In 2016, a sci-fi novel, rather unimaginatively named ‘The Day a Computer Writes a Novel’ and co-created with machine learning by Hitoshi Matsubara and his team at Future University Hakodate, was an entry for the Nikkei Hoshi Shinichi Literary Awards. The contest is named after the famous Japanese novelist and science fiction writer, Shinichi Hoshi, and it accepts entries from non-humans. Of the 1,450 novels that were part of the 2016 edition, 11 were co-written with non-humans.
Today machine learning systems have graduated to become co-creators of content. The New York Times article Computer Stories: A.I. Is Beginning to Assist Novelists describes how machine learning software completes author Robin Sloan’s snippets of text to form interesting sentences. Google already uses machine learning to help you write typical email responses.
Clearly, machine intelligence is moving ever deeper into creative spaces. ‘The Day a Computer Writes a Novel’ did not win but there is growing apprehension that machines will completely take over the creative process and churn out great content — from articles to novels or magnificent symphonies.
It may happen, and even though it won’t be any time soon, the time is right for content writers to get acquainted with, if not embrace, the field of machine learning.
Content and machine learning seem unlikely bed fellows. Content is ensconced in the warmth of human interaction, while machine learning inhabits the cold, technical world of computer science. Nevertheless, there is a symbiotic relationship between the two, especially in the field of communications.
Machine learning has forever changed the way content is created and marketed. In machine learning, a computer analyses data, identifies trends and arrives at decisions without being explicitly programmed to do so. It thinks and it learns! To arrive at optimal conclusions, machine learning systems have to be systematically and skillfully trained, and that is the aspect that concerns content creation.
Let’s decode how machine learning impacts content creation by looking at the three basic tenets of content writing — who you address, what is the message, how do you package it.
All good writers and visualisers know that for an article to be effective, it has to target the right audience and have the right ‘hook’ to attract attention. Machine learning comes into play here. It helps you focus and filter through analytics and search results, to keep you on top of trending topics. It also helps you closely track audience reactions and thus develop a better understanding of what will appeal to your target audience — in terms of language, tonality and ‘hook’.
When it comes to crafting the message, the content writer has to dance even more closely with the machine. The article has to be populated with appropriate keywords so that it stays on top across search engines. Here, SEO optimisation tools such as WordLift, engineered on machine learning, come to your aid. They help you identify gaps in your content and advise you on keywords, tags, links and interlinks, data visualisation, images and so on.
Once you finish your article, machine learning tools can serve as your private editors who are on call to run an eye over your article and suggest qualitative changes. Online content editing tools such as Grammarly and Acrolinx give you grammar, style, orthographic and voice tips to polish your article and make it more relevant to your audience. Once the content is published online, machine learning tracks the data related to your article, analyses it and gives you a fair idea whether it has met its target.
By now the answer to the question ‘What does machine learning have to do with creative content?’ must have become crystal clear. Right from conceptualisation to creation to fine-tuning to results, machine learning is involved at all stages of the modern content creation process. What’s important for the content writer is to remain in the driver’s seat and use machine learning tools to make content creation simpler, faster and better.