Editor’s Note: Today’s post is by Wendy Queen. Wendy is Chief Transformation Officer at Johns Hopkins University Press and former Director of Project MUSE. 

I first met Nadim Sadek at the 2025 Dubai International Library and Publishing Summit, where we served together on a panel titled “Copyright in the Age of AI Strategies.” The session brought together publishers, authors, technologists, and policymakers to examine how generative AI is reshaping authorship, ownership, and creative responsibility.

Held at the Mohammed Bin Rashid Library, the summit reflected a distinctly global perspective on artificial intelligence. AI was not framed as experimental or optional. It was discussed as civic infrastructure, tied to workforce development, cultural preservation, and the long-term sustainability of knowledge systems. One announcement stood out: the UAE’s commitment to train one million people in AI skills across generations and sectors. AI was positioned not simply as a tool, but as national capacity building.

Sadek’s contributions were measured and pragmatic, rooted in creative practice rather than abstraction. They led me to his recent book, Quiver, don’t Quake – How Creativity can Embrace AI, which argues that AI can strengthen human creativity rather than replace it. Sadek is a creative strategist and founder of Shimmr AI, with a career spanning storytelling, technology, and publishing.

Against that backdrop, I asked him to reflect on questions many publishers, including university presses, are actively navigating.

I’ll first ask: what was the most inspiring aspect of the conference that you are still thinking about? For me, it was learning that the government is training a million people (all ages, all genders) in AI.

That learning struck me too — and their strategy for distributing it across all ‘strata’ of society, so whoever you are, however you live, wherever you work, it should be possible to interact with someone expert in using AI. Finland is doing something similar. 

This publishing conference struck me as noteworthy for having the world’s first AI cabinet-level minister as its keynote speaker, too. The UAE is becoming an ‘AI-first’ society. It’s an interesting phenomenon as AI has the ability not only to streamline and make processes efficient in life — whether it’s immigration or car registration or utilities planning — it also is able to democratize information. What once was the guarded repository of an elite is now available pretty freely to just about anybody with a means of accessing the internet. With knowledge comes power, as we know. Compared to Gutenberg, this liberation of previously constrained information is a new, vivifying avalanche. 

Having read my book, you’ll know that I have coined the term, The Panthropic. I use it to suggest that we should see AI as the repository of all human history and accomplishment, in whatever modality it was expressed. I find it hugely exciting that, with 8 billion people on Earth, we now have a technology which, through Socratic dialogue, allows us all to identify, excavate, contour and develop our very own articulations about the world and life. 

Just to be unambiguous — I do not see AI as the means of forming new creative artefacts…making books, drawing pictures, forming music…it’s more about allowing each of us to search, identify and begin to articulate our unique, idiosyncratic, eccentric takes on our world. 

So, to answer more briefly, I’m excited to see AI playing a central role in a country’s life, bringing knowledge, expression and creativity to all.  

Definitions of AI differ across the industry. Whether generative, assistive, or analytical, which forms do you think will most influence publishing over the next three to five years?

I believe we will see two distinct ways of thinking about AI — first, that it enables efficiency; second, that it emancipates expression and creativity.

The Efficiency AI will assist with qualifying submitted manuscripts (I dislike the phrase slush-pile as much as I dislike ‘AI Slop’!), proofreading, editing, printing, distributing, warehousing, legal document management, royalty payments, and so on… 

Emancipation AI will follow more slowly, but it will surely arrive. Publishers will iterate their own versions of House AI as part of their creative collaboration with creators. This will be ethically trained, be full of assistive tools (for example, creative brainstorming, self-editing to house style) and provide services to authors which are either scarce through the publisher’s team (for example, marketing for mid-list titles) or infernally difficult to master for someone who wishes to focus as purely as possible on developing their creative concepts (for example, competitive set analysis, precedents and trends). 

How do you see core publishing infrastructures such as metadata, rights management, accessibility, and multilingual workflows evolving alongside AI?

I’ve got a couple of broad concepts that I think are going to confront publishing. 

One is the notion of Multi-Modal Interactivity. More and more, people are interacting with a single creative idea through multiple modalities — they experience it in different ways. A book becomes a film. A game becomes a book. Music becomes a poem. Podcasts extend an author’s articulations in a book. Live events bring art to life. I love this fecundity. 

To me, this means that publishers will increasingly become Creative Property Managers. It’s more than just IP management. It’s ensuring that if a creative idea has legs, it extends broadly, ensuring its creator reaches far and wide, and their unique contribution to our culture can be experienced in a multiplicity of fashions, by many different audiences. Historically, creatives have expressed their ideas through one favored craft — I write well, you paint, she makes music. In future, it’s more the creative idea, not the original crafted artefact, that will be identified with the author. Of course, original expressions will be the Ur-Text, the fount of all other expressions of the idea. 

Within this broad concept, things like whether a book is printed, made into an e-book or audiobook, or expressed in multiple languages will be trivial considerations. 

The second concept I’ll express with greater brevity: it’s the acceptance that in a time of super-abundance (emanating from the reality that AI will release all 8 billion of us on Earth to be more articulate and expressive), what’s most needed is super-discoverability. Multi-Modal Interactivity will assist discovery. If an idea can be experienced in many different ways, we’ll all learn about it faster and more deeply interact with it. Many AI-assisted discovery tools — like my own company, Shimmr AI, which produces autonomous advertising to sell books – will flourish to help people match themselves with desired (or serendipitous) creativity. 

What cultural, ethical or governance frameworks do you think are missing from current AI adoption in publishing?

The worst problem has been the unauthorized use of copyrighted materials. People who create things must be recognized for that creation and rewarded for it. 

I think we need to accept that a tech-enabled, industrial-volume system for trading rights must be developed, which can work across all modalities (not just books), markets and languages. Creating an effective marketplace where due compensation for the usage of someone else’s thinking is an ethical imperative for rights-creators or rights-holders. 

Beyond that, I think we are going to see AI being trained not so much in the ‘obvious’ corpora of English, Spanish, Chinese, and so on, but in ‘low resource’ languages around the world. The effect of this will be to fill The Panthropic with a truly humanity-expressive set of understanding. It’s not just the languages that are interesting, it’s the fact that within them lie a heterogeneity of values, thinking and cultural mores, that the world would be richer and better for understanding. What do the Bhutanese value most in life? In Swahili, what is regarded as a trivial discourtesy compared to what Germans think? Why do Peruvians regard the potato as sacrosanct, while the Sudanese think the same of a different vegetable? And so on…

For university presses and scholarly publishers especially, what role does change management play in moving from legacy workflows to AI-augmented ones?

I perceive that change management in publishing isn’t primarily about technology adoption — it’s about reimagining roles whilst preserving what makes human judgment irreplaceable. Quiver, don’t Quake articulates the primacy of human input and judgment. I think this pervades the publishing ecosystem.

I’ve watched publishers struggle most when they frame AI as a replacement rather than an augmentation. Their anxiety is palpable as the editors will become mere gatekeepers for algorithmic outputs. Will marketing teams surrender creative judgment to optimization engines?

In my experience, what works best is starting with the tedious. Let AI handle metadata standardization, rights tracking, competitive analysis — the infrastructural work that can quickly overwhelm or exhaust human capacity. This builds trust. People see their time liberated for what they do distinctively well: curatorial judgment, developmental editing, and understanding why a particular manuscript matters.

For scholarly publishers especially, there’s a different imperative. You’re stewarding knowledge across generations. AI can assist with peer review coordination, citation analysis, accessibility compliance — but the determination of what constitutes rigorous scholarship, what advances human understanding? That remains fundamentally human work, really not threatened by the advent of AI within publishing.

My observation has been that transition succeeds when leadership clearly articulates the path to be followed, within boundaries that are unthreatening yet progressive: for example, expressed in my words, “Here’s what we’re automating (the mechanical), here’s what we’re amplifying (your expertise), and here’s what remains inviolably yours (curatorial authority).”

So, start small. Prove value. Then expand deliberately. The technology should serve the mission, not determine it.