A year and a half after I was appointed global AI lead at Pan Macmillan, I’ve been reflecting on where we are –as an industry – with AI. 

AI developments continue at a dizzying pace. Meanwhile, the UK government’s consultation on AI, serious concern about copyright theft and stories like this one in The Atlantic continue to cause huge dismay, pitting AI and creativity into an unhelpful opposition.

A few industry commentators have hardened themselves into one corner or the other – as for or against AI in general – but that is not the majority and nor does it help; most of us, I expect, are trying to tread at once a strategic and sensitive path.

As sensible as that might sound, how does it translate when it comes to setting the right AI strategy as publishers? And, how do we ensure we are effectively navigating these changes and bringing our colleagues with us as we go?

Here are some things I have learned.

The need to acknowledge and keep a variety of different attitudes in play at once continues: we must address the concerns of our constituencies about the seeming disregard for copyright that big tech companies have shown; we must acknowledge the existential fear of how far AI might disrupt or destroy creative work and incomes and make plans to mitigate this; and we must also maintain an understanding of the fast-developing capabilities of AI, engaging with it and adopting it where it makes sense for the business, our authors and illustrators and our readers and listeners.

This inherent friction is perhaps leading to a relative dearth of public stories about how publishers are leveraging AI; and without positive stories there is a vacuum which leaves room for all sorts of (often incorrect) assumptions to run rife.

Understanding the views, ideas and concerns of our constituencies has therefore been a vital primary aspect of the work we have undertaken at Pan Mac. We’ve held open sessions for all our people, run surveys and hosted AI roundtables with authors, illustrators, agents and other industry contacts, and we have definitely been "on a journey", as they say.

What we have found is that, perhaps unsurprisingly, somewhere between hype and scepticism lies reality. As we have developed our strategy and begun to leverage AI in the business, we have found it helpful to acknowledge that people are in different places about AI, and not to assume homogeneity of views or approach; and adopting a variety of communication strategies, engagement programmes and formats has helped to support our work. 

We have to support people to live with a radical acceptance of continuous change; it’s vital not to be buffeted by every new announcement, irrespective of how distracting these continuous updates can be

At our most recent stakeholder AI roundtable, and in the halls of this year’s London Book Fair, there was a general sense of emotions settling when it comes to AI, a greater sense of pragmatism about its potential and a growing number of tools and services on offer.

Nonetheless, it can still be challenging to distinguish which ideas, tools and technologies to adopt and which to ignore, how much we should build versus buy, where to deploy tools tactically and where to make larger strategic investments. How to select the right partners and tools to work with also preoccupies us – especially given the ethical questions surrounding AI and its impact on the environment.

Alongside this, there is the need to keep our people focused on delivering the day job (books! campaigns! targets!) at the same time as learning and upskilling for an increasingly AI-dominated future, and a sense of enormous responsibility to manage that transition effectively.

This is all a long way of saying that managing this shift is… complicated. So, what can I share about how we have learned to navigate these changes at Pan Mac?

First of all, we have to support people to live with a radical acceptance of continuous change; it’s vital not to be affected by every new announcement, irrespective of how distracting these continuous updates can be. One way to ground everyone and to set the right pace is to tie AI strategy to the needs and values of the business: where can AI support us in the areas we want to grow, deliver value or offer points of distinction? Where new AI capabilities don’t yet deliver on any of these, it is probably safe to wait and see how things develop before rushing in, or ringfence a smaller pilot approach.

It is equally important to remember that change doesn’t happen in a straight line; learning to follow wiggly paths towards a goal and facilitating a culture which can feel more comfortable with this is key. Implementing a programme of continuous learning and helping our people to pursue their curiosity supports this. Developing more agile structures and frameworks for projects and learning to kill projects which don’t deliver, or are replaced by something better arriving on the market, is a must.

Modelling and facilitating innovation and giving people the space for it is challenging, but is non-negotiable if we are to emerge successfully through this transition. The need for constant communication, even over-communication, cannot be underestimated, as is the requirement to deliver data and evidence to monitor impact, support progress and communicate efficacy.

To bring people with us we must ensure they understand that automation does not have to sit in direct opposition to human work, but can be introduced where it supplements and enhances. We are not pursuing a strategy of binary choices. And it’s important to stay optimistic about the power of books, of human-told stories and the impact they can have on all of us.

Let’s not forget: Sarah Wynn-Williams’ Careless People could not have been written by AI…