X is considering downvotes for replies to posts

As Elon Musk’s X continues to research the most “truth-seeking” systems to ensure accuracy in online debate, he is revisiting the use of upvotes and downvotes on post replies as a way to better rank engagement and surface the most relevant comments in each thread.

Well, negative votes anyway.

As you can see in this example, X is experimenting with a downvote option on post replieswhich would give users another way to share their feedback about users.

The option was spotted in the app’s backend code by a researcher Aaron Perriswith engineer X Jay Baxter then confirming the internal test.

Although it’s probably not exactly what you think.

According to Baxter:

“Community ratings wouldn’t work well without negative feedback signals. But you have to be smart about how you use them. If you naively add them up, you’ll end up with a hive mind like Reddit. One solution: Only downvote if you see negative feedback from people who generally disagree.”.”

So the downvote option for answers would be weighted based on political perspective, so that users of a certain ideology or viewpoint wouldn’t be able to nitpick answers and downvote everything they disagree with into oblivion.

The end result would then ideally be a system that demotes responses without political bias, so that when there are unwanted comments and spam, or clearly misleading and false statements, those are the ones that are relegated to the bottom of the list, or potentially even hidden, instead of silencing dissenting opinions.

It’s an interesting concept, similar to how the old Twitter team also experimented with downvoting replies, not as a way to express an opinion on the comment itself, but to reduce spam and junk.

In 2021, Twitter had already tried almost the same process, but without taking into account political orientation.

The idea then was that the downvote arrows would indicate the relevance and value of the replies to the messages, and not the personal opinion on the comments expressed, in order to improve the quality of its rankings.

As you can see in this image, Twitter actually tested several variations of the upvote and downvote UI, although part of the problem at the time was that users weren’t entirely clear on what these buttons meant.

This led many people to use them as an indicator of disagreement, as they are used on Reddit, which was not the intended purpose.

X now takes this into account by simplifying the process, in that you will, in effect, downvote comments you don’t like. X recognizes that those on one side of the political divide will undoubtedly vote against views they disagree with, just as they will the other, but the extra weighting will take this into account and ensure that impacts only apply where there is agreement.

Which should ultimately ensure that only the worst examples of spam and junk mail are penalized, not opinions as such. Which is a better approach, although a lot also depends on how X determines political orientation, and whether that process is precise enough to maximize relevance.

Either way, it’s an interesting experiment, which could help maximize in-app engagement by helping to highlight the most valuable post replies.

I mean, weighting political preferences is a slippery slope in itself, which can also lead to certain types of misinformation being kept in the app, because some topics will never be universally agreed upon.

But in this context at least, it makes sense as a way to weed out unwanted comments.

X hasn’t said this will be tested live, but it’s in an early experimental phase.