How to get the most out of ChatGPT

While ChatGPT likely won’t provide identical word-for-word responses to every online user who asks the same question, the fundamental information and facts that are generated remain mostly the same. This means the internet now finds itself awash with swathes of almost identical content, sending Google’s algorithm into a tailspin as it tries to rank content that by appearance is perfectly optimised for SEO, but is in fact unoriginal. And originality and authenticity of voice has always been a critical factor in ranking well on Google.

What ChatGPT is great for is getting the ball running with content creation if you’re drawing a blank. Think of it as writer’s block’s best friend - filling your initially blank page with lots of relevant ideas you can explore further in your own voice.

TIP: Chat GPT responds well to clear and specific queries. So make sure your prompts are as precise as possible, clearly stating the context, target audience and your expectations so that it can generate the most accurate and relevant responses.

While ChatGPT will spit out human-like content that looks perfectly coherent and relevant at first glance, it lacks the depth and personality that come through in genuinely original work.

Brand voice is what makes your organisation recognisable and relatable. All content produced by your brand (and for your brand) should have the same brand voice running through it. This is why it’s so important to interrogate ChatGPT’s outputs and use it to craft something original, from your own brand’s perspective.

Like any AI tool, ChatGPT has flaws in accuracy. The tool has been widely known to generate misinformation – in one case it even invented a sexual harassment scandal and named a real law professor as the accused. A fake Washington Post article was cited, which appeared to corroborate the false claim. Gulp. So not only is the tool capable of giving inaccurate information but it can also fabricate primary sources to back up the claims.

Always cross-reference any facts or information generated by ChatGPT with other trusted online sources. This will help avoid falling into the trap of accidentally misinforming your readers or spreading fake news.

Several studies have reported that ChatGPT’s responses have a clear left-leaning political bias. This is partly down to the training data used – made up of 60% internet crawled content, 22% from curated internet content, 16% from books and 3% from Wikipedia. A large portion of this training data will be from biased sources. This bias also lies in the fact ChatGPT is shaped by reinforcement learning with human feedback (RLHF). This process uses the views of people providing feedback to help shape the model, who will inevitably have their own biases.

Being aware of ChatGPT’s underlying bias is important, particularly if you’re an organisation who needs to remain politically neutral in the content you create.

While ChatGPT is a groundbreaking tool in aiding the content creation process, it’s clear that the content it generates should be treated very much as a springboard rather than a final destination. By using it in this way, we can harness the potential of AI while also preserving the values of authenticity and originality in the content we create.

If you found this article interesting, check out our blog on how to write powerful and engaging copy.