that’s how it usually works. unless you use an AI, then it’s right after one or two reprompts. but you don’t get any of the satisfaction and you feel like a cheater.
I think we’re using AI very differently. I only use it for stuff I really can’t figure out myself and then it needs 15 runs before it closes the loop and recommends the first thing that didn’t work. I don’t use it for everyday tasks, because I see how my performance and capabilities drop. AI is mostly frustrating for me.
you’re right, we DO use it very differently. i use it to code things i don’t know how to code. but my ideas aren’t that original, so it always produces a working product as long as a reprompt it enough. your work involves something that hasn’t really been thought of yet, so i can see why AI wouldn’t do anything
Or something more complicated than it can handle. Ask it to implement an ASN.1 floating point parser. It looks good but the closer you look at the code the more trash it is.
whenever people say “try asking it to do THIS!” they’re usually talking about something the AI doesn’t get right after one prompt. just prompt it long enough and it’ll work
I tried for a good 20-30 minutes to get it to generate something that worked. It just gave me a sequence of garbage. Each one looked plausible but as soon as I really read the code it was clear there were issues. Different issues each time. I wasted more time trying to get it to produce correct code than it took to write it myself, once I gave up on the AI.
For me the most productive way to use AI is to ask it to suggest a strategy and to implement that myself. If the code is simple enough for the AI to get it right, I can often write it myself with less effort. If it’s something that requires multiple reprompts to get right, I can almost always do it faster myself.
if it’s something that requires multiple prompts to get right
that’s almost always the case. it’s so dumb how all these AI critics are like “well it didn’t do everything right the first time, therefore LLMs are garbage.” ffs just have some patience, it only takes a few more prompts most of the time
programming either makes you feel like a genius (when shit finally works) or a complete moron (when it doesn’t). it’s usually the latter :(
10 hours of feeling like an idiot for 1 minute of feeling like a genius.
Sometimes, after that 1 minute, you realize that you were still an idiot, you just didn’t understand it for a moment
On the bright side, after 10 years of doing it, you might improve the ratio to 1 hour of feeling like an idiot and 1 minute of feeling like a genius.
Wait… Isn’t that bad… Going from 10:1 to 60:1???
Check again. Going from 600:1 to 60:1
He’s a programmer, not a mathematician.
You could have let him have his minute.
Some people call that an abusive relationship. I call it M-F.
that’s how it usually works. unless you use an AI, then it’s right after one or two reprompts. but you don’t get any of the satisfaction and you feel like a cheater.
I think we’re using AI very differently. I only use it for stuff I really can’t figure out myself and then it needs 15 runs before it closes the loop and recommends the first thing that didn’t work. I don’t use it for everyday tasks, because I see how my performance and capabilities drop. AI is mostly frustrating for me.
you’re right, we DO use it very differently. i use it to code things i don’t know how to code. but my ideas aren’t that original, so it always produces a working product as long as a reprompt it enough. your work involves something that hasn’t really been thought of yet, so i can see why AI wouldn’t do anything
Or something more complicated than it can handle. Ask it to implement an ASN.1 floating point parser. It looks good but the closer you look at the code the more trash it is.
whenever people say “try asking it to do THIS!” they’re usually talking about something the AI doesn’t get right after one prompt. just prompt it long enough and it’ll work
I tried for a good 20-30 minutes to get it to generate something that worked. It just gave me a sequence of garbage. Each one looked plausible but as soon as I really read the code it was clear there were issues. Different issues each time. I wasted more time trying to get it to produce correct code than it took to write it myself, once I gave up on the AI.
For me the most productive way to use AI is to ask it to suggest a strategy and to implement that myself. If the code is simple enough for the AI to get it right, I can often write it myself with less effort. If it’s something that requires multiple reprompts to get right, I can almost always do it faster myself.
that’s almost always the case. it’s so dumb how all these AI critics are like “well it didn’t do everything right the first time, therefore LLMs are garbage.” ffs just have some patience, it only takes a few more prompts most of the time