Reflections about Internet streaming and television

The million dollar discovery after ~10 years of work with multiple video platforms is the following: the next step for Internet video streaming is not about content, it’s about curation.

When you engage the Internet through the video streaming lens, engaging with most public popular video streaming platforms your realize that the average quality is really poor. It’s like 80 to 90% of the content is terrible or a complete loss of your time.

You usually stay hooked because every once in a while you find something genuinely entertaining and useful which convinces you that the platform is not complete garbage. But overall you’re loosing time scrolling a gigantic pile of crap. That’s a reason why a lot of people tend to consider Youtube as a loss of time or an activity you’re doing when you’re lazy, although there is really good content on the platform occasionaly. The algorithms strategy obviously makes this worse to maximize attention and profit.

As a result, television today stays relevant because although the content will never peak like something rare you’d find on the Internet, the overall quality is steady and you hardly ever fall into terrible quality or a complete loss of the time. There’s consistency on the television that the autoplay feature of a site like Youtube couldn’t reach in a million years. Because for a few good videos you get 80% of attention begging garbage. Even famous and talented video producers fall into that trend.

What’s lacking is essentially what we call a “directeur des programmes” in french television. A guy who’s only mission is to curate as best as it can quality content and assemble it in a coherent way for people to enjoy it and stop loosing their time. Major video platforms thought they’d crack it through the “algorithm” or maybe they believe AI will eventually solve this. I believe that’s wrong, you cannot emulate good taste via algorithm, because good taste precedes algorithms: that’s the data they’re feeding from.

The bottom line is this: we don’t have a problem of a lack of content, we have a quality issue that only human driven curation will be able to solve. With the current data sets at our disposal and via intelligent aggregation we could build fanstatic medias.

We cannot solve this on the web, we need to come up with a post platform solution. Platforms want control over curation to maximize profit and I won’t blame them: they’re companies. But I would say that the overall loss of time they imply is immoral at this point.

The minute you try to aggregate videos on the web you get sued or blocked by standards. That’s an issue we have to solve on the client side. That’s also a genuine argument for why the web is not the end game for software in general.

When it comes to online video, accessing the entire world is not enough you need human based curation.

Benjamin Arnaud