As I typed in the heading for this article, my LibreOffice Writer informed me that ChatGPT is not a word, and I should reconsider my choice of using it. This made me chuckle somewhat wryly as I thought — is this a case of older generation (software) disowning its newer generation? I checked the suggestions, and it offered: chatting? Well, grandpa, ChatGPT might end up being a word soon, and quite a popular one at that.
Being a writer I have at least 10 years’ experience of tackling non-sense suggestions offered by different software. There was a time when I’d use Microsoft Office’ thesaurus feature proudly. Then my love for literature taught me better. I decided I knew better than a mere software, and would depend on it only to avoid typos. A little fling with South Asian literature later, I started using Roman Urdu words. At this point, I decided to ignore the red lines for good. However, a career in (content) writing and editing made my introduction to software such as Grammarly and Hemmingway inevitable. But even with the premium versions, I knew I can’t rely on them entirely — not even for editing! Maybe on a lazy day and with a good responsible writer, I’d let it take the reigns, and even then, it would always leave a few bumps in the way for me. So it was always with some skepticism that I’d read and reflect over all the “AI bots are out to take our jobs” articles.
Finding out about ChatGPT and dodging a bullet to my ego
I will admit I am not a tech-savvy nerd or anything, but I am not a recluse either. Besides my partner is definitely one cute geek, so I learn relevant news (willingly or unwillingly) even before it turns mainstream. Thus, I found out about ChatGPT, and it was not the first time my man was telling me excitedly about how impressive a new AI chat bot was. I have had my share of fun with Google Assistant and Pandorabots (aka Mitsuki) previously, and had read the conversations of the so-called “sentient” AI, LaMDA. I was not expecting any mind-blowing, revolutionizing technology.
So I checked out the latest sensation — ChatGPT, and to be honest, I found it quite impressive. I could see its merits as a time-saving software that can churn out somewhat well-structured articles and blogs. I mean it was much better than the likes of QuillBot (which I have seen writers use and create heaps of mess for the editors to clean). So unless it gets monetized or equally good AI-plagiarism detecting software are created, I can see many mediocre and a few good writers lose the race to AI, if not now, definitely in the future. I mean the few qualms such as it is not creative or funny enough (although as news has it, Anthropic’s Claude is already making great jokes), makes factual mistakes, and simply lacks uniqueness can be defeated eventually. It has come so far, who is to stop it from going the extra step too?
So do writers (I know it is affecting other fields vastly too but I am confining the scope of this article only to writing industry) have competition? Of course!
Do creators need to up their game? Definitely!
Does that mean we will stop writing? A hard no!
After all, AlphaZero had conquered Chess once and for all, but that didn’t eliminate the game from our culture. And so these advances indeed hurt human ego a little, not unlike the discovery of heliocentric model of the solar system. But we always find a way to console ourselves, nourishing our egos back to the center of the Universe with one reason or another.
I, too, despite my inclination towards pessimism and nihilism,worked my way to the realm of hope. When I was a kid, my father introduced me to a Pakistani poet, Nasir Kazmi. In his collection of poems, I found a gem, which I realized I was already familiar with, thanks to Nusrat Fateh Ali Khan, rendering it into a famous Qawwali. There is a line in the poem:
Doston ke darmyan, wajeh dosti hai tu
(you are the cause of friendship between friends)
It had happened at least two decades ago, but I have realized I value Art, Literature, and Knowledge above all else to this day. I am of the mind that it is these values that are eternal and so my ray of hope was: “sure like any search engine, ChatGPT can definitely come in handy, but it cannot do what humans can — create Art!”
Was I in for a surprise though?
Passing the Turing test with flying colors
Naturally my next step was to ask ChatGPT to write a poem. However, before I could do that my dear partner had already asked it to create several poems (one of them being a really cute one about our relationship) and even had the AI compose a song.
I must admit, that was hurtful!
I had just finished writing an essay on how salvaging our values and knowledge is the only reason why humans shouldn’t just go extinct, and here was an AI carrying forward those same values allegedly untarnished! Well I had to see this for myself and I did — giving prompts to ChatGPT which it happily turned into not-so-bad poems. I wouldn’t share those on social media at first but then I did, this once, and this happened:
Did this AI just got a distinction in Turing exam?
Is my Identity (let alone my job) being threatened here?
Of course, I am building this entire narrative on my previous assumption that the true purpose and value of humans lie in their capacity to keep the values going. If you value humans for just being humans or for some other reason, you might see my stance as a tad bit dramatic. Also, it is true that I have so far only talked about Literature but with Dall-E Mini, MuseNet, and other similar software, Art in other forms is also at stake.
With the basic assumption being “if what makes humans human is their capacity to think and create”, can ChatGPT and its successors truly replace “humans”?
To answer this question, I would like to address another ChatGPT-related elephant on the internet, the problem of originality.
The Problem of Originality
ChatGPT’s ability to create informative essays in coherent English has made Language and Literature professors all over the world quite unhappy. A similar kind of discontent is found among (content) writers, designers, programmers, and their employers. It is all about: how to assess the merit of a student or how to assess a candidate for a job?
In other words, is the portfolio someone shows even real or is it made using ChatGPT or similar software?
There was a time when I would have been annoyed with such technology on the loose. Not anymore! Because originality is a scam. If someone can use ChatGPT to get the job done and get it done well, what is the big deal? The ability to use it is in itself a skill. As for the students, well at least at present, I don’t think even ChatGPT can help a mediocre student get an A-grade. There are too many caveats in there.
Coming back to what I meant by calling originality is a scam is that nothing is truly original — there is only innovation. Everyone — scientist, artist, philosopher, etc, is always building on something that has already been there, which is exactly what AI does, albeit in a lot faster, accurate, and cleaner way.
So does that mean a time might come for humans to step-aside and give the torch to the AI? Is this too some step in evolution?
I mean sure, the current AI is only as original as we are, only as creative as we are, only as funny as we are, and maybe in the future would be as emotional, vulnerable, and moral as we render them to be. Currently, it may not be creating plays like Shakespeare’s, poems like Mirza Ghalib’s, art like Dali’s or theories like Relativity. However, it might learn to do so. That is a thought to ponder but I’d like to think it won’t come to that. So, what ChatGPT is doing now, is being a tool in the hands of humans, and is thus, creating poems for us. As such, it is not passing the Turing test at all.
Redefining the Turing test
To really pass the Turing test, an AI would have to create art for the sake of art (earning itself a touch of existential crisis won’t hurt either). I say this because when I revealed to my friend that I didn’t write the poem, she showed disappointment in a very human and thus non-AI way too.
The reason she valued the poem, not knowing it wasn’t actually written by me, was the fact that “I” had given the prompt. It was the human reflected in the product that made it valuable at all.
What does it mean for the writers?
We have to keep striving to keep our creativity game up. For many, that might mean writing less, which is fine. As Bukowski said in his famous poem, ‘So you want to be a writer’:
Unless it comes out of
your soul like a rocket,
unless being still would
drive you to madness or
suicide or murder,
don’t do it.
unless the sun inside you is
burning your gut,
don’t do it.
But what to do about all the content the (corporate) world is constantly asking you to deliver? Well as ChatGPT said in its rendition of Buowski’s famous poem:
Jokes aside, my request to every writer out there is: do not go about adding unnecessary complexity to your articles only to be distinct from the AI bots. Infuse creativity and uniqueness but never at the risk of compromising simplicity unless you are offering an original (or highly innovative) thought. We don’t want to produce work distinct from the ChatGPTs of the world only in the sense of being worse than that.
Icklings 