Like many, I’ve been flabbergasted by the huge advances we’ve seen in recent years with artificial intelligence. I’ve spent hours just playing with ChatGPT and Stable Diffusion and am consistently impressed.
I’m also aware of issues surrounding this breed of technology, like with intellectual property and black box biases. On top of that, I’m trying to avoid hype and grift. Where are you seeing AI excel, or where does it have a potential to excel? Are there places where it is being shoehorned into that should be avoided?
It excels most as a tool for summarizing, formating or calculating things that other people somewhere might’ve needed at one time, in areas where you yourself have expertise enough to quickly check its work.
Pretend it’s an army of undergrads at your command. It’s probably good stuff, but check their work. Some of its cheated, some is wrong, but as a whole most of it is probably fine for mundane uses.
For some reason (maybe due to federation problems caused by Rexxit) this post only appeared now here on programming.dev.
So in my opinion it excels at extracting meaning from fuzzy human input, and transforming it into another representation like text, code or images, with surprising accuracy. I’m consistently impressed by GPT-4’s coding ability, and also its ability to understand and explain nuance and implied meaning in text, which is very useful to me as a non-native English speaker.
I’ve been working on a blog post about the use cases I’ve found for ChatGPT so far, but as so many other things I do, it’s 95% done and it’s just collecting dust in a folder on my computer. Maybe I’ll publish it soon and link to it here.
I’d love to see a blog post when you’re ready with it. Thank you for your comments!
