💾 Archived View for chesteryang.flounder.online › artificialintel.gmi captured on 2023-12-28 at 14:59:42. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-11-14)
-=-=-=-=-=-=-
I had a brilliant idea for a series of stories I was going to write on Flounder, and if it was successful, other places. I sketched out my story and wrote a few chapters. However, when I plugged my work into ChatGPT4 and asked it to "clean this up", what it spit back at me was astonishing. Humans just can't compete.
Me: The wind blew as he opened the old oak door. It was stained with time. He looked down the hallway and saw the most astonishing thing. It was a monster with horns, growling, salvating, and ready to pounce!
ChatGPT4s version: The wind howled as he pushed open the ancient oak door, its surface stained by the passage of time. Peering down the hallway, he encountered an astonishing sight: a monstrous creature, adorned with menacing horns, growled softly, saliva dripping from its jaws, poised to pounce.
ChatGPT4s version is clearly better. One could make the philosophical argument that we've been using technology, whether it is oil-based paints, GIMP, or even a human editor, for a long time, and that using ChatGPT4 is no different. Somehow, it is though. It's one thing if I thought of the original idea, but chat GPT4 polishes it up so good, it ceases to be my creation. At what point does that ownership end? Does it even matter?
In adddition, it can just spam top-tier stories all day long. Humans just can't keep up and write as well. Well, I suppose it's possible, but the work will be so tedious and slow that it's just not worth it unless one has a true passion. And if that does happen, it'll just appeal to niche audiences, diehards, and hobbyists.
On a similar note, I've been reading some recent research and can notice when the researcher used ChatGPT4 to smooth out their writing. Some author's even forget to delete the introductory and obligatory "Here is the answer your looking for, sir!" opening bit. One can also notice it when they compare the flow of reserach articles to articles that were published as little as a year or two ago. It's too smooth and comes off unnatural to the native ear. I notice that a majority of the authors with foreign names tend to use it more. I know it's anecdotal, but it makes sense; I'd use a crutch or a boost too if English wasn't my native language. The language change can be very jarring. One minute their research is written in X style, and then it just takes a switch. If you've been reading research for as long as I have, you can pick up on the subtle differences too. I'm happy I completed my research in the days before "artificial intelligence" because I know that I just couldn't resist the temptation to use it to "clean up my writing" a bit. It's an awesome tool, and its ethics are questionable, but either way, it just nuked writing and research into the stone age.