Re:Ask HN: Is anyone else bearish on OpenAI?
by Rastonbury on Sunday November 12, 2023 @12:27AM
(#38237583)
As a non-professional coder, I'm honestly surprised there seems quite a few professional coders in this thread who don't do the first Stackoverflow first scenario. As someone learning code, shoehorning SO code when I don't understand 20% of the replies made the learning curve so steep, chatGPT has made this so much easier by changing it and telling me why will/won't work, even when watching out for or correcting hallucinations, overall its a time save
Re:Ask HN: Is anyone else bearish on OpenAI?
by meiraleal on Sunday November 12, 2023 @04:27AM
(#38238722)
> I'm honestly surprised there seems quite a few professional coders in this thread who don't do the first Stackoverflow first scenario
You just found your path to become a better developer than the majority then
Re:Ask HN: Is anyone else bearish on OpenAI?
by smrq on Sunday November 12, 2023 @07:36AM
(#38239595)
As I've matured in my career I've gone to SO less and less, to the point where I mostly avoid it now. I use tools, languages, and libraries with which I am now highly familiar--perhaps more so than the average SO answerer, although I guess it would be arrogant to say for sure--and so the bulk of my questions are edge cases in my knowledge which are more easily answered by reading source code or specs.
I suspect this is one of the differences between people that find GPT helpful and people that don't--it seems much more potentially helpful when you are in a less familiar environment. In my work, when I run into unexpected behavior, if I knew what question to ask then I would probably already know the answer too. Until the day I can feed in an entire codebase and ask "What's wrong with this? vaguely gestures at everything", I don't expect that to change significantly.
Re:Ask HN: Is anyone else bearish on OpenAI?
by Rastonbury on Sunday November 12, 2023 @08:21AM
(#38239823)
That makes sense, and its funny because the other reply I got a joke was about how bad coders don't even bother going to SO
Re:Ask HN: Is anyone else bearish on OpenAI?
by aprdm on Sunday November 12, 2023 @04:36PM
(#38244309)
Probably because you're deep specialized in something or have stagnated on a company for a long time ? I cannot imagine knowing everything about what I am working on, that has changed dramatically every 3-4 months even within the same company
Re:Ask HN: Is anyone else bearish on OpenAI?
by sangnoir on Sunday November 12, 2023 @08:31PM
(#38245895)
When you've been working for a while - not necessarily even for the same company- you gather an internal library of knowledge that you can pattern match against for solutions. You may not have seen the exact same thing, but you may have seen something like it. There are hardly ever an entirely novel class of bugs anymore
Re:Ask HN: Is anyone else bearish on OpenAI?
by aprdm on Sunday November 12, 2023 @08:41PM
(#38245951)
Yeah definitely. But the tooling is very diverse in big companies. They might use every database under the sun, one day you might be doing React another one AngularJS, one day Django another day FastAPI, one day k8s, other day terraform.. at least my brain is very limited and it can be helpful
Re:Ask HN: Is anyone else bearish on OpenAI?
by horns4lyfe on Sunday November 12, 2023 @10:22AM
(#38240877)
You’re surprised most professionals would want to actually try and diagnose a problem instead of trying to copy and paste a shoehorned solution? What? Most professionals I work with actually try to debug code, look at docs, etc. You know, research, like a real job.
Seconded.
Let me tell you what I used to do.
First, Imagine I have an error executing code or an error running some bit of 3rd party software. I go to Stack Overflow and search. I find posts related to my problem and I spend a great deal of time trying to shoehorn existing answers to my specific issue. Sometimes, my shoehorning works and I fix my problem. Other times, it doesn't work, and then I post on Stack Overflow myself. And I wait ... and wait ... for a response. Sometimes I get a response.
Now, when I have this type of problem, I tell ChatGPT, "Hey, I'm trying to do <xyz> and I'm getting this error <abc>. Help me troubleshoot this. And it almost always helps me fix my problem. And it's ~10× faster than Stack Overflow.
=============
Second, there are times where I have write code to do some relatively 'complex' data manipulation--nothing sophisticated, mind you, but stuff like, "I need these data columns rearranged based on complicated logic. And I need the text in columns A, X, AQ, and F merged, but only if <blah blah is true>. Otherwise, Just merge text in Columns A and AQ, except if the date in column ZZ is after January 1, 2019.". I can do this stuff on my own, but: a] it's cognitively draining, b] it takes time, c] I often make silly errors due to the complexity.
ChatGPT is, again, an order of magnitude faster than I am. And it makes _fewer_ errors.
It still makes errors. And I still have to know what to look for to catch those errors, but it decreases my cognitive load tremendously.
edit: I haven't used Stack Overflow in 6 months. And "Googling" is Plan B.
=============
Edit 2: I recently had to write a sympathy letter to someone whose husband died.
I knew the general ideas behind what I wanted to say, but I knew I wasn't going to write anything great.
I fired up ChatGPT and said,
"write a sympathy letter to <x>. Tell her that I didn't know her husband well, but the few times I met him, I could tell that he cared deeply about you<x> and his daughter. I know his daughter well and I think she gets a lot of her great qualities from him and you. Tell her I don't know what to say in times like this. Keep it short-ish. Avoid schmaltz and sentimentality because <x> isn't that kind of person."
It gave me about as perfect a letter as I could have asked for.