back to article Audible hasn't even launched its AI-powered book subtitles and publishers have already fired off a sueball

Book publishers are suing audio book specialist Audible for using machine learning technology that automatically transcribes audiobooks to text, claiming that this violates copyright law. The lawsuit [PDF} was filed by seven top US publishing companies, including Penguin Random House, HarperCollins Publishers, and Macmillan …

  1. veti Silver badge

    Old-skool publishers vs Amazon

    Part of me wants to see both sides lose this, but only because it's good to see both of them lose everything.

    Another part is wondering when the companies that call themselves "publishers" are going to wake up to the fact that their business model is broken, and Amazon is one of very few companies who may have both the capability and the incentive to help them rebuild it.

    1. Psmo

      Re: Old-skool publishers vs Amazon

      Amazon is one of very few companies who may have both the capability and the incentive to help them rebuild it hit them over the head and nick their wallet.

      Amazon haven't become one of the richest distributors in the world by helping people.

      Even the large publishers have learnt to be very careful dealing with Amazon.

      1. veti Silver badge

        Re: Old-skool publishers vs Amazon

        It's not Amazon who's eaten their lunch. Amazon's business model depends on a reasonably healthy publishing industry.

        The danger to those publishers is from the new publishers, Facebook and Google. Amazon could be an ally, and they certainly need one.

        1. Ian Michael Gumby

          Re: Old-skool publishers vs Amazon

          It's not Amazon who's eaten their lunch. Amazon's business model depends on a reasonably healthy publishing industry.

          That is not true.

          Amazon no longer needs publishers.

    2. big_D Silver badge

      Re: Old-skool publishers vs Amazon

      I would say that they have woken up, they've been cheated and dumped upon by Amazon all to often. I'm not saying the publishing industry are saints, but Amazon is no saviour, it is the devourer.

      1. veti Silver badge

        Re: Old-skool publishers vs Amazon

        The publishing industry deserves zero sympathy. I began my career in that business, so I've watched with interest their shenanigans over the past 20 years (since I got out),and if ever a bunch of companies deserved to go to the wall, they do.

    3. Ken Moorhouse Silver badge

      Re: Old-skool publishers vs Amazon

      I wonder what would have happened if the Net Book Agreement were still in force. Not just this scenario, but the whole publishing industry.

  2. Kernel

    I'm having trouble seeing what the publishers are complaining about - from reading the article I get the impression that the text is only created and viewable by the listener while the same phrase or sentence is being spoken, in which case the text is as transitory as the audio component.

    Surely the publishers should have raised their objections at the stage when audio books were first created.

    1. Charles 9 Silver badge

      No, because audio book publishers ink deals for those audiobooks. To the publishers, Amazon's stunt smacks of on-the-fly photocopying which is a foul. After all, if one wanted to read the text, why not get the text version?

      1. MiguelC Silver badge

        Is speech-to-text that much different from text-to-speech? Did anybody ever got sued about the latter?

        (note that I'm not defending Amazon, just questioning the legal action's logic)

      2. John Robson Silver badge

        Maybe because you have a mix of visual and auditory disabilities that make doing either exclusively hard work - but the combination much easier...

        I'd welcome this, but I'd want the reading speed to be 'tuneable'

        1. Charles 9 Silver badge

          Amazon takes that into consideration (such as for people wanting to keep up with stories in a car or whatever). Buy one format, and the other format is offered at discount, encouraging bundling.

    2. Roland6 Silver badge

      I wonder when Amazon will do the same with streaming music...

  3. Mr Booth

    Just ask first :)

    I work in an organisation that deals with publishers and copyrights all the time. We rarely have problems because we let our publishers, and narrators know what we are doing in advance. This is good because:

    1. We use the feedback to decide if something is worth pursuing further before putting more time and money into a project

    2. Any objections can be raised and subsequently worked through before the lawyers get involved; if it looks like lawyers are going to be involved, we can then go back to step 1 above.

    3. Sometimes we get really good feedback and incorporate that into our planning.

    4. Everyone appreciates being kept in the loop and over time it builds good business relationships.

    It seems like tech companies these days are just doing things..."because we can" and not really thinking things through. Is this an Agile thing or just a Lack of Common Sense thing?

    At least there should be plenty of egg to go around. Oh, and lawyers too.

    1. big_D Silver badge

      Re: Just ask first :)

      I think it is an arrogance thing.

  4. Forget It

    Why don't Audible just *write* the books on the fly - if they're so d**** clever?

  5. Neil Barnes Silver badge

    Have I got this right?

    There are complaints that software is converting audio conversions of written text to, er, written text?

    Isn't that how they started out in the first place? And is it in any way different from e.g. automated subtitles on broadcast TV or youtube and friends?

    Maybe it's me, but the whole thing sounds like 'damn, I wish we'd thought of that'.

    (Personally, I dislike 'audio books' but I would find this handy to help with a foreign language - one of the biggest difficulties I have is separating spoken words).

    1. Roland6 Silver badge

      Re: Have I got this right?

      >And is it in any way different from e.g. automated subtitles on broadcast TV or youtube and friends?

      Err no very different.

      What disability is Audible's tech seeking to address that can't be addressed by buying a text version in the first place?

      1. Anonymous Coward
        Anonymous Coward

        Re: Have I got this right?

        Wow. Thanks for forcing people to not share things together? Want to listen to a book with a deaf friend? Not allowed. Illegal. Go to court.

        Nice one.

        1. Roland6 Silver badge

          Re: Have I got this right?

          >Want to listen to a book with a deaf friend?

          My deaf cousin, loves modern TV etc. - because of the audio description. Together we watch many of the films and shows (eg. A-team) from his youth, that at the time didn't have audio description - to the amusement of my teenage children, because now he is able to understand the plot and story.

          However, we do have a real problem, his first language is BSL, not English. Which leads at times to much fun particularly when I ask him to try and explain a deaf ie. signed joke...

          1. Robert Carnegie Silver badge

            Re: Have I got this right?

            I think you don't mean "audio description", because that is a voice explaining what is happening on screen, for blind people. I presume you mean subtitles ("closed caption") or on-screen sign language.

            Your friend might be interested in this week's "Fry's English Delight" with Stephen Fry, but utterly frustrated because it's a programme on radio. Except that there is an English transcript (this time only, I suspect). Both are here

            Anyway, it's about the history of sign language, and how it is nothing to do with English at all actually. Apparently this is Series 10 so a lot of subjects have been done, and at the same time "Word of Mouth" and "The Verb" are mowing the same lawn... you see we even run out of metaphors.

            According to this, if you have a deaf friend then you have a sign language nickname. Donald Trump has a celebrity one. Madonna's is intriguing because I would have thought she was famous enough to have one already before... that...

    2. big_D Silver badge

      Re: Have I got this right?

      The text version already exists, so there is no need to transcribe the audio and, if the user wants to see the text, they just need to buy the book at the discounted rate (with my Amazon/Audible account, I can usually get the book/audiobook pair at a big discount).

      Subtitles? But that is for film for the hard of hearing. If you are hard of hearing, you are unlikely to have bought the audiobook in the first place.

      I agree with you about the foreign language part, but, again, I generally look it up in the written version of the book, if I really can't understand it after repeatedly listening to the sentence - or ask a friend.

      1. James 51

        Re: Have I got this right?

        If you're hard of hearing it can help you to follow what is being said (and for all the people commenting, why are you listening to an audio book then, sometimes it is the only format available (see David Eddings books currently) and sometimes narrators like Stephen Fry can elevate the source material through their narration (see the stupendous expensive Harry Potter audio books, available from your library on line (at least in the UK)). It is also useful for learning other languages.

        1. big_D Silver badge

          Re: Have I got this right?

          Here the Harry Potter books are read by Rufus Beck, also a great voice actor and really good. I'm currently re-reading as audio books - just finished Harry Portter und der Orden des Phönix and with my next credit I'll be getting der Halbblutprinz.

          I currently have 280 titles in my Audible library - a mixture of German and English versions. I am a big Audible fan, but in this case, I don't see this as good move on their part.

          With the subscription model of Audible (9,99 a month for a credit and all subsequent audiobooks 9,99 or cheaper, I feel that the listeners are already getting a very good deal, compared to the traditional prices for audiobooks. Amazon/Audible already have a good solution to this problem (Whispersync), I don't see how disadvantaging authors and publishers further is a good thing.

          1. big_D Silver badge

            Re: Have I got this right?

            Just to add, that they should have agreed this with the publishers before doing it. It is obvious that this will rub the publishers up the wrong way, if they just present it as a fait accompli.

            The idea has some merit, and as I note, they already have half of this solution, with publisher approval, in WhisperSync. Why not hammer out how to do this with them first, it would save time, effort and money on lawyers.

          2. John Robson Silver badge

            Re: Have I got this right?

            " I don't see how disadvantaging authors and publishers further is a good thing."

            And how is adding accessibility to books "disadvantaging" the publishers?

  6. VonDutch

    Reinventing the wheel

    The source material for an audio book is a book... Why go through machine learning to generate subtitles?

    Just strike a deal with the publisher to supply the audio book and the ebook in one package and you can read along!

    1. big_D Silver badge

      Re: Reinventing the wheel

      They already do that, to a degree, with Whispersync. If you buy the ebook you can get the audiobook at a discount or vice-versa. When listening to the book, it automatically updates the position in the ebook - although I believe not dynamically, I've never tried that, I generally listen to the book in the car, then switch to my Kindle when I'm home. If they made that update dynamic, there wouldn't be any problems.

    2. DwarfPants

      Re: Reinventing the wheel

      Whats the point? (apart from circumventing some legal restriction). The reason you are using an audio book is because you cannot or don't want to read the text. It just sounds like putting a book through google translate a couple of times, which could result in some interesting/confusing plot deviations.

    3. John Robson Silver badge

      Re: Reinventing the wheel

      Quite often audiobooks are not perfect verbatim replicas if the written word.

      Particularly if a book references itself, or the narrator talks to the reader - translated to listener...

  7. big_D Silver badge

    Why AI transcription?

    Amazon/Audible usually have both the orignal paper/e-book and the audio version, surely they just need to provide the text from the book and let the AI match the spoken word to the actual text...

    Oh, wait, that would be blatantly breaking copyright, by providing the actual text. Letting the AI listen and transcribe gives this the halo of respectability.

    As a long-time Audible subscriber, I have to say I find this idea appalling and appealing. This is just another example of big tech trying to circumvent the law. If I want the text to an Audiobook, I can generally buy the text copy at a discount (or vice-versa - Whispersync). Whilst the service is interesting and for hard to understand passages useful, I think this is going too far. At the end of the day, the book is the result of somebody's hard labour (plus editors etc.) and, likewise, the audiobook is the hard work of a voice artist and production team.

    1. Psmo

      Re: Why AI transcription?

      Off the top of my head:

      • Because an AI transcription is a derivative work and so may be covered by the license already paid if not explicitly forbidden.
      • Because they are working on their speech-to-text models and need the results evaluated.

      1. Roland6 Silver badge

        Re: Why AI transcription?

        Probably also, because it provides Amazon with a way around Nuances various speech-to-text and semantic analysis patents...

    2. Anonymous Coward
      Anonymous Coward

      Re: Why AI transcription?

      I love everyone defending *format shifting* being illegal. The arguments for it are amazing.

      I'll be ready to tell you how stupid or broken against the law or "freeloaders" you all are, when someone takes your use case to court because "you are not allowed to run on a VM/backwards compatibility/etc" because you should have purchased the _new_ format version.

      ***An audio book has paid for both the encoding (story) and the art/performance (reader/speaker) in royalties AND wages to both the author and the performer***

      And these people want to double dip and charge a second time for the text/encoding just because someone hard of hearing wishes a little assistance in understanding was that "one time they won" or "they won one time" because of hearing/dyslexia?

      All of you, perhaps one day you'll get isolated, and told "you're illegal" and know how it feels.

      1. big_D Silver badge

        Re: Why AI transcription?

        But, again, Amazon/Audible have both the original text and the audiobook, they don't need to transcribe, they just need to enhance Whispersync.

        And why not approach the publishers and try and work out a good way of doing this, instead of trying to circumvent agreements and the law?

        If big tech put as much effort into doing things properly in the first place, as opposed to spending the effort in trying to find new ways to circumvent the law, they'd have to pay lawyers much less and they wouldn't have to keep re-doing everything because they got caught.

      2. Roland6 Silver badge

        Re: Why AI transcription?

        >was that "one time they won" or "they won one time"

        It's an AI, can you be sure that whatever the audio says, is actually what the words say...

        I can see this being useful in screening content - eg. U rated 50 Shades of Grey?

      3. Not also known as SC

        Re: Why AI transcription?

        Format shifting.

        ***An audio book has paid for both the encoding (story) and the art/performance (reader/speaker) in royalties AND wages to both the author and the performer***

        So by the same argument - I have a copy of Phantom of the Opera on CD. Does that give me permission to see the musical free of charge after all, the writer has already been paid?

        1. John Robson Silver badge

          Re: Why AI transcription?

          No, but seeing the performance I expect certain accommodations for disability.

          When I went to the Sydney opera house they had live subtitling (sur-titling, the words were above the stage).

          That didn’t somehow reduce the earnings of the writers, composers. Producers, directors, actors... it enabled a lot more people to enjoy the performance...

  8. lglethal Silver badge

    Am I understanding the objection here? it seems that the publishers are saying that by creating a transcription Audible are denying them the sale of a written version of the book. Although the owner of the book has already paid for the book once (via the audiobook purchase), so they have already got a sale. Are they upset they're not getting a second sale? Sounds a bit greedy.

    Still as someone who uses audio books (as well as regular books), I cant really see why you would want a transcription of an audio book. I mean having an audio book is great if you're busy doing other stuff with your hands and eyes, but can still listen and concentrate on the story. It's also of great help for those with sight problems. If you have hearing problems then, you're not going to buy an audiobook in the first place. And if you're listening to an audio book and suddenly have the ability to sit down and read instead, well I find it unlikely you're going to want to sit down and pick up the text version of what you were listening to and start reading, as part of the joy of an audible book is in the narrators performance. It can really set the tone for a book. Switching back and forth would give me a very large cognitive dissonace.

    So i really cant see for who's benefit this is. Still as another commentard above mentions this sounds very much like Audible never bothered to speak to the publishers in the first place, and hasnt really engaged with them throughout, so naturally the publishers are annoyed and so are fighting this. Rule number one of business, maintain good relationships with your suppliers, seems to be something that once a company gets past a certain size they seem to forget or simply choose to ignore...

    1. big_D Silver badge

      I agree with you for the most part. With Whispersync, you can already buy both the audio and ebook versions at a discount and carry on on the other medium from where you left off - E.g. I sometimes listen to an audiobook in the car, then switch to the ebook when I go to bed, so I don't disturb my wife; although I generally listen to a different book to the one I'm reading in bed.

    2. Anonymous Coward
      Anonymous Coward

      It is double dipping. I often have to use subtitles on films/tv because of the audio, or just plain old language ambiguity or sound/word similarities.

      I don't listen to many audio books, but when do, I often follow text.

      So this would both force me to pay twice, and suggest I'm "illegal", for having an additional need *after paying for the content already*.

      My solution? I won't actually consume your content if you treat customers that way. I have no qualms over not watching/reading/listening to your media, and thus not paying.

      1. big_D Silver badge

        This is more about Amazon/Audible finding an acceptable way of doing this.

    3. John Robson Silver badge

      " It's also of great help for those with sight problems. If you have hearing problems then, you're not going to buy an audiobook in the first place"

      Assuming that everyone has either perfect hearing or perfect vision, or that everyone with hearing/sight loss has total loss in that department is rather dubious. Disabilities are very rarely binary.

      I can read books, but it get very tiring due to my visual disability, similarly I can listen to audiobooks (assuming they're streamable directly into my hearing aids) but it gets tiring.

      Combine the two and the additional processing my brain needs to do is vastly reduced -> result - more reading (and therefore more sales of books)

      Note - I'm just one person, with one peculiar set of disabilities. As has been pointed out elsewhere this would be really good for learning another language - and that's a whole lot more people.

      Given that the text is available the 'on the fly speech -> text' seems pointless it's not as if we don't have a perfectly good subtitle file format already (sub or ass, I don't mind) - feed the ML system with the book and the audio and it can probably autogenerate a very good subtitle file - which could be easily distributed with the audiobook.

      Now if they are trying to patent the idea of subtitles then that's another thing entirely.

  9. RichardB

    Clearly discriminatory here

    If the technology exists to make a website or technical product accessible it seems reasonable to expect it to be so under both UK and US laws.

    IANAL, but, surely if this is on their website and is distributed by their website for playback in their website or apps, then surely they have a responsibility to try to the best of their ability to comply with anti disability legislation?

    I note that currently Dominos is under fire in the US due to it's 'bricks and mortar' store presence mandating web compliance. Amazon is rumoured to be opening physical stores, so perhaps is equally tied?

    Certainly it would seem sensible for their most important website - the UK one, to be compliant with the reasonable expectation of accessibility?

    1. Anonymous Coward
      Anonymous Coward

      Re: Already paid.

      But that's the thing. Even if it did prevent a "second sale", it's a *second sale*. The original sale already had the content. This is format shifting, not "copying".

      It's no more "copying" than a TV being analog tube vs playing on a Digital LCD "copies" the VHS from tape to silicon.

      The spoken words, already contain the entire text, but the automation (or even non-automation) would have transcribed that back to text.

      It would be akin to suing someone over putting visual descriptions of a movie on a DVD... the DVD already has the visuals included and paid for, so transcribing "there is a mountain with a woman singing "over the hills"" breaks no copywrite!

      1. Charles 9 Silver badge

        Re: Already paid.

        A format shift is still illegal if there is a contract out for the destination format ALREADY. And there is: the PRINT and EBOOK contracts: otherwise, they wouldn't be available on Amazon. If Amazon wishes to provide subtitles, they need to negotiate with the holder of the eBook rights and negotiate first.

        1. Charlie Clark Silver badge

          Re: Already paid.

          Just do a thought experiment with what you describe and apply it to close-captioning for a film, that doesn't include it.

          1. Charles 9 Silver badge

            Re: Already paid.

            Easy peasy. CC/SDH is packaged with the film and considered part of the whole. It doesn't exist separately from the film.

            Audiobooks, OTOH, are derivatives of the original text (either print or ebook), and the AI is meant to transcribe from audio to text: a format which already exists and thus falls foul of the existing copyright on the original text.

            1. John Robson Silver badge

              Re: Already paid.

              No - the CC isn't 'packaged with the film' in all cases.

              Audiobooks are performances of a written work, in the same way that the sound of music film is a performance of that work.

              They deserve their own protection, but the protection of the work does not extend to preventing people adding accessibility features (audio description, subtitles) to that performance.

              1. Charles 9 Silver badge

                Re: Already paid.

                Unless said feature has ALREADY been produced (in this case, the text version).

                1. John Robson Silver badge

                  Re: Already paid.

                  You can buy the libretto to sound of music - so no, the existence of the book doesn’t preclude subtitling.

                  It’s making a specific performance accessible - not being a book.

    2. Charlie Clark Silver badge

      Re: Clearly discriminatory here

      I think courts are likely to follow that argument and, as long as Amazon can demonstrate that the results cannot be accessed separately, then I think they'll be okay.

      As for the idea that transcription is only possible for companies with deep pockets, that's hogwash. Lots of the research on has been publicly funded for assistive technologies as good models are avaiable. If publishers wanted to, they could easily provide their own versions.

      I'm no fan of Amazon (love my Kobo Aura One) but Audible and were excellent buys for the country getting into digitall services, and there's no doubt that Audible has brought audiobooks to a much wider audience.

  10. Tigra 07

    This isn't the first time Amazon has fought with the publishing industry...

    Isn't listening to an audio book, while reading the captions, just watching TV with extra steps?

  11. DrXym Silver badge

    Hang on...

    Since I don't use Audible I don't know what content they provide so perhaps they also offer performances, blogs, streamcasts or whatnot.

    But why would a deaf person want closed captions for an audio version of a normal print book? Why can't they just read the book? I would have thought the experience is far superior than a machine translated (i.e. garbled) transcription of a condensed audio book.

    Or perhaps this is a red herring and audible and trying to put search functions into their books or similar and the publishers are a bit odd about that.

    1. Anonymous Coward
      Anonymous Coward

      Re: Hang on...

      Not everyone is 100% deaf. So having no way to check a single word/sentence when 99% of the rest of the audiobook is fine, is very very discriminatory.

      Almost predatory and attacking the user.

      Imagine if you could only read slowly. Then imagine if Reg or their advertisers got so Copyright fearful to protect their "content" and money, they forced the webpage to close after 10 seconds. That stops people transcribing the website and stealing the content... but then prevents people reading it.

      I literally left websites that had autoscroll and autoplay and auto skip to next page (BBC news was one for a while). Why? Because I may decide to wait to read that page. Skipping it is both annoying and discriminatory (I don't have a screen reader, but know others do).

      Likewise, the same applies here. Forcing others to purchase twice, when they already purchased the *content* is unfair and malice of a practice.

      Like charging you once for the program on CD, then asking for payment again, when you transfer it to HDD to run the program.

      1. Roland6 Silver badge

        Re: Hang on...

        >Like charging you once for the program on CD, then asking for payment again, when you transfer it to HDD to run the program.

        But if that 'program' was music or a film...

  12. davenewman

    On another site discussing this, they explain that do don't get the full text of what is being said. From time to time you get an automatic transcription of the speech that you can use as a place marker when navigating. It is not continuous.

  13. Dan McIntyre

    Deaf People and Audio Books

    I've read a couple of comments here along the lines of "deaf people won't buy audio books".

    As a Deaf person, and a long time Audible subscriber, I can assure you that is nonsense. Deafness is more of a spectrum than a simple yes-you-can or no-you-can't and the vast majority of Deaf people do have some hearing ability.

    For myself I use 2 hearing aids, but I do enjoy listening to audio books whilst in the shower or on a long car journey.

    Just saying...

  14. Moosh

    Duck Duck Goose

    "The publishing giants believe that rolling out the Audible Captions feature gives the Amazon-owned company a competitive advantage against other audiobook providers, who do not have the technical resources and prowess to utilize speech-to-text technology."

    Irrelevant and in no way against the law

    "It unfairly slashes publishers’ profits"

    Also irrelevant and not against the law

    "it illegally reproduces and distributes books"

    The only actual argument they have, which is a hard sell unless audible actually allow you to download the transcripts.

  15. Anonymous Coward
    Anonymous Coward

    Amazon / Audible

    Are in the wrong here, imho. They should simply have licensed the text for the use from the copyright holders. And does anyone want on the fly AI approximations to the original text, with the ludicrous errors that will be larded through?

  16. ratfox

    I'm with audible on this. I find supremely annoying the copyright owners who insist that any new use of their IP should be a brand new source of revenue.

    And if users were to play an audiobook in front of a speech-to-text machine, they'd probably sue the users, claiming they only bought the license to listen, and not to read. And not in that country. And not on a balcony where passersby could catch a glimpse. Etc. Etc.

  17. Anthidote

    Maybe I'm an idiot but

    Isn't a close captioned audio-book just a you know ... book

  18. Anonymous Coward
    Anonymous Coward

    I think this is hard to find out a good solution for this question. We want to protect the publisher but Amazon wants efficiency. This is hard for me. This is just like I want to listen to Audible on my smartwatch but Audible does not support it. So I have to convert Audible to MP3 with TunesKit Audible AA/AAX Converter.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2022