AI is now narrating audiobooks. Does that bode well for reading?

With the rise of audiobooks, especially at the behest of the pandemic-driven lockdowns, the introduction of artificial intelligence adds a dimension that further reduces human intervention

Imagine listening to a voice that’s impeccably modulated. Every syllable spoken perfectly while the intonation is absolutely flawless? Plus, it's friendly on the pocket!

This is the lodestar for audiobooks of the future, driven by the promise of artificial intelligence. While the rise of audiobooks has been grabbing eyeballs with an estimated sale of US$35 billion by 2030, it is now a frontier that will be conquered by the advent of AI, after writing, coding and creating art. 

CONVENIENCE, THE KEY DRIVER 

What has driven the sale of audiobooks and made them omniscient is their ease of use. It started in trickles and today most articles online have a ‘listen’ option, only proving how commonplace the audio mode has become. The biggest selling point of audiobooks is their convenience and ease in enabling the user to multitask. 

Audiobooks, while being handy, come with their own set of challenges–at times the accent of the narrator might pose a hindrance, it comes at a higher cost, and readers face difficulties with lengthy books or re-reads, as there is no option to skip portions and move ahead.

Ambesh Tiwari, business head, Pratilipi FM and Studios, says a major challenge they have had to deal with as an audio publishing platform is that of speed. “When it comes to reading a book, everyone has their own reading speed, but when it comes to audiobooks, the voices are recorded at a specific speed. While audiobooks are recorded at an ideal speed, it may or may not suit the individual listeners, therefore making engagement tougher.”

This is where AI steps in, with its many uses and dangers. There is no LinkedIn feed without its merits and demerits discussed endlessly. Can the new technology then iron out the imperfections in audiobooks, making them more accessible, cost-optimal and replace the lure of physical books?

THE RISE AND RISE OF AI

Big Tech (Google and Apple) have been the early starters in integrating audiobooks with AI. While Google has personalised AI narrators including Archie and Santiago—who sound British and Spanish respectively—Apple has a catalogue of books narrated by AI across different genres. The format in which people access books has changed over the years and will continue to change with every leap in technology. Publishing firms look at AI as an asset rather than a threat. 

Tara Khandelwal, founder of Bound, a strategic storytelling company, says, “I think there is a role to play when it comes to voiceover acting. If technologies can emulate human voices and are able to emote, that may cut a considerable amount of time and money. We have not used AI yet, but are excited to see where it goes and how it can cut production costs.”

AI is also preferred when there are multiple characters (think of the fantasy genre like Harry Potter or Lord of the Rings) involved in narration which might help bring down costs. Avid reader Prashanth Rao says AI might get good enough in the future if it can replicate the nuances, pauses and emotions of a human voice.  “It shouldn't be monotonous, like the voices used in various navigation apps.” he notes. 

Similarly, the creation of an audiobook is not simply reading the text of the book. Narrators have to immerse themselves within the narrative; almost ‘become’ the characters, adopt their mannerisms, voices, attributes. Author and editor Monica Singh feels that AI will never fill in for human empathy. “With podcasts and audiobooks, there’s a strange sort of intimacy between the narrator and the listener. When a narrator joins hands with the author to bring out the story in vivid detail for the listeners, something extremely unique is created. I am not even sure if that level of intimacy is possible with AI.”

AI can mark a great shift for the future of audiobooks. But not everyone is on board. Tiwari views AI as having multiple uses. He adds, “Currently, AI has been really helpful for recording purposes but gradually, we believe that we can make use of AI to offer more personalised engagement via speed and content discovery. More than using AI to build audiobooks while keeping that human touch/emotion intact in stories, we would see AI be more helpful in discovering and packaging the best suited offerings as per the individual’s taste.”

The AI community though is upbeat. A new entrant, DeepZen turns text into audio without needing one to spend resources on studios and speakers. Others promise AI that can match the emotional resonance of a human. Khandelwal looks at the value-added services. “In terms of voiceover costs, enhancements to the quality and experience, and a lot of the post-production work, AI can definitely help produce the best quality work at a much cheaper price.” she says. 

Singh feels that time will answer if AI-induced audiobooks will replace the physical novel, adding, “For quite a lot of the reading population, ‘listening’ wasn’t the same as ‘reading’. It evolved with time. So, I guess I would embrace AI audiobooks too. Though just like paperbacks and ebooks and Kindle, human narrators would always hold a special place for readers of audiobooks.”

Also Read: Is audio porn the perfect feminist offering?

Also Read: 15 books to keep your summer reading going strong

Also Read: Why does India read more non-fiction than fiction?