Why is this interesting? - Thursday, April 4
On YouTube, algorithms, and bad recommendations
Noah here. One of the stories of the week is surely the big Bloomberg piece about YouTube’s inability to curtail runaway recommendations for all sorts of terrible videos. To me the real meat of the piece comes about two-thirds in:
“YouTube doesn’t give an exact recipe for virality. But in the race to one billion hours, a formula emerged: Outrage equals attention. It’s one that people on the political fringes have easily exploited, said Brittan Heller, a fellow at Harvard University’s Carr Center. ‘They don’t know how the algorithm works,’ she said. ‘But they do know that the more outrageous the content is, the more views.’”
Why is this interesting?
As Colin can attest, I’ve been talking up YouTube Premium (formerly Red) for a few years now. While I’m sure there are other benefits, the main perk is zero commercials. This is particularly helpful if you want to watch a bunch of short videos without waiting for thirty seconds for each to start. No commercial living also led me to the endless supply of lectures from universities and other institutions (I’m particularly fond of the Royal Institution’s science lectures), which I have spent hours watching at the gym on 1.5x speed (something you can set in the app).
Anyway, with all that YouTube time I’ve learned quite a bit about how they make recommendations. While it’s always dangerous to over-extrapolate from your own personal experience, I’ve been surprised at how quickly they can send you off the track you’re currently on. If I’m watching a physics lecture from a well-established institution, for instance, instead of continuing with more lectures from that institution, YouTube will instead recommend me a dude in his living room explaining relativity. It very much feels as if the goal is to share the wealth by exposing you to new users and channels, which will naturally be less established than the very small number of global institutions posting high quality videos on the platform (there’s just more randos than Royal Institutions).
All of this makes me think of a few things. First, for a really long time Google hid behind the idea that its algorithm told some objective truth when in reality it’s made up of the opinions from a bunch of engineers about what should rank where in search results. Second, this James Bridle post from 2017 about the craziness that is YouTube kids is a must read (this recent Reply All piece is a good companion). Third, there’s a whole thing in Jaron Lanier’s book about quitting social media that’s all about how the people that built the internet prioritized democratization over everything else, which at the time probably seemed fine but keeps coming back to bite us in the ass when we continually find that giving everyone a chance to get their YouTube video in front of the world no matter how insane it may be isn’t such a good idea after all. Finally, maybe the most interesting nugget in the whole Bloomberg pieces is how the law actually encourages YouTube to turn a blind eye:
“Yet, in the past, YouTube actively dissuaded staff from being proactive. Lawyers verbally advised employees not assigned to handle moderation to avoid searching on their own for questionable videos, like viral lies about Supreme Court Justice Ruth Bader Ginsburg, according to one former executive upset by the practice. The person said the directive was never put in writing, but the message was clear: If YouTube knew these videos existed, its legal grounding grew thinner. Federal law shields YouTube, and other tech giants, from liability for the content on their sites, yet the companies risk losing the protections of this law if they take too active an editorial role.”
That’s not good. I hope for the sake of YouTube and the great service it could be that they find a way to deal with this stuff, because I can personally say I’ve learned an incredible amount from the site over the last year and the great majority of videos I watched I never would have found without it. (NRB)
Chart of the Day:
The quick rise of Netflix original programming: Q1 2018 vs Q1 2019. (NRB)
Quick Links:
A Japanese style inn, afloat at sea. (CJN)
Thanks for reading,
Noah (NRB) & Colin (CJN)