Once upon a time I was a teenager without much money to spend on games, nor the computer to handle new ones, so I spent a lot of time playing demos of older games. One which was particularly fun was a RPG placed in Roman times Britain called Nethergate. The demo ended with my party burning alive in a wickerman, due to a curse from some witches. It's been a while, so details are fuzzy.
I wanted to play more of the game, to explore an area beyond the witch hut, but to go past I'd have to deal with them. The obvious solution is to kill them (actually this is kinda dumb, since it's unlikely that the game mechanics were set up that killing the witches so early would just remove the curse). My party was far too weak to defeat them. So I did what anyone would do: I leveled up.
One problem: the demo wasn't designed for me to level up far enough to beat them. The game didn't have unlimited enemies, or it wasn't supposed to, but I wasn't going to let that stand in my way. Instead I found a triggered Roman patrol, which I could kill for XP. Except after defeating it, I would hit a wall, a message about "it's too dangerous to go any further".
Here I combine brilliance with stupidity. My initial logic, defeat witches to play the whole game, was flawed. But my solution to the XP barrier, well that was my first exploitation, or should I say "creative use of game mechanics".
Rather than fully defeating the patrol, I'd instead kill all but one and then flee from the battle. I got a depressing message, but I also got the XP from the kills. Retreating itself was somewhat slow, having to be done manually for 4 characters, and the battle itself only netted 4 kills, so it would take a lot of fight-retreat cycles to level up much. I did it anyway and got my characters to a much higher level than was intended. And so was my first grind, a process for which I did not yet have a word, and my first exploit, which also seemed to be nothing more than innovation.
I never did kill the witches. Eventually I got bored of the process, but I definitely spent many hours on it. I was encouraged by my ability to put a noticeable dent in their health, meaning 10% or so. And maybe that was only one of them. But it indicated mortality! Still, at some point I realized that the game wasn't written such that killing them would remove the curse, that the demo ends how it ends.
While I was trying to find this game, since I'd forgotten the name, I found that the company remade it, called Nethergate: Ressurection. I've not played that yet, but if by chance you're looking for an oldish RPG (original is 1999, so not old), I imagine it's a good bet. There's still a demo, which I've downloaded but not played yet.
Incidentally this was also where I first learned about class scaling. Somehow my melee characters were gaining power faster than my casters, though I think they were all gaining the same XP, but eventually they were able to spend point on cross-class skills, learning some casting, which was disproportionately powerful for the point cost, due to the other stats they'd gained from leveling.