As the AI evolution continues to surge, Affiliate Leaders magazine gained the Martyn Hannah, Founder & MD of Comparasino and Sue Dawson, Head of Content at FTD Digital who revealed whether they have re-evaluated their approach to content.
Firstly, can you tell us more about how AI has impacted what you do on a day-to-day basis?
Martyn Hannah: It hasn’t really. At Comparasino, we have a team of online casino experts that produce the content we publish for players in the UK. This might make us less efficient than those who are deploying AI solutions for content writing, but we prioritise quality over quantity.
People use Comparasino to find new online casinos and bonuses because we have a reputation for being trustworthy and reputable, and we would absolutely do damage to this perception if we were to roll out AI-generated content. Having read plenty of AI content over the past few months, it’s clear to me that content written by human experts will always offer more value to consumers.
Sue Dawson: It’s caused us to re-evaluate our existing content and when creating new content, to concentrate on authenticity. And not just for Google, but for people. Users are now aware that there is a large amount of suspect AI content out there so author credentials have become even more important.
We have found quite a few use cases for AI other than creating content, however. AI can be very useful as a management tool, for collating and summarising information from different sources.
How do you believe AI can support firms in bringing engaging content to their audiences?
MH: Having just said we don’t use AI to generate content, we do use AI-powered tools to help improve the quality of the content we publish across Comparasino.
Despite some initial reluctance to use it, I have become a huge fan of Grammarly. While we have an extensive proofreading process in place to ensure our content is factually correct and free of errors, it’s impossible to spot everything and this is where Grammarly has proved to be incredibly effective.
For content to engage, it must be written by human experts. This is the only way to bring a deep understanding of the subject matter to the reader and these experts also know how to turn often complex subject matter into a relatable article or guide for players to read. For me, AI tools like Grammarly simply help to ensure such content is error-free.
SD: AI can be used for collating information from competitor and keyword research and helping us ensure that our content covers everything relevant our competitors include. It’s still up to us to check that our content is factually correct and put a unique spin on it to add value.
In what ways can AI evolve the job of writers, could ChatGPT really replace writers when it comes to producing content?
MH: There will be some that say AI and in particular ChatGPT can be used to help writers generate content ideas and find relevant information for their articles, but I’d argue that the team behind any online casino comparison site should be made up of experts that hold this knowledge and understanding of the market without the need to use AI.
For me, ChatGPT and the like will never be able to deliver the deep knowledge held by subject matter experts, nor will it be able to produce copy with a unique turn of phrase in the way that different people have different writing styles. The best content will always be produced by human experts. Period.
SD: ChatGPT can and already has replaced bad content writers, but I would argue that it has made good content writers even more valuable. I expect the job spec for writers to evolve to put much more emphasis on the writer being a specialist (and preferably an expert with demonstrable credentials) in the subject matter and I also expect content production agencies to become more and more specialist in what they cover.
The very first practical task I used ChatGPT for was to help screen prospective freelance writers. The freelancer and ChatGPT were given the same brief, on a subject I already knew well enough to be confident of spotting any factual errors. To be hired, a freelancer needed to produce something at least as thorough as and more engaging than ChatGPT, with no howlers. Very few did.
Having said all this, there are already some situations where ChatGPT can be used successfully to produce a large amount of content quickly – for example, if you’re a large e-commerce site and you need to produce a lot of product descriptions from already documented product specs.
Even that needs a degree of human oversight, though – I’m sure you’ve all seen dodgy AI written product descriptions on Amazon! As AI develops further, it’s likely to get better at producing different types of content and I’m sure there will come a day when the amount of editorial control I have to exercise over ChatGPT is no more than I’d want to exercise over any other writer.
