Georeactor Blog

I'm reflecting on having my New Mexico Oryx video getting 115,000 views on YouTube (my best-performing Alice Guo video got to only 13,000; OSM 2,000; Supreme Court under 1,000). Now I don't mean to go too deep into stuff like, what compels me to seek out the attention economy. But assuming that I have a YouTube hobby and that it could be a positive thing, thoughts:

Algorithm and monetization

• The algorithm drives things and everyone tries to pin it down to rules, but maybe nothing can be learned from one lucky video. It's possible that most of my videos won't put up useful watch times and I just need a breakout video like this 1–2 times per year.
• At one point I thought that history videos would be valuable because they're evergreen, but very few views come from search. The algorithm is beaming out new content and figuring out if it will be accepted by larger and larger audiences, or maybe specific audience clusters, maybe?
• I joked about the Oryx video "destroying months of human productivity" but viewers were already online watching random YouTube videos (Oryx is mostly homepage clicks).
• From the start the Oryx video was better at % clicking, liking, commenting, and completing the video than any of my previous work. Still 70% of viewers click away from my video (think channel surfing vs. sitting through a podcast). I don't think I could fix this without making the thumbnail super-accurate or the content totally different.
• Even if I was monetized, at current math it wouldn't be a good $/hour (it would be nice if it paid for my YouTube Premium next year).

My content

• It makes sense to me to make a podcast-style narrative. And it should not be current politics. Coding and maps, IDK maybe.
• I'd like it to be topics which aren't covered elsewhere in pop science. Family members suggested topics which already circulate in the trivia-verse (capybaras as fish, Thomas Midgley Jr.). Reviewing my cytogenetics post I thought about a video on just the XYY panic, but then I found a Vox video with almost exactly the same points.
• Most people will be unfamiliar with the topic, but experts should be satisfied with the facts presented in the video, too (I was relieved to find hunters' comments didn't find the video to be flawed or critical of them).
• Some topics exist in local consciousness or popular history with ~3 facts which you hear over and over again without any other context. Supreme Court cases are the peak of this. The oryx video also fits into this space.
• I always edit the subtitles (21% of my viewers have English subtitles on, 61% of viewers on computers).
• I'm psyched if I can find a timely photo, newspaper archive, etc. first it might not have been connected anywhere else, second it's a real primary source, third it's a flex and tells the audience that I did the research.
• Not using AI here. Many other narrative videos are AI-narrated and have AI images, so that stands out.
• The remaining barrier to monetization is 1,000 subscribers, so even if a future video is a perfect fit for subscribers from the Oryx video, it won't necessarily help me reach monetization. It's possible that future topics will be too boring, and drive away subscribers who are focused on oryx, hunting, the American southwest, and exotic animals. I'm not super worried about this if I post regularly, appeal to multiple niches, and YouTube viewers continue to use the algorithm more than their subscriptions.
• The concept of the median viewer being someone in their 40s or older who watches AI narration on TV (here's an example on red onions vs. white onions), is something that doesn't settle well. They're part of the viewership, and I can imagine creating more videos which run in the background or appeal to someone 65+, but I want the channel to have quality which I can stand by, or share with a co-worker.
  Recently I was thinking, maybe this group includes 'unsatiated readers' where YouTube is interesting in the same way that I listen to podcasts or read the New Yorker. But for whatever reason (cost, reading glasses, attention and focus, personal image) they prefer a stream of articles read out on the TV.

Etc

• Maybe this will lead to writing a script for a mainstream science or travel channel. Or a magazine. That would be a neat side quest.
• If I find a nice YouTube audience, this could help with a future non-tech but more YouTuber-y project such as buying a house and doing permaculture.
• On the YouTube creator channels, people who are defensive about AI content (as their clicks and views win in an attention economy) but AI content is generating new videos daily and the most clickbait-y ones aren't true. Some people will insist that standard audio cleanup is "AI" too. There are other people who consider it to be a niche (i.e. I don't have any AI videos recommended to me, it's just what some people like to watch).
• Back when I did conferences and Hacker News posts, usually you could find comments on Twitter immediately after a talk or article. The oryx video has nothing like this on BlueSky or Twitter.
• I don't like how YouTube Studio handles comments. They default to "most relevant", which I have to switch to newest, and that doesn't pick up replies to others' comment threads.