I've only recently begun using copilot auto-complete in Visual Studio using Claude (doing C# development/maintenance of three SaaS products). I've been a coder since 1999.
The suggestions are correct about 40% of the time, so I'm actually surprised when they're right, rather than becoming reliant on them. It saves me maybe 10 minutes a day.
The only part AI auto complete I found I really like is when I have a function call that takes like a dozen arguments, and the auto complete can just shove it all together for me. Such a nice little improvement.
I have been begging Claude not to write comments at all since day 1 (it's in the docs, Claude.md, i say the words every session, etc) and it just insists anyway. Then it started deleting comments i wrote!
Yeah, it usually gets the required args right based on various pieces of context. It have a big variation though between extension. If the extension can't pull context from the entire project (or at least parts of it) it becomes almost useless.
IntelliJ platform (JetBrains IDEs) has this functionality out of the box without "AI" using regular code intelligence. If all your parameters are strings it may not work well I guess but if you're using types it works quite well IME.
Can't use JetBrains products at work. I also unfortunately do most of my coding at work in Python, which I think can confound things since not everything is typed
... you can't use JetBrains? What logic created a scenario where you can't use arguably the best range of cross platform IDEs, but you can somehow use spicy autocomplete to imitate some of their functionality, poorly?
It's mostly been due to the code quality: everything LLMs have spat I've had to heavily edit to make it readable, good quality code, and that's when the code actually works - the number of iterations it takes me to get code that actually works normally means I could have coded it faster myself without trying to get an LLM to do it.
I am baffled about how each company are jumping into LLMs without considering anything about their own privacy when 10 years ago, just using GitHub with a private repository could have been an issue.
> To be fair the code they produce is dogshit, so it isn't a problem.
That's not a problem for managers and CTO that are just being brainwashed by marketing and LinkedIn posts that all their engineers should use Cursor.
To get the code you want out of LLM/Diffusion AI, you have to literally treat it as a smart text editor that understands English and the syntax of the language you are working it. The expenditure of keystrokes is easily worse than if you just wrote that code yourself in the first place.
I'm the developer of a boxing game called Leather, that I've had in the Play Store and App Store for about 6 or 7 years now.
As I'm working towards a Steam release I've been digesting a lot of this guy's advice - https://howtomarketagame.com/
Whilst much of his guidance is of course marketing rather than design related, he does write about genres and game mechanics that attract players - specifically on desktop rather than mobile. It's worth a few hours of your time to check his stuff out.
The suggestions are correct about 40% of the time, so I'm actually surprised when they're right, rather than becoming reliant on them. It saves me maybe 10 minutes a day.
reply