So, here’s a little throwback. Picture me, a thirty-something wandering back into uni life, already feeling a bit like the “mature age” student in the room. Luckily, growing up with some decent computer skills, I could at least wrangle my way around the internet. But then I had this mate who wasn’t a computer whiz at all, and he was the one who nudged me into trying out these early ChatGPT models, 3 and 3.5, just to see what all the fuss was about.
And honestly, I was kind of intrigued. Not because it was perfect, it definitely wasn’t, but because it felt like a novelty. I quickly realised it didn’t quite read my mind or get my context right, and yeah, maybe I was hoping for a bit of mind-reading magic. But the real comedy kicked in when I tried to use it for high school maths and physics. I’d dropped maths early on, never really did the big exams, and thought, hey, maybe an AI could help me brush up. Spoiler alert: it was always just a little bit off. It was like having a friend who’s almost good at guessing but never quite hitting the mark.
In the end, that was actually the charm. It gave me a safe space to ask those “dumb” questions about jargon or symbols I was embarrassed to ask a human. The maths never got perfect, but it was a fun reminder of how these models were like a friendly, slightly scatterbrained tutor back then. And now? Well, now I just get it to write little programs for me instead. Live and learn!