Community Notes Update

After 3 weeks of being a Community Notes contributor on Twitter , I have an update that you may find interesting, especially if you believe in free speech, and especially if you believe in free speech even when you don’t agree with it.

I’m selective of which tweets I notate and which CNs I rank, ensuring they are familiar topics or where I know of trusted sources.?I’ve also tried to balance them to test my ability to provide fair and ethical responses even when I disagree with the content.

In the process, I’ve noticed a few things:

• Many people are reporting tweets simply because they do not agree with the content.?This is not the purpose of reporting tweets. Tweets should be reported when they break the terms of service. An example is posting a deep fake video as real, which is against the synthetic media policy. Eventually, the tweet will likely be taken down through the moderation process, but in the meantime, using the source of the unedited/source media in contrast is a great way to show users that the video is fake.
• The algorithm seems to identify content that may be incendiary and puts it into the CN queue, even when it’s true or is simply an opinion. This appears to happen both in terms of topics (e.g. vax, votes, etc) and authors (e.g. political satirists, certain social influencers, etc.) The content appears to stay online, and with these hot topics come many, many community notes to sift through. The hope is that only the best ones make it into the platform for users to read. (More below on how CNs are selected and viewable by users.)
• Repeatedly, a CN queue includes a tweet from Elon Musk and other high-profile tweeters that are one-liners or even an emoji, so CNs are being used as weapons to stifle even benign content simply because of who tweets it. I'm hoping as the algorithm gets refined, these will be resolved automatically.

Interestingly, here’s some insight into the algorithm: “Note statuses aren’t reached by majority rule. To identify notes that are helpful to a wide range of people, note statuses require agreement between contributors who have sometimes disagreed in their past ratings. This helps prevent one-sided ratings.”

The inner workings of the algorithm that manages this process are of great interest to me. As an ethical tech SME, I'd love to see/review the logic.

My stats are below as of 3/20 PM, and thus far, after more than 100 CN interactions, I only have good marks in the system.

The Community Notes ethos is important:

So, I also expect CN contributors to be required to perform at a certain level of ethical responsibility and to stick to the rules of the game and the purpose of the role. However, many on the community notes team are using CNs to voice their own opinion about the content. That is *not* the correct use of CNs.?If they have an *opinion* they should put it as a reply to the tweet. The purpose of CNs are to provide trusted sources and to help clarify language if the tweet is unclear. It’s also to provide guidance that a CN isn’t even needed! It’s not to debate content you don’t like or to hate on the original author.

For me, the jury is still out on if CNs are a good use of my time. I want to be a part of helping to ensure free speech even when it may be an unpopular opinion. If the back end of the system turns into a pile of hate soup and unaccountable CN content, I won’t continue to invest my energy there. But if it improves over time and is a helpful resource, I believe it will be worth my time.