With artificial intelligence commonplace, even in content creation, is the day of the human writer coming to an end? Is it time to upgrade or be deleted? The cyber invasion is truly underway, in fact, it has been for years. We encounter artificial intelligence everywhere. Be it smart technology, social media bots, or online shopping, so pervasive is it, that you will meet it everywhere.


Long gone are the days when leaving the office really meant keeping your work and private life separate. Today, smart devices leave us permanently connected to the office and all the cyber aspects of our lives. Everywhere we turn there is something vying for our attention. And it is somebody’s job to write all this content. After all, it cannot just generate itself. Or can it? As the demand for text increases, authors are struggling to keep up. People need fresh and relevant content. So the problem arises as to where these content editors can go to sate their readers’ hunger for material.

Meet your new artificial colleague

What you need is a reliable writer that is happy to sit in the corner churning out text. One that never wants to take the weekend off, drink coffee, or badmouth you on social media. Quite the opposite. It is possible to utilise an AI content-generating solution to tick all of these boxes. Not only can it produce the level of content necessary, it can also learn from the behaviour of those consuming it and adjust accordingly.

Something doesn’t quite add up

Computers are smart, really smart. But they don’t quite grasp all those idiosyncrasies that define what truly great writing is—be it rhetoric, irony, wordplay or just downright corny jokes. To understand this better it is a good idea to turn our attention to NLP, that is, Natural Language Processing. This takes the approach of analysing and synthesising how real speech and language work.

It is possible to split NLP into three types: inquiry, conversational, and reasoning.

Inquiry NLP

Defining when someone is asking for information through text analytic tools. By identifying question words, like “when, why, what, what, where, why, does, do, can, is”, to imperative use, e.g., “Give me a list of men’s clothes shops.” By using these tools, you can break the sentence into parts: subject, verb, object, manner, and place—giving a better understanding of the question’s nature—then crosscheck them against an ontology, for instance, www.schema.org, and use this to reply accordingly.

Conversational NLP (or Natural Language Understanding (NLU)) This technique engages the questioner in conversation to clear up any doubts. Further context is derived from the replies. Examples of this are Amazon Echo, IBM, Watson and Siri, which all use the notion of necessity and sufficiency.

Necessity—these are the factors that must be true for something else to be true: if P, then Q.

Here is an example from Wikipedia:

“For it to be true that "John is a bachelor", it is necessary that it be also true that he is

1. unmarried,
2. male,
3. adult,

since the state "John is a bachelor" implies John has each of those three additional predicates.”

Sufficiency—meaning that, according to the information given, there are enough grounds to assume something else is true: P implies Q. Another example from the same Wikipedia page:

“Stating that "John is a bachelor" implies that John is male. So knowing that it is true that John is a bachelor is sufficient to know that he is a male.”

For a real-life example, we can take Siri, Apple’s flagship AI bot. If I ask Siri, “Hey Siri, call Bob”, Siri answers, “I have found the following Bobs in your address book. Which one would you like me to call?” Through additional questioning, Siri can refine the request’s nature. It then uses the process of Necessity and Sufficiency in order to assume more precisely.

Reasoning NLP

An example of Reasoning NLP, the self-learning form of Natural Language Processing, is the MIT project Open Mind Common Sense (OMCS). This involves AI grasping more than just the physical forces of existence, such as dimensional attributes, mechanics and mass, or linguistic devices, such as pragmatics, syntax and semantics, to handle more abstract concepts like belief, emotion and culture. Using empathy by going beyond the application of pure logic means avoiding cultural predicaments and creating replies that are more “human”.

I am not content enough

Gartner estimates that by 2018, 20% of business content will be authored by machines. Through the use of analysis, the necessary data crunching can be automated to produce less creative content items, like report, and legal and financial documents. But even the best AI tools, such as Wordsmith, aren’t as automated as that. It is still necessary to create your template, add your data, preview your stories, and then issue them.

And where a lot of news items produced these days are machine generated, these machines cannot comprehend things like noticing the sentiment between the lines or voice consistency. These subtleties are either missing from the text or, worse, unnaturally inserted. And when it comes to the news, the human focus of the stories, with true emotional impact, missing these idiosyncrasies take away the humanity necessary to convey the stories with empathy. One quick scan read can highlight immediately that what you are consuming feels clunky and false.

Mr Robot, your desk is over there

Does this spell the end to the need for human copywriter? Can we now just plug your content writers in and let them go? Unless you only need to number and data crunch, then maybe. But truly creative NLP is still relies on human input to sprinkle on those subtle tinges to generate the rapport flesh-and-bone writers achieve. And that is the ultimate goal, connected and engaged readers.


By Duncan Hendy, content strategy manager at Kentico

Visit our website to see events that will help you keep up to speed on; Data protection, cyber security, digital marketing and business growth. View upcoming events here!

comments powered by Disqus