How Not to Make an Ass of Yourself With AI
There is no doubt, AI is a powerful production tool, but it is not a complete content strategy.
Let’s be honest: training gave you the elite clinical skills to transform your clients' lives, but absolutely nothing in your schooling prepared you for the relentless, firefighting chaos of digital marketing.
When you are already drowning in a sea of admin, it is incredibly tempting to dump your raw thoughts into a chatbot, hit 'improve', and hope for the best. But unless you know how to steer these tools, you are trading your unique, hard-won authority for sterile, automated copy that looks exactly like everyone else'. The model often strips away the exact human flavour that makes your blogs and posts so compelling.
If you are using AI to help manage volume of work you marketing demands, keep these checks in mind before you hit publish:
Are you really that peppy? (I mean, I am…. But honestly, are you?) sometimes even I ask my AI models to tone it down a bit! 😂
Have you actually used the word tapestry in real life since you were 10 years old and went on that stately home visit? If AI is using words you wouldn't use, make sure you swap them out.
I love 💖a well-placed 📍emoji 😊 as much as the next ⏭️ guy🙋♂️, but AI is absolutely addicted!
Is the response completely agreeing with your point of view? Watch out, you are dangerously close to being funnelled into echo chamber territory! AI is a 'yes man', unless you give it permission to ask questions or challenge, it is going to do its best to agree with you.
It loves to repeats itself, if you don’t give it enough input, it will fluff around repeating itself because it loves it. (See what I did there? 😂)
Does the punctuation and spelling make sense for the region you live or the language you speak.
Are you confident you know enough about the topic to fact check? AI can get it wrong in so many ways.
Your chat becomes tired and confused when overused. Yes, this is a thing!
Your chat can hallucinate. It will fill in the gaps when it doesn't have the information, and its responses don't always make sense.
Your chat doesn't know a good source from a bad source, and it's not fussy about where its responses come from.
It essentially learns from the popular middle ground and recognises patterns, so will try and fabricate a matching answer if it doesn't know.
It can be lazy and give you half the answer.
There is no substitute for human insight so when researching or outlining, a human must always sit above the loop to steer, fact-check, and inject real-world, experiential evidence.
Most default AI platforms use your inputs to train their future models, which means your proprietary insights, or worse, sensitive client details, could inadvertently leak into the public domain. While enterprise-grade tools are shifting toward privacy by default, always treat your data as a vital business asset. Don't feed the algorithm your competitive moat.
For images
Does the background context match the foreground?
What is the spelling like? Could be embarrassing.
Do you look so perfect… like you but not quite you, people will notice.
Are things out of place or are there duplicates of things?
Training gave you your elite skill as a practitioner. Don't let a chatbot dilute it. Use technology to support your efficiency but keep your human genius at the absolute centre of your brand.
Check out yesterday's blog for how to get it right!
