Full rundown of the 21M

How a video gets counted. When a fellow, mentor, or team member submits a video to the portal, there is an optional checkbox: "This video is about AI x-risk communication or education." Every video where the creator ticked it is listed. Views are summed across all the links they gave for that post.

• X-risk relevance was self-reported by creators. As such, some entries are loose, but after some later review I (Ronny Fernandez) think every video on this list with a large view count deserves the title "x-risk relevant".
• The largest fellow-submitted video by view count is a debunking of a popular AI water-use video, which helps raise the sanity of discourse around AI's negative externalities generally. The second most viewed fellow-submitted video raises the salience of the general feeling that things are starting to feel crazy and generally moving too quickly.
• Views are highly Pareto-distributed, as you would expect. This is true of any distribution of videos on social media.
• Fellows were sometimes overly conservative about what counts as x-risk relevant. Some genuine x-risk relevant videos are missing because the box was left unticked, or because a fellow did not want their content on the PDKU website.
• Videos from July 1 to 4 predate the checkbox and were tagged by an automated pass on the titles. This was fairly conservative.
• Instagram view counts come from our own scrape; TikTok, YouTube, X and Facebook views come from the portal. Only public portal posts are included.
• Mentors and team members for whom I (Ronny Fernandez) think PDKU significantly counterfactually contributed to their posting in July are included. For example, Rob Miles posted every day in July, which is very unlike his usual posting schedule.
• The number is an undercount in one specific way: it only sees links that were posted in the portal, so cross-posts nobody linked are invisible. For example, Nate Soares's July 29 reel has 336k views on Instagram but is counted at 3,856 from the TikTok link that was submitted.