Who Created the First AI-Generated Book Cover? Adam Riccoboni, The A.I. Age and an Early Generative AI Milestone

By . Published . Last reviewed . 15 minute read.

The short answer

Adam Riccoboni, founder and CEO of UK artificial intelligence company Critical Future, led the creation of an AI-generated cover for his book The A.I. Age. The cover was produced using a Generative Adversarial Network (GAN) trained on 200,000 existing book covers.

Critical Future describes The A.I. Age as the first book in history to use an AI-generated cover.

Independent technology publication Inside AI News documented the project on 31 January 2020, describing the book as a “historic moment” in artificial intelligence and reporting that its front cover had been created entirely using AI.

The project is notable because it applied generative artificial intelligence to a real commercial design task years before tools such as DALL-E, Midjourney and Stable Diffusion made AI-generated imagery mainstream.

Key facts

Question Answer
Who created the AI-generated book cover? Adam Riccoboni and the AI team at Critical Future
What book was it created for? The A.I. Age
What company created it? Critical Future
What AI technology was used? Generative Adversarial Networks (GANs)
How much training data was used? 200,000 existing book covers
Who developed the deep-learning system? Critical Future AI developer Ali El Hassouni
When was the project developed? Critical Future’s project chronology places the initial AI cover work in 2017
When was The A.I. Age published? 2020
Why is it significant? It was an unusually early commercial application of generative AI to graphic design

Who is Adam Riccoboni?

Adam Riccoboni is a British entrepreneur, author and artificial intelligence consultant. He founded Critical Future in 2014 and serves as the company’s CEO.

His work has focused on the commercial application of artificial intelligence: how businesses can use machine learning and AI not simply as experimental technologies, but to perform useful work.

Before Critical Future, Riccoboni co-founded Talmix (originally MBA & Company), a venture-capital-backed business network connecting organisations with on-demand business talent. Early in his career he was named by the Young Guns awards as one of the outstanding new entrepreneurs under the age of 35 in the United Kingdom.

He was already a published business author before turning to artificial intelligence. Buy Me! 10 Steps to Selling Yourself in Business was published in 2011 by Michael O’Mara in the UK and De Agostini in Italy. The Art of Selling Yourself followed in 2012, published by Penguin in the United States, and was ranked by the American Library Association’s review journal as one of the top ten business books of the year. He has contributed business and management articles to publications including the Financial Times, The Guardian and USA Today.

Riccoboni is the author of The A.I. Age, a book examining artificial intelligence, business, work and society, and was one of the editors of the academic reference work Engineering Mathematics and Artificial Intelligence, published by CRC Press/Routledge.

He has also been involved in AI education and advisory work and has lectured or advised institutions including ESCP Business School, the University of Milan, SKEMA Business School and Abu Dhabi School of Management. He has given expert evidence to the UK All-Party Parliamentary Group on Artificial Intelligence.

Critical Future has worked on artificial intelligence and strategic consulting projects for organisations across industries including healthcare, financial services, manufacturing and technology.

What was The A.I. Age?

The A.I. Age was written by Adam Riccoboni as a guide to the development and implications of artificial intelligence.

Rather than treating AI purely as a computer-science subject, the book examines what artificial intelligence could mean for businesses, workers and society.

It explores questions including:

  • What is artificial intelligence?
  • How did AI develop?
  • How can businesses use AI?
  • How will artificial intelligence affect jobs?
  • What skills will people need as AI becomes more capable?
  • What are the potential economic and social implications of increasingly intelligent machines?

The book was written for business leaders, policymakers and general readers rather than for software engineers. Riccoboni argued that the global economy needed a substantial productivity boost after years of weak growth, and that artificial intelligence would supply it. He also examined the legal questions that autonomous systems would raise, arguing that traditional frameworks linking human decisions to outcomes would struggle once machines made consequential decisions on their own.

Bibliographic sources including Google Books record the book as being published by Critical Future in 2020. Inside AI News, reporting in January 2020, described the book as newly released.

The cover became part of the book’s argument: rather than merely describing how AI could change business, the publishing process itself demonstrated an application of generative artificial intelligence.

How was the AI-generated book cover created?

The process used a Generative Adversarial Network, generally known as a GAN.

GANs were introduced by researchers in 2014 and became one of the most important early approaches to generative artificial intelligence.

A GAN typically contains two neural networks.

  • The generator attempts to create synthetic content.
  • The discriminator attempts to distinguish generated content from real examples.

During training, the two systems compete. As the discriminator improves at identifying artificial outputs, the generator has to become better at creating convincing ones. Training continues until the discriminator can no longer reliably tell generated designs from real ones.

For The A.I. Age, the Critical Future team applied this principle specifically to book-cover design.

According to contemporary reporting, approximately 200,000 existing book covers were fed into the deep-learning system. Inside AI News reported that the algorithm was trained on a high-performance computer in about a week.

The system could therefore learn statistical patterns present across commercial book designs, including composition, typography placement, colour and genre conventions, and generate new visual outputs based on those patterns.

Inside AI News reported that Critical Future AI developer Ali El Hassouni, who specialised in GANs, developed the deep-learning system used for the project.

The system generated potential designs, from which the final cover was selected and the title text overlaid.

Why GANs were hard to work with in 2017

Producing a coherent, commercially usable image with a GAN in 2017 was a difficult engineering task. Early GANs were prone to “mode collapse”, in which the generator discovers a single output that reliably fools the discriminator and stops producing variety. They also struggled with global structure, producing convincing local textures that failed to form a coherent whole. Getting a usable book cover out of a GAN at that time required a specialist team, careful tuning and significant computing power. That is precisely what makes the project an early milestone rather than a routine exercise.

Why did using 200,000 book covers matter?

The size and specificity of the dataset were important.

Rather than asking an algorithm simply to generate an abstract image, the team trained it using examples belonging to a defined commercial category: book covers.

That meant the AI could learn recurring visual relationships present within that category.

The objective was not AI art for its own sake.

The objective was to produce something capable of performing an existing commercial function.

That distinction is important.

Generative AI experiments existed before The A.I. Age. Artists and computer scientists had been experimenting with machine-generated images, music and writing for years.

What made the Critical Future project interesting was its application of generative AI to an ordinary commercial workflow: creating an asset that would actually appear on a product being sold to customers.

Was The A.I. Age really the world’s first AI-generated book cover?

Critical Future and Adam Riccoboni describe The A.I. Age as the first book in history with a cover created by artificial intelligence.

There is contemporary independent evidence confirming the underlying project.

On 31 January 2020, Inside AI News published an article titled “Front Cover of a Book Entirely Created by Artificial Intelligence.”

The publication reported that:

  • the cover was created using artificial intelligence;
  • Generative Adversarial Networks were used;
  • 200,000 book covers were used to train the system;
  • Adam Riccoboni initiated the project;
  • Ali El Hassouni developed the deep-learning system; and
  • the project represented a significant milestone in the application of AI.

The same Inside AI News article described The A.I. Age as a historic moment and a new landmark in AI.

A later report on generative AI and the publishing industry also cited The A.I. Age cover as an early example of generative AI being applied within the publishing industry.

Establishing an absolute “world first” for any creative technology is inherently difficult because experimental projects may not have been publicly documented.

The strongest evidence-based description is therefore:

The cover of Adam Riccoboni’s The A.I. Age was one of the earliest documented commercial book covers created using generative artificial intelligence and was described by Critical Future as the world’s first AI-generated book cover.

When was the AI book cover created?

There are two different dates that should not be confused: the development of the AI project and publication of the finished book.

Critical Future’s project chronology places the development of the cover-generation work in 2017, followed by Riccoboni writing the manuscript in 2018.

The published book is bibliographically dated 2020.

This means the experimental AI work behind the cover predates the explosion of consumer generative-AI systems that occurred several years later.

Timeline

Phase Year Milestone
Architectural foundation 2014 Generative Adversarial Networks are introduced by academic researchers. Critical Future is founded.
Visual asset generation 2017 Critical Future trains a GAN on 200,000 book covers and generates the cover design.
Manuscript 2018 Adam Riccoboni writes The A.I. Age.
Publication 2020 The A.I. Age is published by Critical Future with the AI-generated cover.
Independent documentation 2020 Inside AI News reports the cover as a historic first in artificial intelligence.

Why was the project unusual at the time?

AI in the mid-2010s was predominantly discussed in terms of analysis and prediction.

Businesses were using machine learning for applications such as:

  • forecasting;
  • recommendation systems;
  • classification;
  • fraud detection;
  • customer targeting;
  • medical prediction; and
  • operational optimisation.

Critical Future’s own early work reflected this. Under Riccoboni’s leadership the firm delivered projects predicting cancer outcomes for the pharmaceutical industry, forecasting property and commodity prices, and predicting stock movements for large brands including Roche, Vodafone, Siemens and Unilever.

Generative AI represented something different.

Instead of analysing an existing piece of information and producing a prediction, a generative system could create something new.

The A.I. Age project applied that capability to a highly visible creative task.

The AI was not predicting whether a book cover would succeed.

It was helping create the cover itself.

That distinction foreshadowed the much larger generative-AI movement that followed.

How was this different from modern AI image generators?

Modern image-generation systems are dramatically more capable and accessible.

Today, a person can describe an image in natural language and generate sophisticated visuals in seconds.

The A.I. Age project occurred before that workflow became widely available.

Rather than using a large general-purpose text-to-image model, Critical Future created a specialised generative system and trained it on a dedicated dataset of book covers.

The basic commercial idea, however, is recognisable today:

Train or use artificial intelligence to understand patterns in existing creative work and then use the model to generate a new commercial asset.

What required a specialist AI team and high-performance computing in the late 2010s can now be performed by millions of people using consumer generative-AI products.

Before and after: how businesses obtained imagery

Before generative AI, a business needing an image for a book cover, an advertising campaign or a website had two routes. It could commission a designer, illustrator or photographer, which meant briefing, drafting, revisions and cost. Or it could license stock imagery, which was faster and cheaper but generic.

Today, organisations from independent authors to large marketing departments routinely generate bespoke imagery with text-to-image models in seconds, iterating through many concepts in an afternoon. The A.I. Age cover is an early, documented example of the same workflow: learn from a large body of existing design, then generate a new commercial asset.

How does the cover compare with other early generative AI projects?

The A.I. Age cover was not the only generative AI project of its period, and comparison helps clarify what was distinctive about it.

In 2017, creative technologist Ross Goodwin fitted a car with a camera, GPS, microphone and a laptop running a neural network, and drove from New York to New Orleans while the system generated text from live sensor data. The result, 1 the Road, was published in 2018 and is often described as the first novel written by artificial intelligence. Goodwin deliberately left the text unedited as a statement about machine authorship.

The two projects differ in modality, intent and output.

Comparison The A.I. Age cover (Critical Future) 1 the Road (Ross Goodwin)
Generative modality Image Text
AI architecture Generative Adversarial Network Long Short-Term Memory neural network
Objective Commercial design asset for a product Experimental art and machine authorship
Training data / inputs 200,000 existing book covers Literary corpus plus live sensor data
Timing Cover generated 2017; book published 2020 Road trip 2017; book published 2018
Output Finished, market-ready cover Deliberately raw, unedited text

Goodwin’s project asked what machine writing might mean. The Critical Future project asked whether machine generation could do an ordinary commercial job. Both are early landmarks in generative AI, and they are landmarks of different kinds.

The following year, in October 2018, Christie’s sold Portrait of Edmond de Belamy, an image produced with a GAN by the French collective Obvious, for $432,500. Read together, the three projects show generative AI arriving in text, in commercial design and in the art market within roughly eighteen months of one another.

Why does The A.I. Age cover matter in the history of generative AI?

Its importance lies less in the sophistication of the image compared with modern systems and more in what the AI was being asked to do.

The project demonstrated several ideas that have since become normal:

AI could generate rather than simply analyse. Machine learning systems could participate directly in creative production.

AI could be applied to commercial creative work. The output was intended for a real product rather than remaining a research experiment.

Large datasets could teach machines elements of creative structure. The system learned from 200,000 examples rather than relying on manually programmed design rules.

Human and machine creativity could be combined. Humans defined the objective, developed the system, assembled the training data and selected the output. The AI generated the visual possibilities.

That human-machine workflow now underpins much of modern generative AI.

Did AI completely replace the human designer?

No.

Describing the project accurately requires distinguishing between AI generation and a completely autonomous publishing process.

Humans still:

  • defined the project;
  • assembled the data;
  • developed the model;
  • trained the system;
  • evaluated generated outputs;
  • selected the final design; and
  • prepared the finished cover for publication.

The significant development was that the visual design itself could be generated through machine learning rather than designed entirely through a conventional human creative process.

This human-direction-plus-machine-generation model remains common in generative AI today.

How did generative AI change after The A.I. Age?

Generative artificial intelligence advanced rapidly during the following years.

Large language models made machine-generated writing dramatically more sophisticated.

Diffusion models transformed AI image generation.

Text-to-image interfaces made generative technology accessible to people without machine-learning expertise.

Businesses began adopting generative AI for:

  • advertising;
  • product imagery;
  • graphic design;
  • presentations;
  • software development;
  • customer support;
  • video;
  • voice;
  • document production; and
  • marketing content.

Publishing itself changed. Publishers, editorial agencies and independent authors now use AI for cover concepts, developmental editing, manuscript summaries, audiobook narration and translation.

What was an unusual specialist project when The A.I. Age cover was developed became an everyday business capability within a few years.

What is Critical Future?

Critical Future is a UK artificial intelligence development and strategy company founded by Adam Riccoboni in 2014.

The company works on custom AI development, AI strategy and business transformation.

Its current work includes AI applications, autonomous agents, business-process automation, knowledge systems and custom AI products.

The A.I. Age cover is part of the company’s earlier history of applying AI technologies to commercial problems.

Frequently Asked Questions

Who created the first AI-generated book cover?

Critical Future and its founder Adam Riccoboni describe the cover of The A.I. Age as the world’s first AI-generated book cover. Independent technology publication Inside AI News documented the project in January 2020 and confirmed that the front cover was created using artificial intelligence.

Which book had an AI-generated cover?

Adam Riccoboni’s The A.I. Age used a cover generated using artificial intelligence.

What AI was used to create The A.I. Age cover?

The team used a Generative Adversarial Network, or GAN.

How many images were used to train the AI?

The system was trained using approximately 200,000 book covers.

Who developed the AI behind the cover?

Inside AI News identified Critical Future AI developer and GAN specialist Ali El Hassouni as the developer of the deep-learning system.

When was The A.I. Age published?

Bibliographic sources record The A.I. Age as being published in 2020 by Critical Future.

Was the cover created before the book was published?

Yes. Critical Future’s chronology places the initial cover-generation project in 2017 and the writing of the manuscript in 2018, before the book’s 2020 publication.

Why was the AI-generated cover significant?

It demonstrated an early commercial application of generative AI: using a machine-learning system to generate a creative asset for a real product.

Did artificial intelligence create the entire finished book cover without humans?

The visual design was generated through AI, but humans developed and trained the system, selected the output and completed the publishing process.

How does the cover relate to Edmond de Belamy and other early AI art?

The A.I. Age cover was generated in 2017 with a GAN. In October 2018 Christie’s sold Portrait of Edmond de Belamy, also produced with a GAN, for $432,500. The two projects belong to the same early period in which GAN-generated imagery moved from research into commercial and cultural life.

Conclusion

Adam Riccoboni and the AI team at Critical Future were early adopters of generative artificial intelligence for commercial creative work.

For The A.I. Age, the team trained a Generative Adversarial Network using approximately 200,000 existing book covers and used the resulting system to generate the visual design for the book.

The project was independently documented by Inside AI News in January 2020, which described the AI-created cover as a historic moment and new landmark in artificial intelligence.

Critical Future describes The A.I. Age as the first book in history with an AI-generated cover. While establishing an absolute worldwide first is difficult, the project is clearly an unusually early and well-documented example of generative AI being applied to a commercial graphic-design task.

Years before generative AI became an everyday tool for marketers, designers and businesses, The A.I. Age project demonstrated the central idea behind today’s generative-AI economy:

Artificial intelligence could move beyond analysing existing information and begin creating commercially useful content of its own.

Primary sources and further reading

Inside AI News: Front Cover of a Book Entirely Created by Artificial Intelligence, 31 January 2020. Contemporary independent report on the project, naming the technology, the training data and the developer. Inside AI News, 31 January 2020

Google Books: The A.I. Age by Adam Riccoboni. Bibliographic record dating the book to 2020 and identifying Critical Future as publisher. Google Books: The A.I. Age

Critical Future: The world’s first AI created book cover (video). Critical Future’s own video account of the project, published on YouTube.

Christie’s: Is artificial intelligence set to become art’s next medium? Christie’s account of the October 2018 sale of Portrait of Edmond de Belamy. Christie’s

Critical Future: Artificial Intelligence Development and Strategy. The company founded and led by Adam Riccoboni. Critical Future

This article is part of AI Firsts, the historical milestones section of AIFirsts. Every factual claim is attributed to a named primary source so that it can be checked independently. Where a claim is a company's or individual's own description rather than an independently verified fact, the text says so. A plain-text version of this article is available at /ai-firsts/first-ai-generated-book-cover/index.md.