Originally published at: https://boingboing.net/2019/10/24/computer-says-no.html
…
12 Likes
“Youtube, pioneering the gig economy since mid 2000’s”
7 Likes
It would be funny if Google used machine learning for moderation, so it’s really a black box to them as well.
4 Likes
Given all the You Tube drama why do people continue to use it?
Why has no one created a usable competitor?
1 Like
First, there are competitors like Vimeo as direct competitor, and Twitch for live streaming.
Second, there are several hurdles for competition with Youtube:
- Video streaming consume massive amounts of bandwidth, which is quite expensive.
- Badly written legislation favors established systems that had time and money to develop the necessary mecanisms to filter the content , to comply with things like DMCA and current EU shitty legislation about copyright protection. The latter is really encumbering as it puts the responsability on the content provider, not the uploader, so it’s a costly proposition to make a public platform for videos.
- You compete with a player that has the money to put the app on every cellphone, lowering the cost of entry, a bit like microsoft with windows in the 90’s - 00’s.
- People expects you to have a perfect solution; there is little tolerance for bugs given the maturity of YT platform. Oh sure, they complain about the algorithm, but there are much, much worse alternatives.
All of this costs money and certainly prevents “grassroot” as the environment has changes a lot since youtube started.
As for the first question… well: My guess is people like the platform enough: it is a mature product and works quite decently, and most importantly, it was working fine a year or two ago. People don’t understand why it stopped working, and Google is not very forthcoming with answers.
12 Likes
There’s a lot of niche competitors to youtube, but nothing with YouTube’s scope.
You have places like LiveLeak where there’s a lot of ad unfriendly stuff posted (eg military airstrikes in active theaters, and fatal car wrecks), or imgur where you essentially have a re-make of Vine with minute long “Gifs with sound”.
1 Like
I think it’s pretty well understood that most things Google does work exactly that way. The machine learning optimizes ‘eyeballs for advertisers’ - this includes both the raw numbers of views (and likes, shares, comments and subscriptions are likely among many inputs to the algorithm as predictors of future views) and the acceptability of the video to the advertisers. All of Google’s work on speech recognition, OCR, image understanding, generative adversarial networks, and so on is devoted to modeling not only YouTube eyeballs, but all of human behaviour that can possibly be monetized.
But that does imply that decisions are indeed made by a black box that nobody understands. Moreover, while said black box isn’t intended to be biased (any kind of bias cuts off a potential market!), it can’t help but reflect institutional and societal biases.
Doing it with humans wouldn’t work at scale. I don’t know if there is a better answer. I surely hope there is, because this is how the AI-pocalypse starts.
1 Like
Here’s a thought: Maybe if YouTubers didn’t consistently find new ways to be assholes, YouTube wouldn’t always having to be updating their rules.
1 Like
There are competitors but they are smaller cause all the eyeballs and ad revenue are on YouTube. Everyone wants competition for YouTube, but the cost is very high and long term with chance of success slim. Obviously few with any substantial following are willing to take the hit in viewership and ad money to help grow an alternative.
It only takes one person to shit in the public pool. Should everyone be punished for it? Should the rules that state that shitting in the pool is forbidden be hidden in a kafkaesque beaurocracy?
I’ve seen fascists get away with racist shit, while antifascist historians get videos banned for saying that the Nazis did some bad shit. Should Three Arrows pay the price for the actions of some arsehole with 1488 in their name?
4 Likes
From Google/Twitter/Facebook’s point of view: Yes, they should. Because if the rules are too clear and/or too effective, you alienate conservatives.
If a large fraction of the population voted for a president who says that there are “very fine people” who shit in pools, then you’d lose a good chunk of users and revenue if you make your no-pool-shitting rule too absolute.
Isn’t the moderation of large online platforms basically impossible? If YouTube/Facebook/Twitter has a rulebook that is thorough enough to ban anything that makes the world a worse place for being posted, then freeze-peach folks will complain about all the pointless rules being mindlessly enforced to protect overly sentitive people… but if the platform in question relies on human judgment rather than on a strict rulebook, then folks will complain about kafkaesque bureaucracies and whimsical/contradictory/theatrical decisions to take this or that down, about such decisions being abuses of power or disproportionately unfair to oppressed groups. I’m not saying either of those sets of complaining folks would be wrong; they would both have good points. Which is why I fear that the idea of moderating online platforms is hopeless, if these platforms hope to be a welcoming space to all (or nearly all) people. (I am a part of small online communities that are true communities in part due to good moderation, but those communities do not claim to be for everyone).
Wasn’t there an article about this, earlier this year? When Facebook’s rulebook was leaked, some people nitpicked some specific rules, some people complained about how vague it was, and finally some people realized that Facebook is in a “Damned if you do, damned if you don’t” situation when it comes to this kind of thing.
The rulebook:
4 Likes
The current alternative just pisses off everyone, and frankly I’m not particularly compelled to cede much opinion currency on the freeze peach crowd anyway, given how it’s often just dogwhistle for openly hostile methodologies.
If it’s manual, a clear rulebook at least gives people the ability to say ‘okay cool I’m going to try and engage with this in good faith’ and do content that doesn’t disagree with it. Yeah there’s always going to be the grey area stuff…but so what? Everything has to deal with it. It’s unavoidable, and they’re not avoiding it now.
Even if it’s algorithmic, there’s no reason the algorithm metrics can’t be published regularly. This all turns into ‘oh it’s the SECRET SAUCE’ but in reality the secret sauce is how you make and establish and implement those metrics. It’s just some youtube leadership going ‘oh god what happens if someone uses this data’ instead of realizing it would improve customer experience to do so.
1 Like
This topic was automatically closed after 5 days. New replies are no longer allowed.
