If a company doesn't have the kind of technology to effectively transcribe their large volume of videos and the daily rate of uploads, then must they simply start taking down videos? If a company started hiring inhouse to develop such tech, how long would it take?
Given the performance of Google's auto-captioning, I suspect developing worthwhile auto-captioning is pretty difficult; according to [1] the better Youtube channels use gig economy captioning at $1 per minute.
Of course, there would be some scope for efficiency - no need to pay for captioning when the performers aren't speaking!
While Youtube auto captioning has a superior performance compared to Google Translate output. In youtube translations there is much more language specific nuance. In many European languages a more formal word for the English word "you" exists, in German it is "Sie", French "Vous".
Correct interpretations of these nuances are applied by Youtube, Google however translates "you" by default to the most formal option.
So to me $1 dollar per minute sounds like an awesome deal because there is not much to adjust.
Out of curiosity, can you link to such a gig company?
I don’t know if it’s a gig company necessarily. As I understand it they don’t use Amazon Turk anymore but I use CastingWords and have been very happy with them. I think they’re just English though.
I just mentioned in this respect Youtube performance is better than Google.
I do think the formal nuance is outdated and should be deleted. We know from Korean pilots that after switching to English as default in the cockpit a significant decrease in crash rates was accomplished. This was 100% ascribed to the lack of formal nuance in English. I have a hunch a lot of European languages are handicapped by similar conditions.
Presumably an option would be, if they want to self-host, is to run the audio through an ML service and post the link. Doesn’t really work for user-uploaded video as well. But it would be pretty easy to build into a content upload pipeline at low cost.
As others say it’s not perfect but I have to believe it would be good enough and probably isn’t a bad idea in any case.
That’s what I do for podcasts. Machine translation is getting pretty good but, for publishing an interview transcript, I’d still be spending a lot of my time to clean it up. For a human transcript it’s maybe 15 minutes work.
The word accuracy for auto-generated captions is highly variable, sometimes it's good enough, sometimes it's not. In addition, proper captions should identify speaker changes, significant non-verbal sound (e.g. [car honking]), and include punctuation, all things that most auto-generated caption services don't even attempt to do.
Again, what’s the standard for good enough? And I just ran an interview through an AI service. It caught speaker changes. And did punctuation. As well as a person, no? I use human transcription personally. But it seemed plenty adequate as minimal transcription.
What are you looking for, a percentage? I can't give you one.
Think about what the point is; the point is to give hearing-impaired people the same experience as hearing people. If you quizzed some people who only had the audio and others who only read the captions, both groups should be able to answer questions at the same rate of success (caption readers may get more questions right, like naming a speaker).
Some video captions could have numerous errors but of a type that the reader can easily tell what was meant. Other videos might have highly accurate captions but one essential word was missed, changing the whole meaning.
I don't see the similarity. If we go from blind people suffering from not having access to the audio transcription, to everybody suffering from not having access to either the audio or video (because all the videos were taken down), isn't that a lose-lose for everybody? In worker protection at least the workers win from having more safety.
That's the whole point of disability legislation across the world, including the ADA
If I were to build a new shopping centre, but decided not to put in wheelchair access, would you argue that everyone will suffer from not having access to the shops, and thus it's a lose-lose for everyone?
Are wheelchair ramps an example of cutting-edge / frontier technology in this metaphor? Shouldn't society catch up a little before announcing that a technology becomes a standard? Why shouldn't an advanced AI Siri guide any blind or deaf person through any institution, store, or software experience?
In this situation we are in fact saying that a large volume of user-submitted or professional porn should not be accessible because it might not be worth the money to transcribe. I don't think any stores are closing because a mall had to add wheelchair ramps. It's not like an inspiring technological undertaking.
Relevant precedent: Harvard recently settled with National Association of the Deaf, resulting in all videos made publically available after Dec 2019 requiring captions, and a required 5 day turnaround for captioning of older videos.
https://dredf.org/2019/11/27/landmark-settlement-with-harvar...
That sounds positive. In contrast, a similar situation at Berkeley had a disastrous outcome -- all Berkeley's historical non-captioned videos of lectures were taken down.
> That sounds positive. In contrast, a similar situation at Berkeley had a disastrous outcome -- all Berkeley's historical non-captioned videos of lectures were taken down.
It's worth re-iterating that it didn't have to happen this way: Berkeley could have captioned the videos; but there's no indication that they seriously considered this option. In fact, I cannot find any indication that Berkeley tried to find any sort of compromise with the DOJ - despite the letter from the DOJ strongly urging Berkeley to work with the department towards a solution.
The DOJ letter is also worth a read, in that they find the management of UC Berkeley did not seriously attempt to enforce any sort of compliance with accessibility standards.
It would have been expensive, sure. But it's wise to keep in mind that the UC system operates with a yearly budget exceeding nine billion dollars - and as much as we want to worry about the cost of transcribing those old courses, at the end of the day it is a drop in the bucket of their overall expenditures.
I find their CC to be pretty accurate most of the time with sometimes some words misunderstood. It's not perfect, but for me it often is very usable as aid in a noisy environment (mind you I have no hearing disability, just don't want to crank the volume to 11 when my kids are playing in the same room).
TBH, even human transcribers have their limits. I often get transcripts of my podcasts. But if I have someone with a strong accent, I’ll often skip because I know I’ll be charged for a difficult transcription especially around technical terms and I’ll still have to do a lot of cleanup.
Surely the ADA requirement isn’t for a near perfect transcript.
That actually makes me wonder re podcasts and other audio. Is there any reason they’d have a different requirement or are these lawsuits specifically about video for some reason?
> Surely the ADA requirement isn’t for a near perfect transcript.
The law requires "reasonable accommodations" to provide a equivalent experience. Expecting perfection would not be reasonable. Also with heavy accents, the hearing audience probably doesn't understand every word so if some words in the transcript were wrong, the hearing impaired have the same experience.
There's nothing special about video, the law can also require podcasts and other audio-only content to provide transcripts. But the law doesn't apply to everything and everybody, it applies to "places of public accommodation," and many podcasts are personal projects, to which it doesn't apply. The legal precedent for what is a place of public accommodation is evolving, even including anything online is not universally held, I don't know what counts; having an ad spot in a podcast is probably not sufficient to suddenly make it a business to which the law applies. It almost certainly applies to an operation like Gimlet.
And it’s possibly complicated by the fact that many podcasts don’t even carry advertising but are clearly content marketing for businesses. Mind you for SEO and other reasons, transcripts are often a good idea but probably the vast majority of even quasi-pro podcasts don’t do them.
The phonetic alphabet would probably be helpful here, given the "text" material. Quite a challenge. But since this is for pornography, I do not doubt it will exist in a short amount of time.
Honest question here
If you are building a system that is focused on a specific audience, i.e you're building a video site not intended for those with disabilities, how does that work?
Can you be forced to provide additional content outside of your systems intended use case?
We build sites to meet WCAG standards when doing Government work, but for other sites that are personal/startup projects do these same rules apply and if so why do they apply?
As a somewhat contrived example, are Spotify required to provide captions/lyrics with each song they stream?
> If you are building a system that is focused on a specific audience, i.e you're building a video site not intended for those with disabilities
That in itself would be discriminatory, excluding people belonging to a protected class (i.e. people with disabilities).
The law does not apply to personal sites, it applies to "places of public accommodation." Most court case outcomes have agreed that it's not only physical places, websites and apps of businesses are included.
A private club might be exempt but I'm sure there are limitations on what can be a private club; i.e. Dollar Shave Club can't call themselves a private club so their website can be inaccessible to people with disabilities.
> As a somewhat contrived example, are Spotify required to provide captions/lyrics with each song they stream?
Probably lyrics, which are basically a transcript, they wouldn't have to be synchronized to the music, as captions are. The law requires "reasonable accommodations" to be made and lyrics to songs are readily available from the rights holders; if those rights holders wanted a lot more money for the lyrics along with the songs, that could be an undue burden on Spotify and mean they don't have to do it.
I thought Pornhub hosted videos uploaded by users. If they're just a platform, they shouldn't be liable for the lack of captions. The article mentions there is a section for captioned videos (though I wonder if they're actually subtitles translating a language, not captions for the hearing-impaired) so I assume their video player supports caption files.
This is funny (who cares about what do they speak in porn, really? I know some people learn German just to understand that but that's hardly essential for the experience) but there have been more alike case which were sad: universities removing (!) freely-available lecture videos as there were no subtitles (even those published on YouTube, I don't know why generated subtitles wouldn't qualify).
And here I thought it was Swedish - https://www.youtube.com/watch?v=zPf3kO-p7hc (link to a clip from "Stripes" where John Candy's character speaks some Swedish, learned from watching porn).
In the past German porn used to be the only (or the only reasonable) porn many people could get. Some people felt curious about what do the people in the porn speak (which seems natural when all you have is a couple of VHS movies rather than tons of short clips you can navigate in instants), some of them had too much spare time (that was during the days before widespread Internet availability).
Perhaps they do but we can't see that because PornHub doesn't have the full movies. E.g. I've actually done both the jobs once when I was a [real] cable worker.
Curious: why isn't all porn sites sued for this? I don't quite get why they're only going after PH? Is it because it's harder to win a case against 1000 websites? Is it because it's more overhead to go after multiple sites and they expect the biggest return on investment only going after the biggest one? Anyone smarter than me who has a take?
If "doing the right thing" means banning videos from all online platforms except those with the capability to automatically translate sound into subtitles, I'd rather they don't do the "right thing". The end result will be lawsuit money for a few disabled people and the end of most alternative video platforms.
It must be because I have worked with advocates for disabled people. I don't see the funny aspect. I'm glad for the laws. Either they apply or they do not. Either our hearing-disabled brothers and sisters are equal or they are not. It's juvenile to somehow factor porn out and say it shouldn't count there.
They are obviously not equal. They are hearing disabled. That's a red herring.
Should the society support them to lead normal lives as much as possible? We can pretty much agree on that, at least as long as the costs are reasonable (for example one billion per inner ear implant is definitely not).
If you put the problem like this it's obvious there is room for discussion. Is PH captioning helping them live normal lives? Is the cost reasonable? Is it really the best direction for society to put resources?
People will answer differently, of course, not being identical copies. But the conversation itself is legitimate.
Honestly I'd say you accusing the parent of a red herring is pretty much a red herring, given that the parent is very aware that the people in question are deaf.
Your resource argument doesn't hold up either, especially with money. Things can become cheap if society values them enough to invest time to make them cheap. Most of that time is currently being invested in making middle- and upper-class people's lives more comfortable (eg most unicorns out of SF) but that doesn't mean it _has_ to be that way.
Drugs could be cheaper (or not become absurdly expensive), places could be accessible, people wouldn't have to be homeless... It's clearly a resource allocation problem, but you make it sound like resources are very well allocated right now. I don't think they are.
The problem with this kind of regulation is that it stifles entrepreneurship with unfair financial burden. If the quality level of YouTube auto-generated CCs is considered passable, then don't you think tax money would be better invested in developing client-side captioning software that would be universal and work everywhere ?
Yeah, but what about the case that @threatofrain mentioned? If a company has no resources to auto-generate CCs for every video uploaded, then it should be punished and sued? (Of course this case does not apply to PH)
It's a very bad thing if in real life, everyone has to be in illegality in order to function. This is when the rule of law loses meaning and becomes a mere weapon to be used by those in power.
