James Bridle

While I agree that recommendation are what made YouTube so popular, and also boosts engagement a lot, they are also the most controversial feature of YouTube because they tend to converge toward the most polarizing videos, conspiracy theories, etc. They've also been gamed in a few creepy ways[1] [1]: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

If you somehow don't think there will be accountability for an individual employee who misuses PII, we are living in two very different worlds. You must understand that even a single employee doing this is such a breech of brand trust that Google would deal with it swiftly. OTOH, it is quite clear to me that Google executives are allowed to make virtually unchecked and wide-reaching decisions that impact many and have negative consequences. A simple, unambiguous example is "the YouTube recommendation algorithm is bad for kids and seems to cause their exploitation" [1] To be very clear: executives have the power to change the recommendation engine. To demonetize children's YouTube videos. They choose to be wreckless specifically out of greed. [1] https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

I would suggest looking through that previous link regarding examples of where one can quickly end up in white nationalist content through recommendations. Again, recommendations are for Youtube's own growth-related ends. However, it is emphatically a choice to have them. Regarding kids, this has been very well documented: https://www.ted.com/talks/james_bridle_the_nightmare_videos_of_childre… https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

Also, in case you glazed over the link to the recent Medium article, here it is: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… I read it last night after the wife recommended it ... and wow.

The author seems to be suprised by this, not just founding it on Spotify, but the whole process. E.g.: "I love it, because it’s so strange, so dizzying, and — credit where due–because our mystery producer is truly going with the grain of the medium, in a way that no one merely “making albums” does, at all. What could be more 21st century, more “liquid modernity”, than releasing your music as a haze of variations into the swirling currents of the algorithm?" But tons of artists have done the same in the past, whether in streaming platforms or in the vinyl and CD era. The most famous is probably Brian Eno who has made several albums using algorithmic generation. In general, algorithmic composition is a sizable niche, both in academy (tons of music PhDs on that), live performance, and recording settings. And yes, people release those things too. Even "generated content" to make bucks is a well known thing, both in music platforms like Spotify, in Amazon (tons of auto-generated Kindle books), and in YouTube (https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… The strange thing is, the author should already have known all this, as his previous post goes: "My band The Cotton Modules, formed with the composer Jesse Solomon Clark, also goes with the grain of the 21st century: our process combines AI tools with human skill and imagination, metabolizing a huge archive of recorded music into something genuinely new and exciting."

> Quite frankly it's ridiculous - imagine needing to pay a "fee" to CNN/Fox News/NYTimes/Yahoo just because you have a internet connection and web browser. This comparison is ridiculous. The license fee is a little outdated in certain aspects but it covers more than just broadcast news as the orgs you mentioned do. The BBC has a broad remit to inform, educate, and entertain. Not in itself radically different to a lot of other broadcasters, and typical for a public service broadcaster, but the universal license fee is a key differentiator. In requiring everyone with a television to contribute, the BBC has a duty to serve the entire population of a country, regardless of their income or demographic. This enables the BBC to produce content that may not be commercially viable for privately funded broadcasters that in term can serve communities that would otherwise be overlooked. BBC's Three and Four channels tend to best exemplify this aspect with content that can often be slightly niche and, in the case of Three, regularly have casts and production crews that are far more diverse than the industry averages. The license fee is also independent of the state budget. The Government does have the power and oversight to adjust funding every 5-11 years, but the royal charter ensures that funding is separate from typical taxation and avoids the Treasury. This also ensures (in theory) that there is more autonomy for accurate, fair, and unbiased news reporting. The BBC has also for…

I’m going to say something that will likely get me downvoted like crazy here on HN, to salvage some karma let me first say that I love the first amendment and free speech. Now, as a parent I am absolutely disgusted that there was recently discovered a YouTube “gang” pushing comments that were promoting pedophilia, in plain sight, tagging comments so algorithms could pick them up and other folks could easily find them. Secondly I am disgusted by the fact that there are videos embedded within Peppa pig videos on YouTube, in the middle, pushing kids to harm themselves and it is impossible to detect them. Google these things they are all real. Look I get it, I’m a parent, I’m responsible for my child so the YouTube app is gone, she will not get access to it until she’s old enough. That still does not make this stuff ok, these are terrible human beings creating this stuff. I applaud YouTube for stepping in but I still won’t install their app anymore, this does not mean it is ok for their platform to become a great place for that sort of stuff no platform should ever serve that. When folks on here comment and say one parents voice outnumbers 100 others, why the hell not if this content is targeted at our kids? Edit: Exhibit A: (This one is apperantly fake news, sorry. I was fooled myself a few days ago.) Exhibit B: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

I'm being a cheeky detractor of NYT here. I think "candid, engineering-style blogposts" are the future of news, and ancient vehicles like New York Times are long dead. I think the "general NYT audience" is participating in #FakeNews, and they should radically reconsider their information diet. As vivid example of this, compare James Birdle breaking the "Youtube exploitative kid videos" story way before, and in greater depth, than in any major publication. This is actually the future of news, and pretending like aging institutions like the New York Times are remotely relevant anymore is longshot wishful thinking. Editorialization, fact-checking, and cultural leadership have important roles to play, and I'm excited to see these features unbundled into separate services. I'm long on services like Verrit and Snopes, and wish that I, as an independent publisher, could pay an intern to get official statements, cross-check narratives with history, and perform some of these functions. As is, I think people are operating under the delusion that ONLY NYT-style institutions can perform these functions, which baffles me. (Actually, the future is probably more like James posting on jamesbridle.com, and then aggregating it through sites like Hacker News. But what do I know, I'm just a millennial who doesn't understand all these big partisan topics like modern journalism) https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

What is "demonetization"? I've never heard of this before. Does it mean YouTube stops inserting ads in your videos? As a viewer, that sounds pretty great to me. It sounds like they also stop suggesting your video to random viewers, which is understandable if they're not making money off it. Surely the real problem is that ads are toxic, the ad-supported content model is toxic, and YouTube's automatic recommendations are horrible (see the recent article on terrifying race-to-the-bottom videos for kids: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… Edit to clarify: yes, of course I realize that content creators need the money. I'm just saying that this is a bad system that often rewards bad content and punishes important-but-uncomfortable content. Unfortunately it's almost the only system in town. I would love to see e.g. Patreon rivaling YouTube in revenue. Edit edit: and ads aside, relying on YouTube (Facebook, etc) to tell us what to watch is terrible. We should instead do our own crowd-sourced curation, aka recommend good things to each other. And we need to defend ourselves against the likes of Facebook and Twitter insinuating their own filters between us and our friends.

> "Asked if he keeps passwords to his son and daughters' email and Facebook accounts, Mr Gates said that he doesn’t for Jennifer, 16, who he describes as 'independent'. I mean, honestly, I am very much about giving children their privacy, but when I have children I don't know how I feel about letting them on the internet unsupervised until they're twelve or so. Not to snoop on what they are doing, but because I don't trust what the internet is trying to do to them. There is so much active exploitation and manipulation going on. Not just from malicious individuals trying to get your credit card data, I mean that the most popular platforms themselves are Skinner boxes algorithmically optimised to make you click and consume more. James Bridle had an excellent article about that a few weeks ago, in which he called this "infrastructural violence", and I don't think that is an exaggeration[0]. and children are still innocent in the sense that they are not aware or capable of understanding these tactics. I want to be able to intervene until they are independent enough to judge those things on their own. For off-line usage it's a very different story. [0] https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

too short. keep going!! A lot of abuse results in PTSD. How will that manifest with online? Through the use of llm's, bots now talk just like the people the llm's are trained off of. They have hobbies, opinions, they will aggressively and convincingly argue that they are not a bot and feel insulted to be called one. They also cost non-insignificant resources to configure and operate, so how exactly a bot turns into revenue to pay for those resources is up for a lot of conjecture. I just spotted another article on a similar topic earlier today on hn. It's over six years old, but you may find it relevant. I would imagine the art has changed significantly, and the horror stories have only grown. https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

Your comment also reminds me a lot of the James Bridle piece from a few years ago about creepy, semi-automated YouTube content targeting kids: "Other channels do away with the human actors to create infinite reconfigurable versions of the same videos over and over again. What is occurring here is clearly automated. Stock animations, audio tracks, and lists of keywords being assembled in their thousands to produce an endless stream of videos. The above channel, Videogyan 3D Rhymes — Nursery Rhymes & Baby Songs, posts several videos a week, in increasingly byzantine combinations of keywords." https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

See also, "Something is Wrong on the Internet" "Disturbing Peppa Pig videos, which tend towards extreme violence and fear, with Peppa eating her father or drinking bleach, are, it turns out very widespread. They make up an entire YouTube subculture. Many are obviously parodies, or even satires of themselves, in the pretty common style of the internet’s outrageous, deliberately offensive kind. All the 4chan tropes are there, the trolls are out, we know this." https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

Maybe he is referring to things like the weird videos for kids that emerged from monetization: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

To be fair, this idea has come from credible sources too. From a few years back (2013?) https://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Emp… “According to our estimates, about 47 percent of total US employment is at risk.“ Even with replacement jobs that could be a painful transition. I’m a more worried about the jobs that are _created_ by ai at the moment though. https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

I'll cite a fellow HN user: https://news.ycombinator.com/item?id=15674870 Or this: https://news.ycombinator.com/item?id=15674305 Or how about this medium article? https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… Many people have presented their case on this issue. In this thread or elsewhere. Please don't pretend they don't exist by saying "Explain the actual harm. Enumerate it." Where's your proof that these videos are harmless or even beneficial to children?

This essay[1] by James Bridle points out that YouTube Kids' algorithmic filtering is failing. Thousands (millions maybe) of videos are uploaded containing violent and disturbing imagery and tailored to match the search inputs of children. The entire article is quite interesting. [1]: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

This is your usual boingboing churnalism. Original article is here: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39…

What does "checked the wikipedia page" mean? Reading the linked footnotes? And what about "sexual situations, fetishes, drugs" says "Tom & Jerry" to you? Not that the overview on that WP page does it justice, at all. At one point it even said Elsagate "was" a colloquialism, heh. It's hardly a full list of the themes that are repeated over and over and over again. It involves rape, pregnant infants, (there was one live action thing with Spiderman drugging and raping Elsa and her little child (played by an actual child, though with Elsagate you will also find plenty of adults in diapers or with pacifiers, or otherwise playing a child) as they are unconscious, then delivering both their babies in the form of plastic dolls, and Elsa and her child and Spiderman all being super happy and clapping in delight at the end), a LOT of kidnapping and bondage, children left behind, parents being evil for denying them sweets or staying up late or other things, parents getting hurt by their children or even murdered so children can get something they want, people generally murdering each other out of anger, including with automatic weapons, hands coming out of mirrors or mobile devices, children crying at a graveyard, hands coming out of graves, getting injections to get "made better" regardless of how silly that is (like eating a toy truck and getting an injection because of that), generally eating things and that having some kind of effect, a lot of binge eating, objects or people being…

This presentation briefly mentions but then seems to mostly forget about "Elsagate" which they call the "Peppa Pig scandal". James Bridle argues convincingly that the genre of bizarre YouTube videos which appeals to the toddler reptilian brain ( https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… ) is not created by hostile or evil actors but instead has evolved orgnically based on what stuff toddlers want to click on. Kids' click patterns reward more video themes like "Elsa tied up on train tracks kissing Spiderman", so the content industry crams more of that stuff into its new content. The result, after a few iterations, would not have passed editorial controls at 1990s Nickelodeon (!), which would normally have halted the feedback loop, but with no one at the helm -- to "censor" or otherwise exert editorial control -- YouTube's kid-targeted videos are just a whole forest of weird. Does YouTube want to allow their platform to become a laboratory for rapidly discovering local maxima in very young children's fantasy worlds? Do they have any choice? Should they step in and publish rules for what children's content is allowed? Should they hire some kind of human curator or editor to enforce those rules for child-focused videos? Should Web platforms act in loco parentis? In this cases, in "the Peppa Pig scandal" style situation, the producers are machine-generating content that gets clicks and the consumers are children. When the issue is the…

While it wouldn't solve the problem in and of itself, I think one mechanism that could help mitigate this phenomenon would be to compel companies like YouTube and Facebook to (somehow) disclose what their algorithms select for promotion (that is, for suggestion or automatic presentation). When the citizens don't and can't know what YouTube is playing for people, they can't really object. On the other hand, if the data was there and enough citizens were interested for somebody to analyze the data, then we could know: Oh, during the 6 weeks leading up to the Brexit vote, among the top 1000 videos X% strongly advocated for, Y% strongly advocated against, Z% didn't fit into strong advocacy. Oh, and, this false disproven pro-Brexit piece from Boris Johnson was played 18,960,000 times to an estimated 14,000,000 unique viewers. That would provide another vector of incentive/consequence that content promoters — algorithmic or otherwise — would have to factor into their decisions. I think this would also help with the other, extremely related problem of "algorithmic child abuse" described by the viral article "Something is wrong on the internet" from last year[1]. As we shift from a mass broadcast media landscape (TV, radio, print), increasing algorithmic content selection has huge implications, for society, democracy, children, etc. Because each person (or child) potentially receives their own custom-tailored feed of content (articles, videos) and ads, it is extremely…

> He knew exactly what to say to the responding officer to cover his ass Replacing human judgment with algorithms is a terrible idea for many reasons, but this might be the worst. Algorithms are deterministic, and thus they can be manipulated. "SEO" services exist in spite of Google's efforts to protect their ranking algorithm. Does anybody seriously believe their local government will be more successful than Google at preventing people (i.e. police, prosecutors, etc) from "optimizing" their input to the algorithm? edit - justification for the claim that a bail algorithm would be easy to manipulate: The recent single-pixel attack[1] against image classifiers. https://arxiv.org/abs/1710.08864 Google's failure to recognize obviously inapropriate videos in their "youtube kids" app ("elsagate") https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… (a good summary & discussion https://www.youtube.com/watch?v=LKp2gikIkD8 )

ElsaGate, enabled by Google and Disney (which owns at least network which owns at least one involved channel) and consisting faked videos for toddlers for strategic destruction and some sort of weird grooming. needles? taping people up? humiliating parents? kids driving cars? adults drugging people and getting them pregnant including minors? eating loads of feces? brand products left, right and center? verified accounts? monetized videos? people begging for removal since last year? but oh yeah look what China is doing. anyone feel like data crunching? crunch this. https://gitlab.com/elsagate/blacklist/raw/master/blacklist.txt and while the data is downloading -- because it's not possible that a community of hackers would look the other way when blatant child abuse (that's defined by what happens to the child, not what the perpetrator gets out of it) is going down directly under their nose or even hands, right? right? it's brain hacking, it's technology, it's mystery, it's capitalism, it's SV, what more do you want? nevermind the material responsibility, the real harm done to, the needless additional pain in the lifes of kids who 'just' get left to autoplay on youtube for long enough -- read this maybe: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39… edit: whoever downvoted this without argument, you'll think back to this and feel bad, if you're lucky and both informed and human enough. no more needs to be said. the info is there, CRUNCH…