Harnessing the Power of AI: Shaping the Future of Research Publishing and Content Innovation

Exploring the Future of Research Publishing

It is a pretty new team, we are going into our 2nd year only. My role is heading up a team of leaders who aim to grow, accelerate, and unlock research content generation from short article highlights to longer forms such as papers, books, audio or video with a strong focus on Open Access content and Open Science in general. We design innovative content formats and user experiences which add value, both from a customer and commercial perspective. My role is also to bring people from various departments of Springer Nature together to collaborate on content innovation and explore the future of research publishing. While we continuously monitor industry trends and are in close contact with the communities we serve, our overall mission is to make research content generation a delight for researchers to drive growth and impact.

Growing Influence and Potential of AI

I have been an early advocate of Artificial Intelligence, exploring its potential in many areas of Springer Nature for almost ten years now, be it innovative products and services or internal processes and workflows. I remember there was a time when the Internet was new, now it is everywhere, to the point that we don’t even mention it anymore. The same goes for terms such as online or digital. AI will equally become ubiquitous. It is great to see how AI can help researchers in so many areas and with my team I’m eager to explore this sheer use case heaven. Being part of driving AI-related projects at Springer Nature in this exciting transition period of AI to become part of everyone’s normal life, is rewarding in itself.

AI in the Publishing Industry

AI will become ubiquitous in our industry. Everyone needs to become AI-savvy. I firmly believe that every single job in research publishing and the media industry will not only be affected but rather augmented by AI. For anybody in their early careers or career advancement, please consider AI as a key competence in your profile. And since AI will free up time, skills such as creativity and problem solving are becoming much more important than before, in line with what IBM’s Matthew Candy said in a recent interview in Fortune that if you want to work in tech, you won’t need a computer science degree anymore. AI will allow anyone to convert ideas into products without the need to code.

Current Projects: Human-Machine Synergy

One of the main projects I am currently working on is AI-enabled content generation, both on the journals and the book side. On the journal side we introduced a service which summarises research papers into various different flavours. On the book side we are preparing a tool which allows authors to generate their book manuscript in an iterative human-machine interaction.

Opening New Possibilities: AI Assisted Books

One of the projects that I am particularly proud of was the publication of the first GPT-generated research book in 2023. The book “Einsatzmöglichkeiten von GPT in Finance, Compliance und Audit” was developed during a hack day in spring last year, where three expert authors collaborated with Springer Nature editors on the generation of a book manuscript using GPT. It then took less than five months to publish the book. Not only were we able to cut the time it normally takes to publish a book by half, we were also able to lower barriers for authors who, for various reasons, find it difficult to write due to time constraints and would not be able to author a book without assistance. We are now aiming to build this approach into a tool which allows authors to generate their book manuscript in an iterative human-machine interaction while keeping the entire ownership of the manuscript with the authors themselves, who play an essential role in ensuring high quality content.

The Team’s Passion for AI

I am privileged that my team consists of AI evangelists. We are continuously inspiring each other, sharing latest news in our Slack channel, celebrating breakthroughs, sharing AI training resources, such as the latest best practices for prompt engineering, or organise lightning talks, webinars, hackdays and much more. And - of course - we celebrate our successful innovations and run demo sessions. Let me share one example. My team colleague Stephanie Preuss has managed the development of a service that summarises research papers into various different flavours. It’s an excellent tool that involves human fact checking, an in-built human-machine handshake in a lean and understandable tool, really demonstrating how AI can help with repetitive tasks and can make life easier while enhancing the creativity of the users. I think that is truly phenomenal.

Supporting Authors without Losing the Human Touch

It is the same core problem on the journal as on the book side: How to create more relevant content in less time, in other words, faster and more seamless content generation without losing human oversight and expertise while keeping the entire ownership of the manuscript with the authors. Many researchers consider the writing process of research articles and book manuscripts as a pain. The more we can support them, the better they can focus on their actual research in the lab. And the more we can convert their hassle with the manuscript writing process into a delightful content generation experience, the more we will be able to convert them into loyal authors and successfully address authors who wouldn’t have written a manuscript in the first place due to time constraints.

Future Omnipresence of AI and the Necessity of Human Oversight

In the future, there won’t be any research content generation without generative AI anymore. We need to understand this from the bottom up to build a sustainable business perspective and to be ready to support our customers along the journey. There is a considerable risk that AI enables researchers to bypass publishers as trusted gatekeepers and guarantors of sound science and create research content that appears meaningful while potentially containing all sorts of flawed and biased information and results. We are in the position to provide an infrastructure and services for researchers to create reliable content faster and more efficiently while also providing guidance for best practice which sustains trust, integrity and reputation in line with AI safety. This is only possible through human oversight of technology.

Exploring the Future of Research: AI Assisted Publishing

I was driving a number of industry first products. That in itself has been a great motivation. But more important is why we are doing the things we are doing. I envision a future where researchers in many cases will only need to submit their research results, which will then be used to auto-generate articles for varied knowledge levels and different audiences. This will allow researchers to publish faster and focus on their research. In fact, research and content generation will very much synchronise instead of being a sequential process of conducting the research followed by writing the research paper as it is now. Being active to help define this area in terms of opportunities and guardrails, is a huge motivational driver for me.

Transparency and Accountability: the Future of AI

I started to work very early in the field of AI content generation, programming neural networks years ago with then state-of-art Long-Short-Term-Memory approaches that could generate simple language. Generative AI expanded content generation greatly which is like a paradise for me. Finally, we are able to do what I was waiting for a long time already. And we can do it with a strong emphasis on AI ethics such as human-centricity, transparency, accountability and minimising harm which allows us to find the right balance. This is important as it concerns also my role as a responsible citizen.