Would a time-traveling author from the past centuries stumble upon our everyday read/write tools, he would envision a techno-utopia that allows anyone to act as an archivist, librarian, content curator, or publisher. But the electronic publishing disruption comes with a couple of side-effects: print-on-demand spam is sneaking into our search queries, massively distributed authorship is taking the infinite monkey theorem at face value, while a generation of writers is turning SEO-aware. In that context, Greyscale Press – a post-digital publishing house – is crafting book-like artifacts, merging the toolsets inherited from 20th century modernist avantgardes, post-structuralism, the free software and copyleft movement, up to the latest crop of crypto- and cypherpunk activists.

Download Slides (PDF)


[Video begins midway through introduction] ripped. That he published. That he used, and in open source manner, he put it on his website. Probably it's out of date right now because the Twitter API changes every day, and this is another example of a

publisher [video cuts out]

[Video resumes] form that was accepting entries. It was launched in the end of 2011. It was running during 2012, and it's not working anymore. Again, those APIs are unstable and dangerous. Those are a few examples of books that have been done by people.

This is an amazing book by [video cuts out]

[Video resumes] doesn't necessarily have to be a vanity press 2.0, but you can very well react to political events in the world, and you can put meaning into this.

This is a recent project that was done earlier this year in Geneva with students, and the objective was to make a useful book for graphic designers working with Libra open source fonts. So this is a specimen book designed entirely by students during one week of work using the book sprint methodology that Adam Hyde, who is in this room, is pioneering across the world.

So after the workshop, the book has been put on Amazon. Anybody can order it [video cuts out]

[Video resumes] mostly those are web fonts. There are about 200 of them in the book.

After those examples of exciting and enthusiastic and experiments with the technology, I'm going to look a little bit at the dark side because every technology that appears on the world also brings with it some dangers and some responsibilities. So actually, print on demand quite unexpectedly has unleashed very very strange phenomena that not many people I think have noticed but that have been invading the world of books and publishing during the past few years.

Who of you is aware of book spam? Roughly 30 percent. Who of you has printed or seen a spam book physically? Good. Excellent. Any of you has made a spam book? I want to talk with you after. Thanks.

So, I found this out. I came across this in early 2010 [video cuts out]

[Video resumes] thousand and some books in this inventory. All of them on very very specific and focused topics with interesting cover art sourced from stock photography databases. Here is a brilliant example of the history of the country Georgia in the former Soviet republic, which is adorned by a stock photograph of Atlanta the capital of Georgia. It gained a lot of attention, and there was a comment of an Amazon user who wrote, “When a book has the wrong image for the cover, you know this is a big indication that something is wrong.”

Another example, and you see a pattern familiar from internet spam where the title or the content contains a maximum of keywords. Those books spread from one platform to the other, so once it has appeared on Amazon it will infect virally other book platforms. Sometimes it will lose some metadata along the way. For instance, the cover will be missing, which will make it more and more hard to detect what is actually about.

So if we approach this scientifically, we can see that there are different types of book spam. Some of you have maybe seen the great talk by Nikolai today, so this is a kind of continuation of it. We have algorithmic spam where the entire book has been generated by a computer. Some of those authors are engineers, and they are proud of their invention, and they will talk about it openly like Icon Group International, which is producing those world reports on amazing topics like the “World Forecasts of Artificial Guts Made of Hardened Protein or Cellulosic Materials”. If you would tell me that this is an example of algorithmic poetry, I would certainly agree. This is a book search from this morning for this publisher. He has 870,000 and more books.

Another category even more intriguing is human-made spam where actually there weren’t algorithms in play, but there were real human agents who were involved in the creation of those books. It may seem unbelievable, but we have what is called outsourcing. That’s a type of work. This is from the front page of one of those publishers advertising its location on the island of Mauritius, and as an investigative data researcher, I figured out that there was a Facebook group launched by the editors themselves. For those understanding French, the subtitle “nu trop bon” is an interesting case of querization. Those are photographs of the workers who are churning out the books, and those are some of the comments that they shared on this group, which was of course deleted shortly later. By the way, it's also an interesting insight of what is happening behind all the data centers and all the digital services that surround us. To some people, if you have a service that is transcribing your text messages, this might well be happening in some other country through outsourced work.

In a way those workers are an incarnation of [video cuts out]

[Video resumes] a quick overview of the different elements in the real world that have enabled this practice. One is the convention of academic publishing that producers very very formalized books and items that have a strict design that can be easily imitated by algorithmic or by simple intervention. This is a legitimate scientific publisher publishing books probably also with print-on-demand, and you can see that with a quick glance, it’s hard to differentiate them actually from an artificial spam book that has been harvested or to figure out that is a legitimate scientific work.

Another key element is that open source free open license content is largely available on the web. Wikipedia is the biggest example, which is used here by those books, and those books comply to the rules. On the last page, they will show, in very very small print, all the names of the contributors and the full license on one page.

PediaPress is one cornerstone. This is a start up in Germany who made an amazing job, and they created a way of getting books out of Wikipedia. It's a legitimate business. It's an extraordinary idea. You can use it on Wikipedia itself or use it as a plugin on your own Wiki, and they are being used actually by those spam printers, as I found in this little Facebook statement: “PediaPress pa bon!” The servers of PediaPress were slow that day.

The last thing is that big book industries now have print-on-demand built in. Amazon, for instance, is printing a lot of books themselves. They will show them as being available, but actually once they get ordered, they will be printed in one or two days in one of their warehouses.

So, the print-on-demand spammer actually abuses those four cornerstones and exploits the system in a brilliant way.

Those books, unbelievably, have been infiltrating academia. You can find them in all university libraries around the world. Here in Switzerland I made a quick search. I found that there were 25 books by the publisher Alphascript about various topics: the judicial system of the People's Republic of China for instance or the legal history of China. So, those are printed Wikipedia articles on the shelf of that University. I borrowed some of those books from the University, and I brought them to the tombstone of Jorge Louise Borges in the hope [video cuts out]

[Video resumes] happening in our century.

It also means that the people who acquired those books failed the spam detection test. We all know the CAPTCHAs that we encounter every day on the web, so those are the touring tests that [video cuts out]

[Video resumes] thing is human-made or bot-made.

Now, what has Amazon done to counter this? I noticed that recently they managed to decrease the number of spam books in their database. They took countermeasures, and they have started to eliminate those books.

I had one interesting experience with this project. This was a book that I made. Again, with the original book sprint technology, I gathered a team during an art festival who was working during five days and producing an improved Wikipedia article on the art form of audiovisual performance (aka VJing). After one week of editing, we had a large article that we printed in book form, and I attempted later on to place this book also on Amazon through their print-on-demand service. After I submitted this, I received an answer that the title had been blocked and couldn't be published because the content was against the Terms of Service of Amazon create [video cuts out]

[Video resumes] you have to hold the exclusive rights, and the content shouldn't be available on the web. Or, in the original terms, we will not accept content that is freely available on the web unless you are the copyright owner of that content. The way it is phrased is interesting because the first part of the sentence has more weight. The freely available on the web content, by default, is banned unless you can prove that you are the copyright owner, and they insisted it should be exclusive rights. This is intriguing, and I think there would be many reactions here that would question the usefulness of this rule.

I think I'm reaching the send of my presentation and [video cuts out]

[Video resumes] it’s a bit too early in the morning, and I'm a practitioner and I'm saying thank you therefore.