Old school ChatGPT

Old school ChatGPT

By Steven Orpwood

June 2023

 

When I was at school a long time ago, my headmaster, now sadly departed, taught us sociology. Being a student focussed more on mathematics, science and sleeping, I found writing long essays about things that didn’t really interest me to be mindrendingly awful. Hence, like all teenagers before and after me, I looked for a way round the problem. And a solution was found, it was ChatGPT 80’s style. This AI was no 4k GUI with life like colours, the ability to multi-task, and create works of art, no, this AI was paper based, sent to me once a month in an envelope using a stamp. This was the age of Readers Digest. No top shelf fare for me, my smorgasbord of delight was contained in a handily sized magazine.

 

But how, I hear you ask, do you have the audacity to compare ChatGPT, a state-of-the-art AI, with a magazine of sometimes dubious content that offers mountain dwellers adverts for motorboats and teases 60 somethings with a ‘Stannah’ stair free life. Simples. My headmaster would set us an essay, for example, ‘how did ancient people organise their societies’. I would then attack a pile of RDs, and voilà, an article on Inca civilisation in Peru, or ‘the impact of violence on society’ was swiftly despatched using an article about the effects of gun crime in New York. And even if there were no exact matches, a short introductory paragraph would make an article about the décor in Liberace’s Sherman Oaks home, fit with themes ranging from famine in Ethiopia, to Neil Armstrong walking on the moon.

 

No hold on, I hear you cry, you must be joking, this isn’t what ChatGPT does, it carefully selects material from many sources and weaves them together into a seamless and focused article. I suppose I could agree, but no, not a chance. The writers at RD carefully selected their material using many reliable sources and sometimes their own expertise, crafted it into an absorbing and informative whole and passed it to the wider public via a comprehensive QA mechanism; at least I’ve always assumed someone else read them prior to printing.

 

Ok, but you still had to do some work yourself, it wasn’t the finished article. Actually no, in most cases they were copied word for word. Based on the vocabulary used I have always believed the writers of RD were children who didn’t like inclement weather and so wouldn’t want paper rounds. As a result no amendments were required. As for the introductions - weaving together titles and articles with no apparent synergy - 50 words of teenage gobbledegook later and my headmaster, an otherwise intelligent chap, would happily accept that teenage minds are not always on the same plane as anyone else’s. That or he assumed I was either touched or gifted.

 

Plagiarism then. Oh, come on, who else read them? We only had them because my dad won a competition and we got a year free, after which I made it clear that my education would suffer if the subscription were stopped.

 

And there you have it, ChatGPT is not new, it’s just changed its form and doesn’t take a month to arrive, or take up as much physical space. In summary, don’t be worried about ChatGPT, or AI in general. It’s always been there, you just didn’t realise it!

 

At Aim, we too make complex things simple and connect the dots where others might fail, and we’re not averse to a bit of ChatGPT.

 

For more blogs on ChatGPT and AI, please click here.