It’s human to have fears.
There are the basic ones; spiders, heights, airplane travel. Fears can be specific to the environments we move through every day – particularly in our professional lives. A salesperson stresses about forgetting to return that important call, a hairdresser panics about whether they’ve left that bleach on too long.
As a content writer, there aren’t many vocational fears that keep me awake at night. The only reason I’ve sat bolt upright in bed at 2:00am is if I’ve suddenly remembered a synonym that had been on the tip of my tongue all day. Though it’s hard work, I’ve been pleased to enjoy a career where people respect my experience, expertise, and skill, and know the value of having quality content to promote and sell their brand and products.
But now, thanks to AI, there is something to fear in my rosy content world. And it could very well see me out of a job soon.
Unless you’ve been firmly offline, living under a rock, or you don’t have a boss who spends his spare time deep diving into even the quietest whisper of new technology, you should have heard of AI content creation. Whilst this isn’t a new concept – and certainly not one I’ve been afraid of thus far – there’s a lot of buzz right now about one chatbot app in particular.
So, let me – an admittedly bitter, but altogether human writer – tell you about ChatGPT, and question whether it could ever replace humans in the content world.
Aside from my nemesis? ChatGPT is an AI chatbot developed by OpenAI that uses online text to come up with natural language answers to questions you ask it. It is designed to simulate human conversation and provides answers to questions in a conversational style.
OpenAI counts Elon Musk as one of its early founders, however the man responsible for such triumphs as the implosion of Twitter and space littering has since distanced himself from the company, allegedly due to their transition from not-for-profit to, well, profit. He also tweeted about ChatGPT, ‘We are not far from dangerously strong AI’ which we’re all apparently supposed to pretend is a normal and not-concerning thing for a (self-proclaimed) tech genius to say.
To use ChatGPT, you have to create an account with OpenAI. From the very first step, I’m scared – why is it asking me to verify that I’m human? Why is this chatbot worried about other bots? Are they not friends? Brothers-in-arms in the war on humans?
Regardless, I sign up.
The design is simple. A toolbar, query bar, and a blank document ready to be filled with generated responses. Once you type in a question or request, ChatGPT types out the answers.
To begin, I asked ChatGPT to tell me about itself. It gave me a brief overview using some quite technical language, so I instructed it to rewrite its answer more informally. I’ve actually used this answer in one of the earlier paragraphs – could you tell it wasn’t crafted by human hands?
Next, I wanted to see if we had anything in common.
Okay, so not much to bond over there.
As I started writing this blog around lunch time, I asked it for a recipe using some of the items I have in my fridge. Even though I found ‘Grilled Burrata and Basil Skewers’ a bit of a cop-out, I can’t deny that it did sound pretty tasty.
I then asked it to write this article for me. It did a fairly half-hearted job, some excerpts of which I’ve shared below. But I’ve seen ChatGPT formulate full essays, and I know that despite my reservations, being able to instantly access information from across the internet does have its perks. I mean, I can do that too… just a bit slower.
Tricky question. As a human, I want to say no. However, even I can admit that there are some scenarios where AI content is technically better.
Like in this study conducted by Today Online where it seems that, unfortunately, it is error and inconsistency that defines human writing over that created by bots. In this experiment, an English professor was asked to compare three pairs of writing samples covering the same topic and written in the same style (an essay, opinion piece, etc.) and determine which was written by a person and which was AI. Though the professor confessed that he struggled to make his final rulings, it was the errors sprinkled throughout the human piece that had him guessing correctly.
We’re not just talking about spelling or grammatical errors – the professor referred to things like overused metaphors, vocabulary shifts from formal to colloquial, and sometimes it was just a general vibe. When it comes to AI, the content was too … perfect, apparently. But it was also bland, unemotional, and lacking in nuance – which brings us to the key issue.
I could rail on for days about why I think human content is far superior to AI text. However, there’s nothing like getting your opponent to spell it out for you.
Because I am bitter and cruel, and will undoubtedly be Enemy #1 when the machines rise, I forced ChatGPT to explain why AI content will always be inferior to that produced by humans. It may have been pandering to me, but here are some excerpts from that generated article:
‘When it comes to creating content, human creativity can never be replaced by artificial intelligence (AI). Humans have the capacity for creativity, which AI cannot replicate. Content creation requires imagination and out-of-the-box thinking, two things that AI lacks. AI also lacks the capacity to understand emotion, which makes it difficult for it to create content that resonates with people.
Humans are able to think critically and make decisions based on sound judgment. AI lacks the ability to think critically, which makes it difficult for AI to create content that is meaningful and purposeful. AI lacks the ability to understand context, which makes it difficult for it to adjust its content based on a given situation. AI lacks the ability to understand the needs and desires of an audience, which makes it difficult for it to create content that is relevant and engaging.’
Yes, those were all sentences that were technically correct and made sense. But where was the spice? Where was the personality? Where’s the joy? At this point in time, those are needs that AI simply can’t fulfil.
According to ChatGPT, yes.
‘It can be used to automate customer service conversations, helping humans free up time to focus on more complex tasks. ChatGPT can also be used to generate product recommendations, based on customer query data, helping humans make more informed decisions.’
The key message here is that ChatGPT and humans can work together if the AI is used as a tool to assist in content creation, not as a replacement for the creator.
The debate about ChatGPT and AI content in education has been fierce, and understandably so. Already we’re seeing examples of bots writing essays and doing homework, which some educators decry as cheating. However, others are more open to collaboration and discovering how AI can improve comprehension and fact-finding in classrooms.
When given explicit instructions, ChatGPT can be used to kickstart scripts and write code. It can allegedly also sniff out bugs and create programming languages. You’ll still need to go through it with a fine-tooth comb to look for any errors and to personalise the script for your specific needs.
For many people living with a disability, the opportunities presented by ChatGPT are endless. It has the potential to provide users with greater independence with its Q&A style. It can perform vital tasks like writing a cover letter for a job application, or providing options for text or email responses in personal and professional settings.
To end this takedown and begin my reluctant truce with ChatGPT, I’ll share what the chatbot wrote when I forced it to concede defeat.
‘Overall, humans remain the superior content creators, as they have the capacity for creativity, empathy, critical thinking, context awareness, and audience understanding that AI lacks. While AI has its place in content creation, it cannot replace the creativity and skill of a human content creator.’
Now excuse me while I email this to my boss and cling onto my job for a bit longer.
This piece was written by Laura, who is Propelle’s Content Manager and would like to remain so for the foreseeable future. You can read more of her work here.
