Rise of the copywriting machine – Will AI replace writers?

Artificial Intelligence (AI) is getting amazingly good at writing. There’s always been a bit of mystery surrounding copy writing and how it actually works. Recently though, there’s been a lot of hype around the idea of machines and AI helping us to craft the words we use

Is there anything wrong with the above paragraph? Technically, there is nothing wrong with it. It has a somewhat catchy opening sentence followed by a matter-of-fact tone when talking about AI helping “us” craft words. Seems quite ordinary, really, as if it could have been written by any human. But… How would you feel if this writer told you that she did not write them and that the words you just read were written by a bot? That same bot also suggested 5 other introductory paragraphs that I could have used that were logical and befitting of an article that is exploring if AI will one day replace humans.

Very meta indeed.

Copywriting AI exists and is currently being used to help writers, marketers, editors, and anyone who needs content for web or social media purposes to churn out a few words that they can use. At the core of its technology, it combines machine learning and natural language processing to analyze the keywords that one puts into the copywriting AI tool which will then generate the words that one might need. The scary thing about this is that the content that the tool generated made sense. So much so that if I were to give it more specific keywords and let it generate more content, all I probably had to do to churn out an entire article was to do a little fact-checking and tweaking to suit the direction I wanted to take, and an entire decent article could have been generated with only me knowing that I did not write those words.

In an interview with TechCrunch, Hypotenuse AI’s Singaporean founder Joshua Wong explained that he came up with the idea for his version of the copywriting AI after getting frustrated with writing large amounts of effective copy when setting up his friend’s website that sells vegan soap. All they wanted to do was sell products, but they were wasting time trying to describe their products. Thus, Hypotenuse AI was born. It can generate texts using machine learning models or neural network models that could speak very fluently in a very human-like manner. Users are also able to choose the type of style they want for their brand. A few simple clicks and the algorithm takes care of the rest.

Wong’s vision to use a copywriting AI to generate content for product descriptions can be an extremely useful tool for business owners and marketers who have large amounts of copy to generate but do not necessarily need a dedicated copywriter to do so. A machine is there for you at any time you need it, and it can churn out multiple versions of legible content for you to pick and tweak to your fancy.

Ai Writers
An example on hypotenuse.ai showing the efficacy of their technology.

Copy.ai, another copywriting AI service, offers a free trial of their technology and all you have to do is give your email address. I gave it a try and within 2 minutes I was exploring the options they had to generate different types of content from social media ads to web copy to a motto generator. Within 30 seconds of putting in my keywords, I had my introductory paragraph. If I wanted to, I could give the AI more keywords and keep getting it to generate content to the point where I was satisfied with it. Indeed, it was like I was playing a game, picking and choosing what content I felt was appropriate. Does imposter syndrome even apply here?

Far from trying to replace human writers, the technology does have very practical uses and is already being used in major publications with the purpose of making things easier for their staff. In 2014, The Washington Post used its own AI called Heliograf to generate updates on the Rio Olympics so that its reporters had more time to write in-depth analysis and give insights in ways that only they can. Technology like Heliograf could also be used to automatically update a story that was previously written. For example, a story published on Tuesday could have some updates. Instead of getting a writer and editor to rewrite and republish the story, which could have some unpredictable turnaround time resulting in a republishing perhaps only by Wednesday. With just some data, Heliograf will be able to help update it with the most recent facts. At this point, depending on the settings chosen, the story could wait for an editor to approve it, or it could automatically be updated by the system immediately. In today’s fast-paced environment, automation like this could be very useful as it gives readers access to the latest information plus helps reporters free up time so they can concentrate on other tasks.

As this technology evolves due to humans developing it as a way to help improve our lives, will we see a rise of the machines? Will it come to a point where the technology is so sophisticated that full time writers will no longer be needed? Or will it come to a point where a news publication must state whether an article was written by a human or AI? As of now… this writer does not think so. Robots have yet to learn how to truly be human. Yet. We still need a human to capture emotions, sentiment, and context whether it is to sell things or reporting on a major news event that might have nuances that an engineer has not figured out how to teach a computer yet. For now, I say let us soak in this atmosphere where humans and robots are working together to try and make the world a better place.

(This article was written by a human. Really.)


Note: The opinions expressed in this blog post are those of the author. They do not purport to reflect the opinions or views of the Asia Internet Coalition (AIC) or its members.