How Artificial Intelligence is Changing Writing
Will robots eventually have writers out of a job?
Are we going to see computers writing the next Great American Novel?
Probably not … at least, not any time soon.
Over the last decade or so, though, AI (Artificial Intelligence) has become increasingly sophisticated … and it’s influencing the world of writing in a number of interesting ways.
What is AI, Anyway?
AI is all about machines learning and adapting. Instead of simply being programmed in minute detail with everything they need to know to accomplish a particular task, they’re programmed with instructions that allow them to learn from their experience (just as people do).
There’s no one standard definition of AI, but ZDnet suggests some common features:
AI systems will typically demonstrate at least some of the following behaviors associated with human intelligence: planning, learning, reasoning, problem solving, knowledge representation, perception, motion, and manipulation and, to a lesser extent, social intelligence and creativity.
Here are six key ways in which AI is changing the face of writing … and reading.
#1: Translation from One Language to Another
In the past, if you wanted to translate a passage of text from one language to another – say from English to Spanish – you needed to find someone who spoke both languages.
Ten years ago, you could use a service like Google Translate, which essentially ran all the words through an English-Spanish (etc.) dictionary … with questionable and sometimes hilarious results.
In 2016 Google Translate had a major upgrade. Instead of translating word by word, it now translates more accurately by phrase or sentence – through an AI system. It even invented its own language to help.
Writers could potentially use Google Translate to translate their whole book into another language for free. (Note: I’m not recommending you do this, unless you have a native translator lined up to do some extensive editing!) As the technology develops further over the years to come, this could be great news for publishers and self-publishing authors … but worrying for professional translators.
#2: Automatic Editing and Proofreading
You’re probably very familiar with the red squiggly line in Microsoft Word (and other word processing programs) that marks spelling mistakes. There are lots of tools out there, though, that can go far beyond helping you spot typos.
Software like Grammarly, for instance, uses AI to spot overly wordy phrases, vague language, instances of the passive voice, stylistic issues, and much more.
This is great news for writers, particularly non-native speakers, who may need an extra helping hand with what they’re working on – whether it’s an essay, a blog post, or simply an important email. It’s potentially less great news for professional editors – but so far, no tool is able to provide the big-picture substantive editing that a good editor can offer.
#3: Checking for Plagiarism
One problem that universities and publishers deal with is the possibility of plagiarism. While a quick Google search for a couple of lines from a document can be enough to spot egregious forms of plagiarism, if a student or writer has changed, say, one in every five words, it’s a lot tougher to spot.
When it comes to students’ essays, they might be plagiarising from another student’s (unpublished) work, rather than a published source. This could be easy to spot within one institution – but not if the student has borrowed, or even bought, an essay from a friend at another university.
There are solutions out there, like Turnitin, that check submitted work against their vast database, flagging up cases where there’s a match between the submitted work and existing sources.
But AI is also increasingly being used in this area, with an AI bot called Emma Identity (reported on here by LifeHacker) being used to figure out the authorship of a piece of text. While this is essentially just a fun tool for now, it could eventually be used to combat plagiarism.
More worryingly, though, this type of technology could potentially be used to unmask authors writing under a pen name, if they’ve also written under their own name – or to uncover the authorship of anonymous posts on internet forums.
#4: Searching Through Audio Files
Although more and more content has been produced in video and audio format over the past decade, YouTube channels and podcasts haven’t diminished the amount of text online.
One huge advantage to text has always been that it’s searchable – and video and audio aren’t. If you want to find out a specific fact or dig into on a particular point of interest, text is definitely the easiest medium to work with.
However, audio search is becoming a reality – through the power of AI. Computers can increasingly decode sound – think of Siri, for instance, or Alexa – and audio search takes this further. Apps are already available: Castbox.fm, for instance, bills itself as “the search engine for spoken audio”.
What does this mean for writers? It’s not necessarily bad news. Newer types of text, like scripts for videos or outlines for podcasts, might become increasingly important. But as well as helping with audio search, AI could lead to even better dictation apps – potentially making it much faster to create written content, too.
#5: Crafting Breaking News Stories
You might be wondering by this point if AI can be used to actually write. Yes … but it’s not going to be producing works of literary wonder just yet.
Over the past couple of years, some breaking news stories have been written by AI, and there’s a great account of that in Wired here, explaining that a particular story was created by AI:
The dispatch came with the clarity and verve for which Post reporters are known, with one key difference: It was generated by Heliograf, a bot that made its debut on the Post’s website last year and marked the most sophisticated use of artificial intelligence in journalism to date.
The advantages for newspapers and websites are obvious: if AI can report on breaking stories, it makes it possible to get an article published almost instantly. No human needs to type a single word.
There’s a darker side to this use of AI, though: it could not only put journalists out of work, but it could also lead to a lot of low quality, derivative content – similar to content produced by low-paid “content spinners”.
#6: Influencing Readers’ Book Buying
Major online book retailers, like Amazon, rely on complicated algorithms to predict what books someone might be interested in, based on what they’ve already bought. If you’ve ever bought a book on Amazon, I’m sure you’ll have seen this in action!
This is generally seen as a positive use of AI: it helps readers to discover books that they’ll hopefully enjoy, and it helps authors to be discovered by readers who might not otherwise have come across their work.
Even so, it could be a drawback for, say, independent book stores and librarians: why consult an expert about what to read if an algorithm can recommend books to you automatically?
Whatever you write, there’s a good chance that AI will become more and more a part of your writing experience as time goes by – even if you barely notice it. Perhaps you’re already using a tool like Grammarly, for instance, or maybe you rely on dictation software to produce content quickly.
How do you think AI is changing writing? Do you think it’s a net positive or negative for working writers, editors, publishers and journalists? Share your thoughts with us in the comments below.