For the past 2 and a half years I've been running a YouTube channel with semi-automated content. The content in question is threads on AskReddit. I know a lot of people look down on content like that (tbh myself included) but to be fair most content on YouTube is very low-effort. It's not getting close to the views it used to, but now it makes an average of about $1k/month.
Our content is a lot higher quality than most other similar channels, mainly because the software I made automates so much. Which is why I think this one stood out. I effectively spend an hour a month or so just maintaining the software, while the actual work of uploading/making thumbnails is outsourced to someone I know. He spends maybe 15 minutes per video and we upload daily. It's a fun side-business to have but I don't expect it to be viable long-term unless I somehow pivot it into something else.
I didn't realize this was a thing. Who is your audience? I assume people who want background stuff going on while doing something else (like a podcast etc). Otherwise why wouldn't they just read reddit?
It's exactly this. I know of a couple people who will throw on a reading of a thread from a popular subreddit (AskReddit, TIFU, etc.) and treat it like any other podcast.
Sure thing. It has 2 parts, frontend (web gui in React) and backend (Nodejs + Puppeteer)
I have a web GUI for fetching thread comments from the Reddit API and where the user can quickly pick which comments should be filtered out. It also lets the user know if a comment is "bad" (violence etc.) from some automatic string matching so they can be filtered out or edited quickly.
Then on the backend I fill a HTML template with the comment data making it look pretty. Then I use Puppeteer to render and screenshot it into an image. Then I do a "segmented reveal", revealing one sentence at a time using HTML classes and CSS, depending on what part of the text is currently being read. It also scrolls down so the sentence currently being read is in the center of the screen, and does some other stuff like censor bad words using the CSS blur filter. Then I use ffmpeg with a bunch of complex_filter entries to add TTS audio to the screenshots, add transitions and stuff and concat it all together.
It's really not that complicated, but learning ffmpeg and making rendering reasonably quick and correct was a challenge.
Our content is a lot higher quality than most other similar channels, mainly because the software I made automates so much. Which is why I think this one stood out. I effectively spend an hour a month or so just maintaining the software, while the actual work of uploading/making thumbnails is outsourced to someone I know. He spends maybe 15 minutes per video and we upload daily. It's a fun side-business to have but I don't expect it to be viable long-term unless I somehow pivot it into something else.