Generative AI and Educational Publishing

Generative AI and Educational Publishing

?

Having spent over a decade in the publishing industry, I've seen how the transition to digital communication and systems has brought incremental changes to the processes of developing educational books. I've personally tried my best to craft the perfect prompt (a written instruction to ChatGPT) to generate lesson plans, summaries, multiple-choice questions, blurbs, marketing copy and educational content. In doing so, I’ve seen that these AI models have the potential to significantly enhance book development pipelines, thus enabling even smaller, digitally adept teams to compete with the offerings of industry leaders.

?

Understanding Generative AI

But first, what exactly is generative AI? These systems use large language models, trained on vast amounts of data, to generate text by predicting the likelihood of the next word given a specific context. They are sophisticated prediction machines. Responses are generated using prompts. These detailed instructions combine context, examples, data and restrictions. With advanced language capabilities, large language models (such as ChatGPT, GPT-4, Claude and Bing) can impact how books are written and revised. Generative AI assists authors, editors and proofreaders to automate tasks, structure content and refine text, helping publishers and materials developers to streamline their workflows It's exciting times ahead for publishers!

?

The rapid rise of generative AI in the workplace also means new skills are required. There will likely be a growing need for experts who can work with AI systems to develop and curate content. This may lead to new job roles, such as AI managers and AI strategists, who are tasked with adapting models to inhouse workflows in order to ensure optimal interaction between human expertise and AI-generated materials.

?

Applications in Educational Publishing

Using advanced algorithms and a vast knowledge base, AI systems can generate content on various subjects, helping authors to develop their ideas and fill gaps in their manuscripts. These models become really interesting when publishers provide their datasets, such as specific curriculum knowledge, and a prompt to generate content following a particular tone of voice. AI also has the potential to generate diagrams, charts, videos and interactive content.

?

However, prediction models cannot be assumed to be factually accurate at all times. This challenge undermines a primary goal of a textbook, which is to provide accurate, trustworthy educational content. Ensuring that generated content is factually correct continues to need rigorous checking by human experts to maintain integrity and quality.

?

Textbook authors try to maintain consistency in style and structure to ensure the textbook is readable and easy to follow. Language level is critical to ensure understanding. Generative AI tools can assist in language editing by ensuring the content maintains the desired tone and level. AI can provide suggestions on language use and guidance on how to adapt language level as needed. It can also be used to analyse the overall structure of a textbook, identify areas that need improvement and offer recommendations for restructuring or rephrasing.?

?

Condensing complex ideas into concise explanations can be challenging for authors. Generative AI, with its ability to quickly parse vast amounts of data, can create clear and informative explanatory text and summaries. This can be particularly beneficial when preparing abstracts or chapter summaries to highlight critical concepts within a textbook.

?

Integrating prediction models into publishing workflows allows tasks such as formatting and curriculum coverage checks to be automated, freeing the publishing team to focus on content quality and pedagogic innovation. These tools also allow for data-driven decision-making where predictive analytics can be used for forecasting demand and customer preferences, thereby helping the planning for planning new projects.

?

While generative AI has shown great potential in the publishing industry, it is a tool to assist authors and editors and not to replace them. Human intelligence, creativity and expertise remain crucial in creating high-quality, reliable textbooks. By collaborating with AI, publishers can make the development process more efficient and improve the learning experience for learners.?

Intellectual property?

A publisher's business is in generating and packaging intellectual property. It is in this space where issues around authorship, copyright and AI systems still need to be resolved. Key concerns include determining who should be credited as the author of AI-generated content and the extent to which AI-generated content can be considered original. The?U.S. Copyright Office?has stated that AI-generated material is not the product of human authorship and, therefore, not copyrightable. However, if human involvement is significant in guiding the AI, AI-assisted works may be eligible for copyright. Other countries are also working on providing guidance on AI and intellectual property.

?

Another concern about generative AI in textbook publishing is copyright infringement. If AI-generated content draws from copyrighted sources, it could infringe upon the rights of the original copyright holders. Until a clear legal framework is established, textbook publishers and content creators must remain vigilant to avoid copyright infringement in AI-generated works.

?Concluding thoughts

Generative AI should be used to augment human creativity and expertise, which remain at the heart of quality educational content.

?

Niall McNulty, Cambridge University Press

Disclaimer:

?“The views expressed in this article or by any article or comment or blog associated with this article do not necessarily coincide with the views of PASA or any of its members. Authors invited to express their views are responsible for their content and also for having the necessary rights to submit any article for publication or re-publication (with permission), for not infringing any third-party rights, including copyright and intellectual property, privacy and publicity rights. Authors are also responsible and must avoid causing injury or prejudice beyond what is permitted in an open, free and democratic society and under freedom of expression principles applicable in South Africa by virtue of the genuine content they submit for publication and/or public dissemination.”