Gamedev using D here, my main gripe with programming isn't AI, but that I decided to go with my own engine, and now there's barely any documentation on graphics API use that isn't the OpenGL manual, as I'm encountering very interesting bugs recently (textures disappearing, texture glitches, etc.), while most article nowadays being about how I wasn't supposed to learn any programming, but instead art, so I could create the next Undertale, Touhou, etc. all alone, and it would be truly my game with my true intentions.
Some people are often asking "but what about these other tools", when I criticize AI. IMHO, some of the tools that made game development more accessible also made it more atomized, since you have less need for a programmer nowadays, with some tools not even requiring "classical" coding, but is visual in nature instead.
is everyone here a complete beginner? how do so many people relate to this? as soon as you need to do anything halfway interesting the thing just confidently spews nonsense.
There's always been a tendency of folks reading programmer humor to be beginners rather than seasoned devs. I think, there's just more of those in general, as there's lots of fields where entry-level coding skills are good enough...
Including when you have entry level skills and just enough seniority to pretend it's a full time job to tell people who know what they're doing to keep doing it.
Nobody's really serious on a meme sub. Maybe check your humour meter. It must be broken.
And halfway interesting means beyond a medium blog written by someone with 3 months experience trying to pad their resume, and not following standards beyond default linting rules at all
At my workplace eng management won't shut up about AI and vibe coding.
It's because the potential customers for your product are asking about AI adoption. It's become a box to tick for all of the non technical goofballs making decisions.
Nope. 10 years experience at dicking around and collecting a paycheck. ChatGPT helps me dick around better.
Sometimes the more nonsensical it looks the less likely someone will bother taking their time to actually code review it.
My goal is keeping a job. Not getting work done. Then I just get more work to do.
Edit: I don't know why I expected people on this sub to actually be able to tell what a joke was.
I just some a lot of Lemmy are neurodivergent and do in fact require /s.
Which is actually halfway true, the other half being a more international crowd that aren't on the same page as whatever perspective you've got. That said, I got the joke and second you on spending more time with my feet on my desk while the boss pays me for my 20 minutes of work.
When I used to do image and signal processing for embedded systems in C and C++, AI was useless. Now that I do backend web development in Python and Ruby, AI is better than me. It really depends on the problem area and how many sample code and answers are out there for it to steal from.
I do backend development in PHP and Ruby, and AI sometimes has a suggestion that helps me out but is often completely, utterly useless, especially at actually coding the thing from scratch.
Like 95% of the people on here do the latter, though. (Way too often I get in arguments where it seems like people don't realise there's other kinds of coding)
I do a ton of Powershell scripting, and AI is either a half competent programmer, or someone let grandpa respond with the syntax from nineteen dickety two
Can confirm. AI is worse than useless for embedded systems.
I was working with Yocto on a very specialized xilinx chip. I had chatGPT makeup a chapter that didn't exist (in a manual that did), reference non-existent paragraphs from said chapter, and then argue with me quite confidently that that chapter was real and the information it was giving me was accurate.
ChatGPT can make ridiculous claims about what it can do. It sometimes even says that it does real world things when it’s laughably impossible.
And as soon as you enter corporate stuff, LLMs are useless again, because most things are integrated into existing ecosystems which LLMs don't know and/or libraries are only used for closed source code.
Really? I had an app that would autogenarate time sheets for work in Google Sheets. I decided to minimise API calls by doing a single call to Google Drive then parse the HTML and reupload. Not a big Python project but ChatGPT hit a wall pretty fast on that one. Though, tbf the documentation was suprisingly opaque so I suppose that goes back to your point.
That project also produced my finest pile of spaghetti code since I had to account for stretched cells in the HTML parsing. I still have a piece of paper with my innumerate math scribbles. The paper makes sense to me. The code does not.
When I needed to pick up on some basics, it actually did help but ultimately not as much as actual guides and tutorials written online. This image of a chimera certainly matches the kind of Frankenstein code I was getting.
That said, when I was having some very interesting ideas about making automated code in R, it did make for a good sounding board. You don't need to Google when everything in R has documentation but you do when you're combining libraries in unique ways to automate 98% of the stupid shit you do at your data researcher job (e.g. can you look up in our database how many students pick their nose during philosophy class on a Friday?)
ChatGPT is a better programmer than you
Unfortunately very true 😭
It's great for getting off the ground in a language you're unfamiliar with. But a recipe for technical debt when used long term. You're inviting spaghetti code with some real hair pulling bugs.
This was my usecase when learning JS. I've always had a very very difficult time with programming and wrapping my head around those types of patterns. But with the help of AI I was able to get a quicker understanding and being able to ask followup questions, explain what different symbols do, nowadays I use it as a basic reference to get a starting point to some solutions while just being well versed enough to approach a lot of things on my own.
I often try to switch around and experiment with different similar and potentially better approaches unless it's giving me one of those responses that have been the most bogstandard straightforward way of doing it... because of a stackoverflow answer from 12 years ago. But those are often the more simpler queries. But each day I practice my programming I turn to search engines more and more to adapt similar problems to my own so the kickoff of using AI got me to the point where I had a stronger understanding of how things work in a practical learning method that worked really really well for me.
I hate how great AI can be in some use cases while I also am part of the reason for excessive power usage of GPUs :( but I really don't think I would've been able to get as far as I've come today without it.
This is genius
