Can We Make Video Streaming Healthier?
I have a problem with Netflix, and it's not just that you need a dozen subscriptions to watch everything these days. Even if Netflix had everything, I can't shake the feeling that it rests on a fundamentally unhealthy business model.
Netflix promotes binging, one of its key innovations. I'm sure the savvy business folks at Netflix chose that model for sound reasons. For many end-users, however, binging causes unstructured, unplanned media consumption. It's common to have an entire day pass without realizing it.
Unplanned binging is not the user's fault. “Just use the tool more responsively” only works when the tool is truly neutral. Neutral tools do not exist. Autoplay and auto skip are there by default for a reason. A user who wants more intention behind their media consumption has to fight against addictive tendencies.
Netflix causes FOMO. The front page is a near-endless scroll of algorithmically generated content, each thumbnail promising more dopamine rush. Yet nobody has time to watch everything, not even professional critics, who watch up to three a day. The endless sea of recommendations is overwhelming and impossible to tame.
Youtube, Hulu, HBO Max, Disney+, etc., have much of the same user interface, all vying for a pie of the video content market.
The concept of affordable online video streaming is good, and some even consider it a fundamental technology. Can we design augmentations that enhance the user experience? I can think of two ways.
- Hand-curated content for a niche audience.
- Algorithmically recommended content with customizable parameters.
I love that hand-curated lists encourage me to check out new content that I otherwise never would've. In Nebula, for instance, I discovered several quality creators that Youtube, with its highly funneled algorithmic recommendations, wouldn't have shown me.
The latter is the more interesting solution. It gives some control over the recommendations feeds to the users. For instance, users would be able to
- Turn off recommendations.
- Limit the number of hours of recommendations displayed per day/week/month.
- Receive recommendations as a customizable email newsletter.
- Tune how much “experimental” content, which the algorithm is unsure if the user will like, will be shown.
- Select the category of recommendations.
Currently, we need external tools to achieve this. For instance, the alternative Youtube frontend Invidious allows users to turn off recommendations and trending tabs. Browser extensions can also selectively hide elements on a webpage. The only way to achieve something similar to the others is through newsletters. A cinephile newsletter, for instance, would give you a predetermined number of recommendations which may include movies that you otherwise wouldn't have tried.
I quit streaming services entirely after being fatigued by too much content, FOMO, and decision paralysis. I discover new movies and TV shows nowadays through word of mouth. I tend to rent them out digitally. While it's more expensive than a Netflix subscription from a cost-per-entertainment-hour perspective, Netflix tends to make me get addicted to it, and I don't watch many movies and TV shows anyway. I also use an Invidious mirror yewtu.be to subscribe to a select number of favorite creators.
I still use Millie, a Korean ebook subscription service. First, it's a lot harder to mindlessly binge books than visual media. Second, it's difficult to buy Korean paperback books abroad. (I prefer to read books written for the general public in Korean and more technical books in English.) Third, I read a lot, so it has a good value proposition.