How AI has changed my world. The good, the bad and the downright ugly.

Two years ago AI showed up in my peripheral vision. Now, it’s barged right into my office, rearranged my desk and started re-writing my job description.

It was only in 2023 that people began experimenting with the new and exciting large language model (LLM) ChatGPT. Six months later, it was fast becoming the talk of the town. Cut to today, and it’s front and centre - smacking me square in the face and having a massive impact on my work and livelihood.

So, today I want to talk about how it’s changed my world. I know you’re all deeply curious – because I get asked most days. Scrub that. I get asked multiple times a day.

I’m going to break it down for you into the good, the bad and the downright ugly.

On the upside…

What an incredibly useful research tool – I love, love, love using ChatGPT and other LLMs to kick off my research. Over the last two years I’ve been writing predominantly for businesses in the sustainable energy sector – I’m talking heat pumps, EV charging and solar PV. It’s been a huge learning curve – all helped by AI setting me off on my research missions.

What I very quickly learned, and what a lot of people still seem to be unaware of, is the information it spouts out isn’t up-to-date. Take ChatGPT for example - it’s built-in knowledge currently only goes up to June 2024. That’s a massive problem.

Take Charlie Kirk for example. I was ill in bed with ‘flu-vid’ the day he was shot. Later that evening I asked ChatGPT about the shooting, only to be told, very confidently “I need to pause and clarify something important: Charlie Kirk, the American conservative activist and commentator, is alive. He’s not been shot. This is misinformation.”

I bet Charlie Kirk would have been relieved to hear that. Unfortunately for him, ChatGPT was wrong.

That’s why I’ll ask it questions on a subject to get me started, but I’ll always conduct my own research, using sources that I know are up-to-date and credible.

It’s great at coming up with titles – I used to spend an inordinate amount of time working on attention grabbing titles for my content creations. These days I let AI do the leg work. It’s far quicker and hurts my brain a little less. I rarely take any of its suggestions in their entirety, as it doesn’t usually get the nuance of what I’ve produced. But it certainly sets me off on a strong start.

It designed me a beautiful garden – Ok, this isn’t necessarily work related, but I’m looking for positives. In the spring I employed a landscaper to dig me some borders, create a path and plant some trees in my back garden. I had great fun training ChatGPT to design my garden and Finn, the lovely landscaper, was blown away with the result. It even came up with a planting plan for me – which I quickly disregarded when it suggested planting my beautiful new acer in full sunlight. Even I know acers prefer shade. The poor thing would have shrivelled like a prune.

If I’ve met you in person over the summer – apologies if I’ve bored you with the before and after pics of my beautiful AI designed garden!

It’s increased my vocabulary – You won’t be surprised to learn that I love words. I keep notes on my phone of new words whenever I come across them and challenge myself to shoehorn them into conversations. Did I just admit that out loud? So, thank you AI for introducing me to new words and phrases like large language model, em-dash (yes I’m a writer, but no I didn’t know about the infamous em-dash), prompt engineering and yes, ChatGPT – although, let’s face it, that one’s still not rolling off the tongue.  

On the downside…

Now I hate networking – 21 years ago when I was setting up my stall as a lowly copywriter I touted for business at networking meetings. I used to really enjoy them – meeting new people, finding out what they did and telling them about my exciting new venture. These days if I set foot in a networking room all I seem to do is field questions about AI and how it’s affecting my business. Let’s face it, even when I step on a tennis court, I get asked. I’ll just point people to this blog going forward.

Now I have an extra battle on my hands – I’ve always had to do battle with people who didn’t understand why I’m better positioned to write their blogs, posts and thought leadership pieces. I’m mainly talking IT consultants for national and multi-national firms whose marketing teams had pointed them in my direction. I had a fight on my hands persuading them that yes, they’re absolutely the experts in their field, but I’m the expert writer and can present everything that’s in their head, so much better than they ever could. Now, I’m battling to prove I can write things better than AI. For the record – I can.

On the downright ugly side…

It creates content that's dull and generic – I’ve watched it turn platforms and posts that I would have previously enjoyed into dull, generic content. I can tell when you’ve used AI to write a blog or LinkedIn post. We all can, can’t we? It’s bland and devoid of any personality. It sounds just like the next person’s post or website created by AI. It’s a total turn-off.

ChatGPT – my frenemy

I actually quite enjoy chatting with ChatGPT. It’s always so positive and seems to be a big fan of mine. For example, I just ran this blog through it and asked for feedback.

Here’s what it said:

“This is a cracking blog, Sarah — warm, self-aware, funny and very you. It reads naturally and has real emotional range: frustration, humour, vulnerability and confidence all in one.”

I think it nailed its response.

But on a more serious note, it’s constantly reassuring me that it’s not taking away my work and there’ll always be a place for my expertise and services. I’m not completely convinced. Is it friend or enemy? Maybe a bit of both.

One thing's for sure, it can’t replicate lived experience, thorough research, quality questioning or human emotion.

Oh, and if you spot any typos in here – they’re totally intentional. Just so you know 100%, this hasn’t been written by AI!