--Steve Jobs
The algorithms are they way they are because YouTube has figured out what people actually want to watch. Surprise, most people still want to watch "lowest common denominator" content, just like they did before digital media.
"And so the children of the revolution were faced with the age-old problem: it wasn't that you had the wrong kind of government, which was obvious, but that you had the wrong kind of people. As soon as you saw people as things to be measured, they didn't measure up."
Would it not be easier
In that case for the government
To dissolve the people
And elect another?
Poem in Die Welt (1959)"The medium is the message. This is merely to say that the personal and social consequences of any medium - that is, of any extension of ourselves - result from the new scale that is introduced into our affairs by each extension of ourselves, or by any new technology."
--Marshall McLuhan
I evoke this point because I think it is crucial to remember that the content of a form of media is nowhere near as impactful as the medium itself. While TV as a medium was designed to cater to lowest-common-denominator values via video, YouTube and other digital video distribution services enable this same thing - but at enormous scale, brevity, and accuracy unknown to the TV medium.
This is not just like the era prior to digital media - this is a huge expansion of the era prior to digital media. And it isn't incorrect to point out how scary and dangerous that fact is.
And youtube is a much better platform than television for evolving the cheapest, crappiest "content" that is just barely good enough to get people to continue to watch. Twitch is especially interesting in this regard.
I've often noted that these days you get the entire actual content of a "<insert country>'s Funniest Home Videos" TV show in 60 seconds or less.
In better times I'll turn to documentaries or TED talks, but yes, L.C.D. content is sometimes exactly what is needed.
I personally agree with this. One piece of evidence is that thinking adults do go see Marvel movies instead of the Scorsese/Fellini/Antonioni/James Gray/Coppola/Fincher/etc. feature, when those are not options.
The internet media of today is a completely different industry from the television industry of the 1990's. Catering to the lowest common denominator isn't necessarily bad as long as it's not the only option. If, however, YouTube only monetizes crap and removes more thoughtful content, people will go elsewhere.
I don't think it is a conspiracy. It is just what the networks do. It influences society.
In other words, I think it's 50/50. Some times, the networks influence society and other times, society influences the networks.
The networks aren't in business to give people what they want. They're in business to make money, and doing what sells easily is the usual way to make that money. The BBC have a different business model, so they don't have to chase the money that way.
Jobs is talking as if people can afford the things they want - if he instead said that "the construction industry builds tons of cheap, shitty housing, because that's what people want! It's the truth!", it'd be immediately obvious that he was speaking blubbering nonsense.
It's definitely a balance. If a creator only pumps out high amounts of low quality content, they'll never gain a subscriber base. On the other hand, if the content is so short that subscribers only interact with their channel a minute per month, this isn't particularly useful to advertisers either.
I'm not convinced this bias is a bad thing though. My perspective has always been that a closer aligning of YouTube's financial incentives with the financial incentives of creators will definitely be a positive for the platform in the long run.
As more work is done on alternative methods of monetization (YTR, super chat, etc) that have different incentives than just watch time, the biases that come from ad based monetization might become less of a driving force for creators in the future.
> What DOES create a problem for artists is the tuning of the algorithms to be highly preferential to long form frequently posted content
Disclaimer: I'm typing this on mobile so can't look up the sources as the same time, so this is from memory, might get some details wrong
A relatively recent change in YT policy is that you don't see every video you're subscribed to. They select among your subscriptions. How? Well, apparently one very important part of the algorithm is that you have to upload at least one long form video every week. If you break the streak, you're at the bottom of the priority queue.
Which is crazy, because if anything, one reason to subscribe is to not miss out the content of people who only very rarely upload!
Or what if you're a regular content uploader but just so happen to feel like you deserve a holiday for one or two weeks a year?
Now this might have changed since they first started doing it and youtubers rightfully complained about it, but the fact that this was implemented in the first place, and the big effect it has on content creators is pretty disturbing.
(aside: this is why I don't rely on YT subscriptions; NewsBlur has the ability to turn a channel into an RSS feed. Of course, that in turn means I don't show up in the YT subscription statistics, hurting the content creators I follow, so I symbolically subscribe but ignore the YT subscription service. How bonkers is that?)
How do you explain the youtuber phenomenon, then? Unfortunately there seems to be very little in terms of content rather than commentary, as the latter is so much cheaper to throw together with no money.
The result is HOURS of "user attention". Which is specifically not the correct term... :) Sometimes they ignore a video and just chat.
But the point is, they use it as some people use a television. It's just "on".
Because of this, I've seen a lot of "top ten list of..." and "Did you know that...?" videos. Not to mention compilations of traffic mishaps and skateboarding fails.
Quality absolutely doesn't matter. Just ask Budweiser!
Maybe food analogy is good? If you put on the table: potato chips and a green salad with grilled chicken which one would most people choose? Could people agree on the definition of "quality" in this case?
I agree that the objective is definitely to maximize aggregate watch time, but, the practical implication that I've been hearing about is that longer more frequent uploads are key to getting a video to rise in search and recommendation rankings, and thus become discoverable. It's been a while since I've had any inside information, but this guide mentions as much http://techtipswithtea.com/youtube-tips/youtube-changes-algo...
The thing is, what they're doing is clearly working as far as optimizing the key metrics that they're looking for. It's also not necessarily to call high volume game commentary low quality - it's a lot of fun, it just ignores the many dimensions of great videos that may account for the long tail of viewed minutes, but are still a big part of the soul of independent art. Those videos may never reach the critical mass to become viral and make money for their creators, where before, the algorithm gave them a better shot of being seen and shared until they take off under their own momentum.
To make an unnecessary food analogy, it's like a marketplace that rewards a creator that can pump out $1 cheeseburgers at a huge rate - they're tasty and they're good enough to keep people coming back. Bigtime profits. Win! But it's sad, then ,if the marketplace is no longer viable for the artisan who can only make one fancy, excellent, but unconventionally flavored, cake per day. Hosting the cake maker may be a loss for a long time, but, if they never had a chance they'll never be around to have their innovative, challenging, masterwork (viral video) take the world by storm some day.
Mr.Weebl, the animator who made the ubiquitous early-internet viral flash movie about badgers, mushrooms, and snakes was recently talking about the issues he's seeing. It doesn't seem like any of the early viral animations could have gotten featured and taken off in the current environment.
Who knows, maybe everyone will go back to newgrounds?
How's this for a solution? We just return this stuff to the web. How about a service that lets you host and create embeddable video links for a small fee? There would be an API that lets people automatically tip creators per view. Discovery would be through search and through blogs.
At the very least, this would solve most of the censorship issues.
They could have an independent artists section, like the indy corner on their android app store, but which they don't really promote... presumably for the above reason.
It's ironic that YouTube, named in ironic allusion to the denigrating slang term "boob tube" for TV, has become just that, after all.
Once there's enough people that see it as a network of content that they can spend time on, they can then start encouraging and optimizing for specific, high-impact items instead of recurring content. Right now they are in the "let's pump out as many sitcoms as possible" phase, and hopefully they'll progress towards a "let's release a couple mini-series and made-for-TV movies" phase (allowing that they aren't generating the content in either case, just encouraging it's production and dissemination by offering cheap hosting and monetization).
What DOES create a problem for artists is the tuning of the algorithms to be highly preferential to long form frequently posted content E.g video game commentary. This is a big problem for creators who work in short, high effort, highly polished media - E.g animators.
This really really bums me out. I was an employee at YouTube during their period of great growth and I absolutely loved the platform and potential for compensation that they gave to niche and indy animators and musicians, and had great hope for a future where the democratization of video distribution would lead to a flourishing of art and entertainment that caters to a higher common denominator than the over-polished, focus group approved, mass market creativity that passes for main-stream.
Sadly, it looks like the prime goal of increasing return visits and raw view time has lead to algorithmic optimization of that same mass-media race-to-the-bottom in quality and diversity in search of a broadly palatable and compulsively consumable product. It saddens me and I still feel that they could have it both ways if they took a more aware approach to niche content. YouTube offers monetization - the fact that artists even need Patreon means that GoogTube is ignoring a need, and thus a business opportunity. Just because a piece of work may only appeal to 50,000 people world wide doesn't necessarily mean that it has less value or merit than one that appeals to 10 million.
I support quite a few animators on Patreon, and the impression I get is that they are all struggling to keep working on it as a primary job.