find this strange because these notes and accurate content are useful, but if the post is still being pushed out to viewers and shared faster, most people still aren’t reading the notes, which makes misinformation spread faster and makes the issue worse. That is something I would definitely change, and I would have posts with misleading information have major distribution penalties. I feel like Community Notes should be pushed so users see them before even seeing the post, but I don’t know if that would ever happen.
TikTok’s algorithm and approach are so interesting to look at. Someone’s For You page is designed to introduce users to content and creators beyond those they already interact with, and TikTok says its recommendation system learns from users’ previous interactions to predict what else they’ll enjoy. This can be really useful and cool because it helps me discover lots of new information, trends, and ideas, which is useful for my job and entertaining for me. Unlike Instagram, TikTok limits posts with misinformation that could be harmful, and they won’t be put on the For You page. They don’t take the video down completely, but they don’t keep recommending it to more users. They also have a notes system that lets you add more context to their posts, which supports their other fact-checking programs. I think this approach is more effective because it actually stops the spread of misinformation, and there are lots of supporting systems to limit any harm.
More could be done to stop these posts from spreading sooner, especially on Instagram. If a post is concerning or needs review, its recommendation to other users should be stopped or slowed while it is being evaluated. I feel like creators would probably hate having this thought because legitimate content could be slowed down, but if information can be amplified so fast, why can it be verified quickly, or does the whole process just need to slow down a little to create more of a balance?