If I had a dollar every time someone asked me whether I was using ChatGPT to write blog posts and marketing content now, I would be mortgage-free and wearing a new pair of kicks (but probably still working, cause hey, I love it). The answer is a resounding no. We are not anti-AI, per se, but it, like many other technologies, is a tool and will not put copywriters out of a job.
I should note that while I am referencing ChatGPT throughout this article, the same limitations apply to other generative AI platforms like Google Bard AI, Microsoft Bing, ChatSonic, or similar services you have come across.
The reasons that AI-written content isn’t the answer for OLS are extensive, but they include;
- Inconsistencies – around this time last year, Gabe Marzano (inspired by an MIT Technology Review) asked ChatGPT what it knew about her. The answers were worryingly inaccurate, claiming that she was a he, that he (she) had co-founded OpenAi (incorrect) and that he (she) was the CEO of Synthestra Technologies (incorrect). When asked where this information was sourced from, the answer was “various articles and interviews featuring Gabe Marzano, as well as his official LinkedIn.”
These inaccuracies likely relate to ChatGPT’s propensity to predict plausible text or information based on the user’s text. It uses statistical patterns in its training data and the context that the user has provided, but as seen in the example above, this is a long way from foolproof.
- Plagiarism – last month, the New York Times announced that they were considering suing the Chat GPT developer, OpenAI, over the use of copyrighted content to train ChatGPT’s system. This announcement comes in the wake of a change to their terms of service to stop AI companies from using NYT content for training in the future.
- Americanised speech patterns and diversity issues – just like global content doesn’t work in every market without a level of localisation, Chat GPT has specific default settings that dictate what it views as “standard” and what it views as non-normative. While yes, you can input specific instructions to generate different languages, such as Australian English, there are some problematic consequences of having one dominant norm being perpetuated at scale. For more on this topic, this article on The Conversation is definitely worth a read.
- The need to build an authentic brand voice - There is a reason that tools can be deployed to identify AI-generated text vs. human text – it just isn’t sophisticated yet that it can replace human storytelling and experience. I say yet because artificial intelligence is learning and improving exponentially, and who knows what the future holds? But for us to nurture and grow your brand, we need to create content that is interesting, informative and even entertaining. And that sounds like your brand, not generically generated text.
- In-depth industry knowledge - As industry experts in the IT channel and technology sector, we understand the nuances, politics and broader contexts surrounding the APAC market at a level that AI does not.
ChatGPT isn’t the first time that AI has impacted marketing. We use AI frequently in different contexts. Want a quick grammar and spelling check? Grammarly is super handy. Using Google Analytics? It is AI-powered. Plus, let’s not forget the plethora of email marketing automation tools.
But what role does generative AI (the kind that writes content) play in the future of marketing? Here are some of the advantageous areas.
Just like Googling a topic can spark thoughts in different directions, so can ChatGPT.
Brainstorming and ideas
One of the core pillars of marketing is creativity. With around 334 million businesses globally, how can you ensure that you are speaking to the right people at the right time in a relevant and memorable way? There are thousands of books and theories about cultivating creativity. In my experience, much of it comes down to time, collaboration, risk and experimentation (you can read more about my take on controversial and provocative advertising in the B2B space here).
But working in a fast-paced industry, you don’t always have as much time and opportunity for collaboration as you might like. And we all get the odd mental block where we’re not feeling creative. At times such as these, it can be super helpful to bounce ideas around. Sometimes, I jump on Teams and speak to one of the other Little Seeds, but other times it is handy to have a tool to generate different ideas. Just like Googling a topic can spark thoughts in different directions, so can ChatGPT. While the ideas it throws out are unlikely to be the ideas that end up being used (see points 4 and 5 in particular), it can be used to break down that mental block and help you consider alternative directions.
Data-driven analytics
There is huge potential to integrate AI further to leverage data for better decision-making. It is already being used to try to personalise marketing outreach based on each customer’s behaviour. However, this approach relies on high-quality, accurate and integrated data with any bias removed. And unfortunately, a lot of companies struggle to ensure they have a clean, up-to-date, single source of truth, so this is the first port of call before investing in fancy new AI technology.
So, while AI technology certainly has its place in B2B marketing strategies of the future, be wary of assuming that it can write your content and make smarter decisions for you. We are a long way from Terminator or Matrix-esque super intelligence and will be tapping out our blog posts the old-fashioned way – one word at a time.
