Imagine for one moment if, somehow, the Bertelsmann buy-out of Simon & Schuster had been successful and the ever-tuxedoed Count Dohle was still emerging from the grave each night to dictate PRH policy. There would have been no Spotify deal and no unprecedented surge in the fortunes of the audiobook industry.
I’m frequently asked, since I take such an optimistic and forceful view of AI’s future role in the publishing industry, whether I use AI to write the TNPS posts and if I don’t lose sleep over advocating technology that is already costing millions of jobs and will shortly lead to the end of civilisation as we know it.
All of which is a preamble to a review of a powerful post by CITLoB’s Senthil Nathan, who argues that in the field of translation, at least, AI is going to transform the industry for the better, creating jobs as it goes. To which I would expand that argument to say the double whammy of AI audio and AI translation is going to take the industry to new heights hitherto unimaginable, no matter what nonsensical scare tactics the Luddite fringe might invoke.
But first, let’s clear the air around the earlier questions. Is TNPS written using AI?
In my dreams. I’m an old-school journalist and teacher turned book author and publishing industry commentator, that started out as a kid with a clunky manual typewriter and copious quantities of liquid paper to correct the even more copious amounts of typos. For which I blame my school. Girls were taught to type, cook and sew. Boys to become carpenters and welders.
Typewriters Are For Secretaries! Men Don’t Type!
My protests that I wanted to be “a writer” were met with that look of perfected teacherly derision and condescension reserved for those awkward students who think schools should equip kids to make informed choices, not go along with whatever is most convenient for the teacher. Writing is not a job! Typewriters are for secretaries! Men don’t type! Do you seriously want to be in a class full of girls typing letters, while your mates are sanding off a table and four chairs, playing with welding equipment and taking apart a Maserati car engine?
By this time, I was just about old enough to see the merits of being the only boy in a class of girls. But I was as excited by a mortise and tenon joint in woodwork as a butt joint in metalwork, and firmly believed cars were for driving, not fixing. But typing? Cooking? Needlework? Actually I had no problem with any of these, but peer pressure is the enemy of achievement, and to this day I still type with two fingers and cannot thread a needle, although I can open a can of baked beans with the best of them.
But by this time, too, word processing was a thing, although not in schools, and while I was always a step behind with the latest tech hardware, and three steps behind with the software, I religiously followed tech’s progress in authorly things, especially when it came reducing my unhealthy reliance on liquid paper.
A Life-Long Hatred of School
Here to add that I absolutely hated school, and as I came to build a never-ending list of all the actually useful things school failed to teach me, I would later become a teacher just to show them what they were doing wrong.
Not teaching kids to type being one such. Not teaching kids to use a word processor, another. And worst of all, not teaching kids that understanding and embracing changing technology was the key to success in our goals, not the road to unemployment and despair.
Nowadays it’s clear a lot of people in the publishing industry went to the same School of Despair as I did, but whereas I dropped out and forged my own path, they graduated with full Luddite honours.
TNPS and AI
But of course, embracing change doesn’t mean being a prisoner to change. So yes, I still write my own content, but yes too, I do use AI to run a pass on my finished texts to spot the inevitable typos and grammar misses, and lately I’ve been asking AI to throw in some subtitles to aid navigation and readability. Yes, I could write my own, but during my journalism years we had sub-editors paid to do that, so it’s not a habit I ever got into.
I also use AI for research – or more specifically, search (tip: always ask for links to sources to avoid creative responses). And no, I see no evidence that anyone in the sub-headings business, or anyone working in Google’s search engine room, has lost their job because of my callous indifference to their plight.
That said, jobs are lost when things change. That’s the way it’s always been. And nobody cares but those directly impacted, which is why you are reading this on a screen and don’t commute to work on a horse-drawn carriage.
No, wait, some people do care, vocally at least. Those in the soundbite industry, for example, that will be deeply concerned about absolutely anything just to get in the headlines.
A Wave of Fear and Uncertainty Among Translators
The rise of AI in the translation industry has sparked a wave of fear and uncertainty among translators that has been an absolute gift to the soundbite industry. Headlines scream that machines are taking over, leaving many professionals wondering if their careers are doomed, while still others have already lost their jobs and have joined the ranks of candle-makers, gas lamp lighters, horse-drawn carriage drivers, typewriter manufacturers, straw roof thatchers, chimney sweeps and parchment makers in the great jobs graveyard in the sky.
As an industry commentator with no soundbite career to build, I see things differently, and I’m not alone, which brings us to CITLoB’s Senthil Nathan, mentioned above, who recently wrote a compelling defence of AI in the translation industry.
The Double-Barreled Potential of Translation and Audio Working in Tandem
Lately publishing has been obsessing so much about AI and audiobooks that we could be forgiven for forgetting that AI is all-encompassing and not just about audio. More importantly still, it is the double-barreled potential of translation and audio working in tandem that can take the global industry to new heights those in publishing today can barely conceive of.
Consider: we have barely scratched the surface of reproducing existing text-based books to audio, yet already we are reaching new audiences on a scale hitherto unimagined, thanks to Spotify. Or more specifically, thanks to Spotify and the subscription model.
This despite the huge costs involved in real-time, real-people translation, and despite the Luddite resistance from the Soundbite Brigade that have railed against subscription to serve their own narrow interests, and for so many years have delayed audio reaching its current heights.
A Juicy Soundbite of Self-Righteous Indignation
Who could forget (don’t worry, I wont let you!) Nicola Solomon, former CEO of the UK’s Society of Authors, in 2023 telling us so clearly about the “devastating effect” the Spotify deal would have on the industry, not least because it was allegedly giving away audiobooks for free. Never mind the fact that Spotify was doing no such thing. Who cares about mere facts when there’s a juicy soundbite of self-righteous indignation to throw into the arena?
Colossal Misjudgment
Or consider the beyond-colossal misjudgment of PRH CEO Markus Dohle, who in 2021 was telling us, “When it comes to subscription, I am convinced that in the long run it is not good for author income, it is not good for retail.“
More recently, the Count was in Court desperately explaining to a judge that unless PRH was allowed to acquire Simon & Schuster, subscription would cause the complete collapse of the publishing retail sector in as a little as three years.
Imagine for one moment if, somehow, the Bertelsmann buy-out of Simon & Schuster had been successful and the ever-tuxedoed Count Dohle was still emerging from the grave each night to dictate PRH policy. There would have been no Spotify deal and no unprecedented surge in the fortunes of the audiobook industry.
I make that point as a reminder that, while AI offers unparalleled opportunities in the audiobook sector, it has been the embrace of the subscription model, not the use of AI, that has freed the western audiobook industry from the absolute dominance of Amazon’s Audible that Dohle felt so acceptable, and that today has Amazon playing catch-up as Spotify leads the way.
The Double-Barreled Potential of Translation and Audio
But let me return here to the “double-barreled potential of translation and audio working in tandem” I put forward above.
Put simply, audio not only reaches readers who can read but cannot always find the time or opportunity, but also those that love a good story (TV and films are so popular precisely because they tell great stories!) but were failed by the education system and never learned the joy of reading for pleasure.
Overcoming the Barrier of Adult Illiteracy
But it doesn’t stop there. Even in the rich west, adult illiteracy is at worrying levels, and by definition these people are not buying books. But being unable to read is no obstacle to listening to audiobooks, and while being unable to read likely means lower earnings potential, affordable subscription-based audiobooks levels the playing field for audiobook buyers.
But it is in the less-developed world where illiteracy is a way of life for so many, that audio really comes into its own as a game-changer for global publishing.
Shakespeare Wrote Plays to be Performed, Not Read
For us in the rich west that take books for granted, its easy to forget our oral story-telling traditions. We easily forget that Shakespeare wrote plays to be performed in pubic, not read, and least of all in classrooms.
We easily forget that Dickens made his name not so much by being a great author, but by being a great orator, reading his books out loud to eager audiences.
And bringing all these threads together – audiobooks, reading aloud, illiteracy and subscription – let me share this paragraph from World Literature Today:
“Several biographers of Dickens also draw attention to the fact that it was typical for his texts to be read aloud in Victorian England, and thus illiteracy was not an obstacle for reading Dickens. Philip Collins, for example, writes that ‘even some of the illiterate were familiar with [Dickens] through theatrical versions or by attending penny readings,’ while Edgar Johnson also reports that there was an elderly charwoman who could not read but attended every month a tea held by subscription at a snuff shop, where the landlord read [Dickens] aloud.'”
The Game-Changer’s Game-Changer
But it is in the less-developed world,, where books are harder to come by that oral traditions still hold sway, and the potential here for AI-assisted audio to bring to life narratives in literally any and every language, available to anyone with a smartphone, is simply staggering.
Which is why AI-assisted translation is set to become the game-changer’s game-changer.
And this is where I hand over to CITLoB’s Senthil Nathan, who explains in the CITLoB newsletter: “As a language industry expert, I see the transformation differently. AI isn’t the death knell for translators—it’s a call to adapt and thrive.“
“Routine translation tasks like social media posts, press releases, and other short-lived content have already moved into AI’s hands. Platforms like Facebook offer instant translations at the click of a button, making human intervention redundant in these cases.”
Google Translate Stole My Job!
Let me jump in there. When was the last time you hear an industry sound-biter – sorry, spokesperson – complain about translator jobs lost because of instant translation offered on social media (including LinkedIn) and via Google? We probably all use these short-form auto-translate options all the time, despite knowing full well that no-one is being remunerated.
Back to Sentil Nathan:
“Long-term content like books, technical manuals, and business documents still require human oversight for cultural nuance, tone, and precision. Translators are becoming curators, not just creators. The barriers to translation are falling, enabling businesses of all sizes to adopt multilingual strategies. This surge in demand means more work—not less—for skilled professionals who can navigate AI-assisted workflows.”
Let’s hear that again: “More work—not less—for skilled professionals who can navigate AI-assisted workflows.”
Good Enough
It’s an important point. Yes, of course, there will be some operators who will go with an auto-translate rendition to save paying for professional oversight, just as there are some in publishing who will skip editing stages to save money, and hope no-one notices.
And many won’t. Notice, I mean. And we should be absolutely clear on that. Most consumers do not expect, let alone demand, perfection. They just want things to be good enough.
Exceptions are important too, and Sentil Nathan has those covered.
“Critical content like legal, medical, and other high-stakes documents still need human involvement. No company or institution is ready to trust life-or-death decisions or legal risks to an algorithm. Translators in these domains will evolve into hybrid experts, blending human judgment with AI-powered tools to ensure accuracy and reliability. For creative translations, like novels or film scripts, AI is not welcome.”
Hmmm. On that last point I have to disagree. The Soundbite Brigade may not welcome AI, but regular consumers are likely indifferent at best, and will almost certainly warm to the idea as they see prices drop.
A Multilingual, AI-Driven World
But as the OP says, “The real shift is mindset. Move beyond charging by word count. Embrace hourly or project-based pricing that reflects expertise. Expertise will always command value, whether in refining AI outputs, managing complex projects, or navigating cultural nuance.”
Mindset, indeed.
The OP concludes: “Yes, AI is transforming translation—but it’s not the end. It’s a beginning, where translators step up as indispensable partners in a multilingual, AI-driven world. Those who adapt will not just survive; they’ll lead the way.”
“A multilingual, AI-driven world”?
Two points there to wind up this essay.
Multilingual Means Isi-Zulu and Assamese, Not Just English and Spanish
First, multilingual does not here mean English and maybe Spanish and German or French and Dutch. Multilingual means English and French and Dutch and German and Spanish and Hindi and Mandarin and Arabic and Albanian and Lithuanian and Wolof and Mandinka and isi-Zulu and isi-Swahili and Yoruba and Hausa and Punjabi and Pashto and Bangla and Thai and Assamese and…
There are literally thousands of active languages around the globe today, many that are not spoken by enough people to be commercially viable as audio or even text in the traditional publishing world. And some that literally have no written form.
But in every instance, AI translation in combination with AI audio will soon be able to infuse these languages, and their speakers and authors and stories, with new life and new possibilities, that will make the present-day audio market look like a sideshow.
This post first appeared in the TNPS LinkedIn newsletter.
