Listen to the latest articles and insights from our experts.
Listen to the latest articles and insights from our experts.
Book publishers will never abandon human-written books; it’s their entire reason for being.
Although new Artificial Intelligence tools like ChatGPT write impressive text – and probably could even write book-length works – the output is simply content that the engine has scraped, analyzed, and rehashed. In the best case, an AI-produced work is inherently unoriginal. In the worst case, it may repeat someone else’s identifiable, copyrighted, or flawed work.
Yet even these tools have their uses, not for creating books but for promoting them. Book editors, marketers, and publicists can use machine learning tools to analyze their books’ text and write descriptions for their book catalogs, media kits, executive summaries, and abstracts.
Additionally, machine learning may be ideal for editors who want to examine authors’ manuscripts for plagiarism, check their citations, and even reverse-engineer text to ensure that it wasn’t authored by another AI platform.
The audiobook market is booming thanks to distributors like industry pioneer Audible, digital jukebox Spotify, and library distributor OverDrive. This is true not only in English-language markets but also in the Nordic countries, and other markets are likely to follow suit in the coming years.
But this market is dependent on a steady supply of new audiobook titles, which are expensive to produce thanks to the complex production process: professional voice actors, recording studios, and skilled post-production editors.
Digital Text To Speech (TTS) voices have the potential to help fill this void. TTS technology has improved to the point that digital voices are often indistinguishable from human narrators. Companies like Deepzen offer digital production and editing services that will even customize each digitally produced audiobook with custom pronunciations and inflections. While it’s still not the same as having a human actor interpret each text, it’s getting mighty close.
Expect publishers to spend money for actors only for their more important titles (those that they expect to recoup the substantial audio production costs) and employ high-quality TTS for other books, especially nonfiction titles. There may even be new licensing models that allow the listening platforms to apply TTS voices to ebooks, which would massively expand the available catalog of content.
The new European Disability Act (2022) requires all EU member states to pass and enforce updated accessibility laws by 2025. At that point, all ebooks sold in the European Union must be accessible to users with disabilities.
Fortunately, the industry-standard EUPB reading format has allowed publishers to offer “born accessible” ebooks for the past decade. But even these titles may have problems, though new technology can offer a solution. Machine learning has the potential to identify accessibility issues like improperly tagged heading structures and images that are missing alt text. Image-recognition tools could even write alt text, with appropriate human editorial oversight.
Meanwhile, publishers who have PDF-based titles (especially academic and scholarly publishers) may need to undertake extra measures to ensure the accessibility of their titles. Regardless of the format, publishers also should update their titles’ metadata to reflect how accessible their titles are.
The new European Disability Act (2022) requires all EU member states to pass and enforce updated accessibility laws by 2025. At that point, all ebooks sold in the European Union must be accessible to users with disabilities.
Fortunately, the industry-standard EPUB reading format has allowed publishers to offer “born accessible” ebooks for the past decade. But even these titles may have problems, though new technology can offer a solution. Machine learning has the potential to identify accessibility issues like improperly tagged heading structures and images that are missing alt text. Image-recognition tools could even write alt text, with appropriate human editorial oversight.
Meanwhile, publishers who have PDF-based titles (especially academic and scholarly publishers) may need to undertake extra measures to ensure the accessibility of their titles. Regardless of the format, publishers also should update their titles’ metadata to reflect how accessible their titles are.
The new European Disability Act (2022) requires all EU member states to pass and enforce updated accessibility laws by 2025. At that point, all ebooks sold in the European Union must be accessible to users with disabilities.
Fortunately, the industry-standard EPUB reading format has allowed publishers to offer “born accessible” ebooks for the past decade. But even these titles may have problems, though new technology can offer a solution. Machine learning has the potential to identify accessibility issues like improperly tagged heading structures and images that are missing alt text. Image-recognition tools could even write alt text, with appropriate human editorial oversight.
Meanwhile, publishers who have PDF-based titles (especially academic and scholarly publishers) may need to undertake extra measures to ensure the accessibility of their titles. Regardless of the format, publishers also should update their titles’ metadata to reflect how accessible their titles are.
The new European Disability Act (2022) requires all EU member states to pass and enforce updated accessibility laws by 2025. At that point, all ebooks sold in the European Union must be accessible to users with disabilities.
Fortunately, the industry-standard EPUB reading format has allowed publishers to offer “born accessible” ebooks for the past decade. But even these titles may have problems, though new technology can offer a solution. Machine learning has the potential to identify accessibility issues like improperly tagged heading structures and images that are missing alt text. Image-recognition tools could even write alt text, with appropriate human editorial oversight.
Meanwhile, publishers who have PDF-based titles (especially academic and scholarly publishers) may need to undertake extra measures to ensure the accessibility of their titles. Regardless of the format, publishers also should update their titles’ metadata to reflect how accessible their titles are.
The new European Disability Act (2022) requires all EU member states to pass and enforce updated accessibility laws by 2025. At that point, all ebooks sold in the European Union must be accessible to users with disabilities.
Fortunately, the industry-standard EPUB reading format has allowed publishers to offer “born accessible” ebooks for the past decade. But even these titles may have problems, though new technology can offer a solution. Machine learning has the potential to identify accessibility issues like improperly tagged heading structures and images that are missing alt text. Image-recognition tools could even write alt text, with appropriate human editorial oversight.
Meanwhile, publishers who have PDF-based titles (especially academic and scholarly publishers) may need to undertake extra measures to ensure the accessibility of their titles. Regardless of the format, publishers also should update their titles’ metadata to reflect how accessible their titles are.
