This article was published on August 23, 2019

Dear Netflix: Please ditch the algorithm and focus on human-curated ‘Collections’


Dear Netflix: Please ditch the algorithm and focus on human-curated ‘Collections’

Netflix is reportedly testing an all-new recommendation system that harnesses the power of employed biological neural networks – that’s humans if you insist on dumbing it down. Dubbed ‘Collections,’ the new system works like every collection of anything did before about 10 years ago – people sort it out so you don’t have to. I think this is a fabulous idea and so should you.

The way Netflix does things now involves artificial intelligence and necromancy. The company gathers data on everything its users do on the platform, and then feeds that data to a bunch of machine-learning algorithms that continuously funnel users into titles that’ll keep them watching. It’s good old-fashioned psychological manipulation to keep us focused on the screen and not those monthly payments.

I don’t have anything to back up the necromancy part but in light of how we’re being manipulated by AI, death magic doesn’t seem like a big deal.

Now, of course, we could just not use Netflix if we don’t want to be psychologically manipulated by a machine that thinks I want to watch “Leprechaun In Space,” because I enjoyed “Ichi The Killer.” I do but that’s besides the point, which is: not using Netflix isn’t an option. I’ve been psychologically manipulated by so many corporations that my loyalty to anything I’ve spent more than 99 cents on is unwavering.

And that just leaves figuring out why, no matter what I watch first, the algorithm thinks I should watch “Friends” next. Or “Arrested Development,” “Big Mouth,” or “The Office.” Seriously, those can’t logically be the next best-related recommendation out of a zillion other TV Shows. Finished the last episode of “Making A Murderer?” Alright! Let’s check in with Joey and Monica.

That’s why we need human-curated collections. Robots are bad at comparing cinema. Sure, algorithms work like magic in the YouTube arena. Google’s entire modus-operandi is keep em’ watching and see what happens. Netflix, however, is where real movies and television shows live. We watch programs that win Oscars and Emmys on Netflix – and since those are crap so very often, we could use some human experts to curate all the great titles we missed while we were obsessing over whatever movie Ryan Gosling was just in. 

Sure, chances are most of the curated collections will be mainstream pleasers like “Breakup movies for people who love Howard Hughes” or “Our favorite 1980s slashers.” See for yourself, here’s a tweet with a screenshot of the app showing some of the Collections categories:

Yeah, I’m not oohing and ahhing. Maybe “Take a Closer Look” could be interesting? But the important thing is that human-curated collections bring renewed hope for overlooked cinema. At the very least, a human cinema expert will be far more capable of connecting obscure, yet relevant titles to blockbusters than AI is.

Humans are awesome at this because, even though we’re biased, we’re smart. Machines are just biased – try to convince the algorithm you don’t like Tyler Perry movies after you watch any comedy with a predominately black cast, I dare you. It can’t be done.

Collections could save us from this, but it’s just a test right now. We reached out to Netflix, a spokesperson told us:

We’re always looking for new ways to connect our fans with titles we think they’ll love, so we’re testing out a new way to curate Netflix titles into collections on the Netflix iOS app. Our tests generally vary in how long they run for and in which countries they run in, and they may or may not become permanent features on our service.

I sincerely hope Netflix leans on Collections and human-curated recommendations hard. I miss walking into a video store and checking out the “employee recommendations” display to see whether anybody cool worked there.

Well, Netflix, does anybody cool work there?

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with