Computer Generated Fiction

If computers are able to write content that is indistinguishable from human authored writing, what will this mean? If they can one day write anything from travel guides to literary novels, will people read them? Would they trust that a computer knows where the best restaurants in Atlanta are or the best hotels in Paris? What about literature? Humans plumb the depths of their emotions and draw on extraordinary experiences to create great literary works; could a computer possibly do the same? If this comes to fruition and people can distinguish between a human and a computer, which will they prefer? If they prefer the computer-generated content, what would this mean for authors?

Starting with what computers can already do, we can look at information compilations such as travel guides. People will read them because computers can pull human reviews from the internet and compile more reviews more quickly than any human ever could therefore making it the best authority on anything that is subject to reviews. It will likely have a better review for you than your best friend’s actual experience.

Once computers have advanced past that stage and on to simple or formulaic literature, humans will not care who or what wrote it as long as it is entertaining. The most common formulaic genre is romance, and, out of the paperback fiction category in North America, romance novels are the best-selling.  Computers can be fed formulaic plot lines and stock characters to work with, and will be able to read through a million novels to get ideas on which words and phrases are liked best by humans. Humans already read extremely formulaic books and will not mind if a computer writes them instead of a human, especially since the computer is pulling from sources written by humans.

Other formulaic examples come from syndicates like the Stratmeyer Sydicate that put out Nancy Drew and The Hardy Boys. The Stratmeyer Syndicate is an excellent example of why computers could become “fiction factories”, as the syndicate was known. Founder Edward Stratmeyer hired unknown writers, gave them anything from a few sentences to a three-page outline and a plot, and expected to receive a finished book two weeks later complete with chapter cliffhanger endings and consistent sounding dialogue.

Computers will soon be able do even better – as of today they can write fiction that is almost comparable to that written by humans. By having access to every book and online resource possible, they have access to almost all human documentation thus far, giving them the power to not only get ideas and phrases that have received positive human feedback, but also millions of human experiences, and what emotions these evoked. Alexander Prokopovich’s algorithm wrote its own version of War and Peace, entitled True Love, in 2008. It sounds close to a lot of human writing apart from the odd phrase or two: ‘Kitty couldn’t fall asleep for a long time. Her nerves were strained as two tight strings.’ The Georgia Institute of Technology has developed a program called Scheherazade that can write fiction that sounds convincingly human. For example:

John took another deep breath as he wondered if this was really a good idea, and entered the bank. John stepped into line behind the last person and waited his turn. When the person before John had finished, John slowly walked up to Sally. The teller said, “Hello, my name is Sally, how can I help you?” Sally got scared when John approached because he looked suspicious. John pulled out a handgun that was concealed in his jacket pocket. John wore a stern stare as he pointed the gun at Sally. Sally was very scared and screamed out of fear for her life. In a rough, coarse voice, John demanded the money. John threw the empty bag onto the counter. John watched as Sally loaded the bag and then grabbed it from her once she had filled it. Sally felt tears streaming down her face as she let out sorrowful sobs. John strode quickly from the bank and got into his car tossing the money bag on the seat beside him. John slammed the truck door and, with tyres screaming, he pulled out of the parking space and drove away.

Robot fiction reviewer Nicholas Lezard actually thought it was an excerpt from a new Dan Brown novel, but then realised Scheherazade could have been programmed using algorithms based on Brown.

Over the years people have come up with tests to see if a computer can pass as a human. One such test is the Turing Test.  Invented by Alan Turing, it consists of a human sitting in a room at a terminal with a computer, and a computer at a terminal in a separate room. The human corresponds via text with whoever or whatever is in the other room, and then he or she has to figure out if he or she is corresponding with another human or with a computer. So far no computer program has definitively passed this test. People have come up with alternate Turing Tests where people read different articles or stories and try and figure out whether a human or a computer wrote them. One such a test can be found at http://www.nytimes.com/interactive/2015/03/08/opinion/sunday/algorithm-human-quiz.html.

As times moves on, humans will be unlikely to prefer one type of writing over the other. Some people will happily read formulaic, computer generated novels, others will be intrigued and will voraciously read literary novels written by computers, while traditionalists will stick with their human written works.

If the majority of humans ever do prefer computer-generated content this will affect authors because they will be less in demand. If a computer can write the new Dan Brown while authors are working on literary novels, which already don’t sell as well as thrillers, they may lose out on work. That being said, there will always be traditionalists so computers writing fiction may actually push human authors harder to compete, drawing forth some of the best literature we have ever read.

As of right now, although they are close, computers cannot write fiction equal to that authored by humans. “The hardest [for the computers] to crack will be the elements of great writing we ourselves struggle to explain: the poetic force of the sentences, the unique insights of the author, the sense of a connection.”

 

Sources:


http://www.scientificamerican.com/article/computers-vs-brains/

Studying the Romance Novel

http://www.trussel.com/books/strat.htm

www.bbc.com/culture/story/20150122-could-a-robot-write-a-novel

http://www.theguardian.com/books/2014/nov/11/can-computers-write-fiction-artificial-intelligence

http://psych.utoronto.ca/users/reingold/courses/ai/turing.html

 

The Authors Guild Went too Far to Eradicate Amazon’s read-aloud Feature

E-readers like Kindle should be allowed to have a read-aloud feature for their e-books in a text-to-speech (TTS) format. The Authors Guild was wrong and petty in asking them to remove it. Read-aloud TTS features do not violate any copyright laws, and are an aid to the visually impaired and useful to commuters. By asking Amazon to disable the TTS function on its Kindle 2, which was released in 2009, they stifled innovation and made themselves laughable by incorrectly interpreting copyright laws. The Authors Guild took things too far to prove a point. They also claimed that it infringed on audio book rights.

Traditionally, audio rights are a subsidiary right, which “refers to licensing agreement provisions for copyrighted material published in derivative formats, where licensed publishers are granted legal authorization to publish or produce copyrighted media.” Amazon’s Kindle 2 had a function called “read-aloud” in a TTS format, but almost immediately the Author’s Guild accused them of violating copyright and wanted to take Amazon to court. “The Authors Guild objected to the text-to-speech function, saying Amazon doesn’t have the right to essentially turn e-books into audio books.” ““They created a hybrid product,”” Paul Aiken, executive director of the Authors Guild, said when reached by phone late Friday. “It was being used in a way they had not been given permission for.”

Amazon quietly capitulated without any legal proceedings because they didn’t want to deal with the hassle and legal fees. It isn’t hard to see why as “with Google, the Authors Guild managed to score a $125 million settlement and arguably interfered with fair use rights under copyright law.” But had they gone to court, Amazon would have won.

For Amazon to be violating copyright, they would have to create a derivative work of an original work, or create a reproduction, or unlawfully distribute copyrighted materials. A derivative work is “a work based upon one or more preexisting works . . . which, as a whole, represent[s] an original work of authorship. A book being read aloud by a machine is neither creative nor is it independent of the original work. To create a reproduction, said reproduction must be in a fixed format. Since the Kindle TTS application is not being recorded, there is no fixed format for which to reproduce the work. This is like saying that if I read a book out loud I am violating the right of creating a reproduction. Many journalists, scholars, and even authors have jumped on this point. “Lawrence Lessig, founding board member of Creative Commons, points out that by allowing the Authors Guild to prevail, ““publishers get to control a right which Congress hasn’t given them—the right to control whether I can read my book to my kid, or my Kindle can read a book to me.” Amazon was also not unlawfully distributing materials. They have a licence to sell their Kindle e-readers, and they are not distributing audio copies as those are made up on the spot.

There is also the question of TTS itself. “There shouldn’t be anything controversial about TTS: it’s been available on personal computers since the 1970s. It’s important to people who have impaired or no vision, but little used by anyone else. However, the Authors Guild argues that the audio rights for a book are different from the reading rights, even if the audio is provided by a software robot.” Some may argue that photocopiers have been around just as long, but in truth that is an entirely different beast. A photocopier always makes a copy in a tangible, permanent form, whereas as TTS does no such thing.

Many lawyers have upheld the fact that Amazon did not violate any copyright laws. “Ben Sheffner, a Los Angeles copyright attorney and author of the blog Copyrights & Campaigns, said Amazon probably reversed course to maintain good relationships with authors, not because of legal concerns. Sheffner said that Amazon probably wouldn’t need different rights to sell an e-book with the text-to-speech function enabled.” It is sad that Amazon did not fight for TTS on their Kindles as even though no legal precedent was set, a recognizable precedent was set nonetheless.

On March 2nd, 2009 the Author’s Guild website announced that “at the end of the business day on Friday, Amazon announced that it would allow publishers (and thereby many authors) to block text-to-speech audio functionality on a title-by-title basis for its Kindle 2 reading device.” They lauded it as a great victory for authors everywhere.  But was it really? What did they actually achieve? As author John Scalzi noted: “Has it escaped the general notice of folks that the same company that is putting out the Kindle is also the same company that owns Audible.com? Yes, Amazon owns both, and I don’t really see the company trying to put one section of itself out of business with the other.” Why would Amazon undercut their own audio book sales, for which author are compensated.

The Authors Guild also claim that the read-aloud function violated audio book rights. By their argument audio books are a feature of an e-book as well as they a separate product. They cannot be one and the same. An audio book, as defined by Macmillandictionary.com, is “a book that is read out loud, usually by an actor, and recorded as an MP3 file or on a CD etc.” This definition clearly excludes TTS. Not to mention that a real audio book is far superior to a TTS “audio book” (Stephen Fry read all seven Harry Potter books aloud for Audible with a masterful skill at giving all the voices different voices). A machine cannot read with proper pauses or intonation.

Lastly, not only did the Authors Guild wrongfully ask Amazon to disable their TTS feature, they also stifled innovation, which is a worrying trend. As Lawrence Lessig points out “the bigger trend here is much more troubling: Innovative technology company (Amazon (Kindle 2), Google (Google Books)) releases new innovative way to access or use content; so-called “representatives” of rights owners, Corleone-like, baselessly insist on a cut; innovative technology company settles with baseless demanders, and we’re all arguably worse off.”

E-readers like Kindle should be aloud to have a read-aloud TTS function. The Authors Guild was wrong to want to take Amazon to court over this innovation because Amazon did not violate any copyright laws, or any audio book rights. This reversal on Amazon’s part sets a precedent, and makes consuming books harder for the visually impaired and commuters. The Authors Guild took things too far to prove a point.

Sources:

http://www.techopedia.com/definition/24922/subsidiary-rights

http://www.bitlaw.com/source/17usc/101.html

http://consumerist.com/2009/03/02/amazon-allows-publishers-to-kill-text-to-speech-function-on-kindle-2/

www.macmillandictionary.com/dictionary/british/audio-book

http://www.lessig.org/2009/02/caving-into-bullies-aka-here-w/

http://articles.latimes.com/2009/feb/28/business/fi-kindle28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Magazines and Metrics

Everywhere you look in the magazine world people are talking about data and metrics. What kind of data are we talking about? Data on people’s age, their income, their marital status, whether or not they own a dog or a cat or a ferret or a rabbit, how much they time they spend outdoors, which is then broken down even further into which outdoor activities they enjoy most. The list goes on. The data used by different companies is known as metrics. Magazines use key metrics to determine what their audience wants to read and which advertisements to run. Every magazine wants to make money, and in the past publishers strove to create the best content possible in order to sell the most magazines. Of course they had their readers in mind, but they weren’t slaves to data endlessly counting how many “likes” their Facebook page received or how many times their tweet was favourited.

Although data has been collected for years, it was never on the scale that it is today, and it didn’t have the extreme impact on what was published in magazines. Today metrics are used too much and they negatively affect magazine publishing by determining the entire content of a magazine from editorial to advertisements to advertorial. Publishers no longer think in terms of the best content selling their magazines, but rather the content best tailored to their audience to sell the most magazines. The readers have inadvertently become the content creators.

The Print Measurement Bureau (PMB) started collecting data in 1973 and they still collect mass amounts of data today, but the process consists of “a two-stage interview process: a personal, in-the-home interview, followed by a leave-behind questionnaire” (pmb.ca) which is time-consuming and costly. The studies are “based on a national stratified sample of approximately 22,000 Canadians measuring over 100 publications, consumer exposure to other forms of media, and consumer usage of over 2,500 products, services and brands.” (pmb.ca) While this sounds impressive, it is nowhere near what publishers collect now.

Publishers now collect data from many sources, and it is easy to do so as “by one estimate, more than 98 percent of the world’s information is now stored digitally, and the volume of that data has quadrupled since 2007. Ordinary people at work and at home generate much of this data, by sending e-mails, browsing the Internet, using social media, working on crowd-sourced projects, and more.” (theatlantic.com). Companies like Facebook and Google have way more access to data than publishers do through methods like people voluntarily giving their information to Facebook and Google analyzing their gmail users’ emails. Facebook and Google, along other companies of the same ilk, use this information to decide which ads to place where and to whom to direct them. Data has a much more insidious role in magazine publishing: it produces content based on metrics instead of creativity, results in crafted advertising, and produces advertorial, which is literally advertisers paying magazines to write about them.

Publishers have long been thought of and expected to be cultural gatekeepers. Literature is meant to inform people and to help them think more critically. If magazines no longer publish with this in mind, they cannot help but publish tired material simply because they know it will interest their readers in some way. But if you keep publishing the same ideas over and over again, not a lot of new or creative material will emerge. This, of course, excludes news and entertainment magazines, as news magazines publish whatever is current and entertainment magazines publish simply to entertain, and not to encourage critical thought.

People tend to trust what is in print over other forms of media. They know it has been sourced and fact-checked. This means that they unknowingly also trust what advertisements are chosen for the magazines. But choosing advertisements based on readership could have a seriously negative effect. In 2012 the New York Times stated: “Data aggregation has social implications as well. When young people in poor neighborhoods are bombarded with advertisements for trade schools, will they be more likely than others their age to forgo college? And when women are shown articles about celebrities rather than stock market trends, will they be less likely to develop financial savvy? Advertisers are drawing new redlines, limiting people to the roles society expects them to play.”

Magazines are also drawing redlines. They are limiting people as to what they read. One could argue it is up to individuals to read literature that expands their minds and ignore advertisements, but it is all but impossible for this to happen. People have been conditioned over many years to implicitly trust print, especially newspapers and magazines.

With magazine publishers now obsessed with churning out editorial, advertisements, and the dreaded advertorial to conform to their metrics, quality content has been lost. It is now expected and accepted for magazines to have 50% editorial content and 50% advertisements and advertorial. Believe it or not 50% is actually quite a high percentage of editorial content – many magazines are up to 70% advertisements and advertorial.

One magazine that has perfected the art of catering completely to metrics is Vice Magazine. Yes they are enormously successful, but what do they actually publish? Vice has established an empire and “the magazine is [now] less than 5% of [their] total revenue”. (marketingmag.ca). Completely ironically, the previous quote was sourced from an online magazine – marketingmag.com – and the article about Vice is promoting its business model. They espouse Vice’s mantra of creating as much content as possible. But what kind of content is this? In CEO Shane Smith’s own words, the content is shit. He says magazines should ““make piles of content. Make shitloads of content. Music, fashion, food, fucking booze, news, do whatever the fuck you can because all of it is going to be worth money.” The CEO is excited when he hits this part of the sermon: the revelation that Vice’s predilection for “making shit” has become his company’s main revenue source.” (marketingmag.ca)

With the new business model being bowing down and worshipping metrics and relying on them to dictate what editorial content and advertisements to publish, magazines are no longer producing informative and thought-provoking content; it is just a load of crap.

Sources:

“Print-measurement-bureau-canada.” LinkedIn.com. Linked In, n.d. Web.

“PMB Print Measurement Bureau – PMB 2013 Fall Study Overview.” PMB Print Measurement Bureau – PMB 2013 Fall Study Overview. N.p., n.d. Web. 29 Jan. 2015.

Peck, Don. “They’re Watching You at Work.” The Atlantic. Atlantic Media Company, 20 Nov. 2013. Web. 29 Jan. 2015.

Andrews, Lori. “Facebook Is Using You.” The New York Times. The New York Times, 04 Feb. 2012. Web. 29 Jan. 2015.

Gierasimczuk, Tom. “Vice Age.” Marketing Magazine Vice Age Comments. Marketing Magazine, 08 Feb. 2012. Web. 29 Jan. 2015.