Don’t rage against the machine! How scholarly publishers can embrace AI

Don’t rage against the machine! How scholarly publishers can embrace AI

Recent predictions claim that AI will contribute $15.7 trillion to the global economy by 2030. From self-driving cars, personal assistants and chatbots taking on roles as wide ranging as customer services, medicine, law and therapy - AI continues to generate hype, excitement and even fear.  AI will clearly shape our future, but what will this future look like? And what - if anything - does it mean for scholarly publishing?

Stephen Hawking told us that “AI will be either the best, or the worst thing, ever to happen to humanity”. There are certainly some big issues to be tackled, including privacy, transparency, security, ethical concerns around training data containing human bias, and even what jobs the future holds for humans in an automated world.  While we need to keep these factors in mind, I personally believe that AI - if applied appropriately and diligently in the right places - will bring more positives to scholarly publishing than negatives. Although each publisher is unique there are many shared problems to solve and areas of opportunity. Here are five key areas where AI can make a difference.

Peer review

AI is already being utilized by some publishers to augment the current peer review process, performing tasks such as checking the completeness of submissions, assessing suitability for the intended journal, reducing human bias, finding suitable peer reviewers, checking for conflicts of interest and handling the workflow between authors, editors and reviewers.

Fighting fraudulent practices

By using Natural Language Processing AI can detect plagiarism more accurately than traditional software that only detects phrases that has been copied word for word. AI is also being used to detect whether data or images have been fraudulently modified, duplicated or falsified.

Predicting high impact research and emerging subject areas

Imagine being able to alert your marketing team as soon as a high impact piece of content is submitted. AI has been shown to outperform humans in predicting content that is likely to create a buzz. Similarly, AI can be used to predict emerging and growth subject areas to inform your editorial strategy.

Auto-creation of content

AI is already being used for the auto-generation of abstracts and metadata, for example for legacy articles or book chapters you wish to repackage or drive usage of. Another use case is to summarise research papers and automatically generate press releases and other marketing outputs. Some are looking to take this one step further by training an AI to take research data as an input and to output a human readable article.

Delivering personalised user experiences  

AI can be used to power a number of features to boost discoverability and usage of content and deliver personalised experiences for users. For example, by understanding the relationships between documents AI can power recommendation services, personalised content alerts, the ability to automatically create custom content collections as well as value-add features such as the ability to ask a question rather than input a search term or paste in your hypothesis and be presented with a reading list.

With such a large number of opportunities you might be wondering: how should we begin to engage with AI?  First, I recommend that you really understand the business problem and whether AI is the appropriate solution, before thinking about implementation. AI is not a silver bullet that will solve all your business issues, but with enough data it may be trained to perform a particular task or solve a well-defined problem.

Next, be honest about your business’ technology skills. Consider partnerships, or using a trusted vendor, or think about where you need to hire or upskill. AI is a technology, but to deliver real value it needs to be implemented by a cross-functional team. As well as technologists, you will need subject matter experts, operational and commercial stakeholders, and staff currently trying to solve the problems to help train the algorithm and shape the solution.

Finally, give your robot some breathing space!  Expecting AI to solve your business problem perfectly first time without adequate training and input is a common misconception.  Think about the data you already have and the data you would need to collect. AI is only as good as the training data you feed it.

So, what will 2030 look like?  It’s impossible to give an exact vision of the future, but AI is here to stay. Publishers large and small should strive to understand AI and engage with it now. Or they risk being left out of the future altogether.