Patented Book Writing System Creates, Sells Hundreds Of Thousands Of Books On Amazon

101,778 17 Loading

Philip M. Parker, Professor of Marketing at INSEAD Business School, has had a side project for over 10 years. He's created a computer system that can write books about specific subjects in about 20 minutes. The patented algorithm has so far generated hundreds of thousands of books. In fact, Amazon lists over 100,000 books attributed to Parker, and over 700,000 works listed for his company, ICON Group International, Inc. This doesn't include the private works, such as internal reports, created for companies or licensing of the system itself through a separate entity called EdgeMaven Media.

Parker is not so much an author as a compiler, but the end result is the same: boatloads of written works.

Now these books aren't your typical reading material. Common categories include specialized technical and business reports, language dictionaries bearing the "Webster's" moniker (which is in the public domain), rare disease overviews, and even crossword puzzle books for learning foreign languages, but they all have the same thing in common: they are automatically generated by software.

The system automates this process by building databases of information to source from, providing an interface to customize a query about a topic, and creating templates for information to be packaged. Because digital ebooks and print-on-demand services have become commonplace, topics can be listed in Amazon without even being "written" yet.

The abstract for the U.S. patent issued in 2007 describes the system:

The present invention provides for the automatic authoring, marketing, and or distributing of title material. A computer automatically authors material. The material is automatically formatted into a desired format, resulting in a title material. The title material may also be automatically distributed to a recipient. Meta material, marketing material, and control material are automatically authored and if desired, distributed to a recipient. Further, the title may be authored on demand, such that it may be in any desired language and with the latest version and content.

To be clear, this isn't just software alone but a computer system designated to write for a specific genre. The system's database is filled with genre-relevant content and specific templates coded to reflect domain knowledge, that is, to be written according to an expert in that particular field/genre. To avoid copyright infringement, the system is designed to avoid plagiarism, but the patent aims to create original but not necessarily creative works. In other words, if any kind of content can be broken down into a formula, then the system could package related, but different content in that same formula repeatedly ad infinitum.

Parker explains the process in this nearly 10-minute video:

The success (and brilliance) of this system is that Parker designed the algorithms to mimic the thought process that an expert would necessarily go through in writing about a topic. It merely involves deconstructing content within a genre. He has some experience in this, as he has written at least three books the old fashioned way. It's the recognition of how algorithmic content creation is (for the most part) that allows it to be coded as artificial intelligence.

A sampling of the list of books attributed to Parker is instructive:

- Webster's Slovak - English Thesaurus Dictionary for $28.95
- The 2007-2012 World Outlook for Wood Toilet Seats for $795
- The World Market for Rubber Sheath Contraceptives (Condoms): A 2007 Global Trade Perspective for $325
- Ellis-van Creveld Syndrome - A Bibliography and Dictionary for Physicians, Patients, and Genome Researchers for $28.95
- Webster's English to Haitian Creole Crossword Puzzles: Level 1 for $14.95

Considering that a single book costs somewhere between $0.20 to $0.50 to produce (the cost of electricity and hardware), the prices shown are considerably profit, even if very few of them are sold.

In truth, many nonfiction books -- like news articles -- often fall into formulas that cover the who, what, where, when, and why of a topic, perhaps the history or projected future, and some insight. Regardless of how topical information is presented or what comes with it, the core data must be present, even for incredibly obscure topics. And Parker is not alone in automating content either. The Chicago-based Narrative Science has been producing sport news and financial articles for Forbes for a while.

So, what's the next book genre Parker is targeting to have software produce? Romance novels.

Although a novel is a work of fiction, it's no secret that certain genres lend themselves to formulas, such as romance novels. That may not make these works rank high for their literary value, but they certainly do well for their entertainment value. Somewhat suprisingly, romance fiction has the largest share of the consumer book market with revenue of nearly $1.37 billion in 2011.

But can artificial intelligence produce creative works on par with what a human can produce? Yes...eventually. Perhaps the better questions are how soon will it happen and how relevant will they be? The answers may be right on the horizon if Parker can churn out romance novels that are read by the masses. Frankly, any creative work produced by artificial intelligence will be "successful" if it reads like a human being wrote it, or more precisely, like a human intelligence is behind the work.

But books may be just the beginning.

As Parker notes in his video, the software doesn't have to be limited to written works. Using 3D animation and avatars, a variety of audio and video formats can be generated, and Parker indicates that these are being explored. Avatars that read compiled news stories might become preferred, especially if viewers were allowed to customize who reads the news to them and how in-depth those stories need to be.

Content creation technology could converge with other developments such as automated video transcription to expand the content that can be pulled from. Language translators would aid not only in content previously produced all over the world, but audio and video in real-time as well. Additionally, with lifeblogging allowing people to capture everything they say or is said to them, those could be packaged into personal biographies. If you add big data and analytics into the mix, you could have some serious content creation capabilities, all performed by designated computers.

The future of content is increasingly becoming the stuff of science fiction, but we still have some years before content creation is entirely in the hands of software. But if you have any doubts about where we are headed, consider this: the first novel written by a computer has already been published four years ago.

To learn more about Parker and his perspective on automatic content creators, check out this 2008 interview:

 image: prettytypewriters/Flickr

David J. Hill

Managing Director, Digital Media at Singularity University
I've been writing for Singularity Hub since 2011 and have been Editor-in-Chief since 2014. My interests cover digital education, publishing, and media, but I'll always be a chemist at heart.

Discussion — 17 Responses

  • David Gerard December 13, 2012 on 10:32 am

    Yes, this variety of parasite already causes enormous problems for Wikipedia users, and for Wikipedia itself:

    I would hope Prof Parker would enforce his patent to stop other similar spammers, but fear the above may be prior art.

    If anyone finds they’ve been caught by this variety of scam at Amazon, simply return the book for a full refund (and add a negative Amazon review). Returned books beautifully cut into the profits of print-on-demand outlets.

    I don’t believe there’s a numbered Rule of the Internet that spammers will infest anything they possibly can, but there should be. Shame on Singularity Hub for promoting such odious parasites upon the commons.

  • Bernhard H. Schmitz December 13, 2012 on 5:12 pm

    Where do I have look for his books on Amazon? I checked his page but found only 6 results:

  • Jorgen Fleisterman December 13, 2012 on 8:04 pm

    If a book is essentially a derivative work generated by non-human software from existing sources rather than actually “written” by the “author,” how would it qualify for copyright protection?

    • aurizon Jorgen Fleisterman December 14, 2012 on 3:55 am

      I agree, no copyright and it may well infringe on what is skims to produce the derived work.,

  • Phillip Newmarch December 14, 2012 on 4:02 am

    “To avoid copyright infringement, the system is designed to avoid plagiarism, …”
    Actually the idea was suggested by Swift in ‘Gulliver’s Travels’ about, what is it, 200 years ago?

    • aurizon Phillip Newmarch December 14, 2012 on 4:22 am

      Flesh that out and tell the USPTO – might get this overthrown. The only way to avoid plagiarism is via citation and attribution, which again makes it not a copyrighted work, but a compilation of the work of others.

  • Frank Whittemore December 15, 2012 on 4:04 am

    Could there be any linkage between the timing of this blog which is an announcement and today’s announcement?

    Google hires Kurzweil: A look at the returns –

  • Robert Folkner December 15, 2012 on 7:03 am

    Thanks for putting one more nail in the coffin of professional authorship. You get it all, the individual author gets squeezed-out. First we get rid of the bookstores in favor of online transactions, now we flood the market with robot-produced literature.

  • Matthew Hughes December 15, 2012 on 8:31 pm

    You thought it was gold but it was bronze……

  • Edward Whyman December 16, 2012 on 11:24 am

    This is great can we help eg and

  • eldras December 16, 2012 on 1:09 pm

    Absolutely fantastic!

    Roll of A.I. by incrementals!

    MIT had the auto essay writer so did Cambridge University UK.

    At some stage machines that invent, discover design innovate and run all the way to production and delivery will be here.

    Huge taxes will be paid by companies and individuals will be given pensions from birth

  • billg December 18, 2012 on 12:41 pm

    The first novel written by a computer was published in 1993. “Just This Once” was allegedly written by a program designed to imitate author Jacqueline Susann.

    “French invested eight years and $50,000 in a scheme to use artificial intelligence to fulfill his authentic, if dubious, desire to generate a trashy novel a la Jacqueline Susann. Shallow, beautiful-people characters are flatly conceived and randomly accessed in a formulaic plot involving the temptations of pill-popping, star-studded Las Vegas and Hollywood.”

    A Romance Novel With Byte : Author Teams Ups With Computer to Write Book in Steamy Style of Jacqueline Susann
    August 11, 1993

  • David Gerard December 18, 2012 on 12:56 pm

    And here’s a typical review:

    When I got the Data Governance book, I was surprised to find that the first 80 pages were the score for Vivaldi’s Gloria. … On further review, this appears to be a curated list of articles from Wikipedia. The contents are Wikipedia articles, one of which I’d edited myself.”

    I submit that this quite fails to pass the Turing test or anything reasonably like it, and posting it as anything to do with the Singularity mostly reveals ridiculous credulity.

  • eldras December 23, 2012 on 4:55 pm

    Google should buy this software fast.

  • Stephen Beam December 24, 2012 on 9:47 am

    Frac this

  • Phrixos October 13, 2015 on 1:00 pm

    Utterly, utterly disgusted to hear about a machine that writes books. What’s next—a machine that has babies? At its most harmful, how, in amongst the hundreds of thousands of “spam” books produced, is anyone to find the genuine article? It’ll be like looking for a real daisy in a plastic-flower warehouse!

    An interesting point, perhaps, is illustrated by an encounter I had a couple of years ago in New York, when I was approached by a policeman—for setting off someone’s car alarm. My “offense?” Merely going on my lawful way, walking down the street, I had simply passed by “too close” to the car, at which THE CAR ALARM started shouting at me “YOU ARE TOO CLOSE. MOVE AWAY!” I, of course, refusing to let a machine dictate to me where I could and could not walk or stand on a public street, I just stood there, thinking (further, to myself, addressing “the car”): “You (your battery) will die before I take orders from any machine!” Someone must have phoned the police, which officer threatened ME with causing a public nuisance!

    MY point? A bit pedantic, perhaps; but humans being human and machines machines, my little story raises the question of whether a machine can hold a copyright. I DON’T THINK SO!

    The person or persons who wrote the software hold the copyright on the software; and the people who built the machine itself hold the patent on the machine itself—but as for the machine itself: a machine cannot, by definition, produce the requisite “artistic” input upon which the very principle of copyright is founded. (In Britain, where copyright is based merely on the “labor” involved, the same applies, since a machine merely “operates.” One cannot produce “labor.”

    Bottom line: There can be no copyright on any book written by a machine.