821
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 25 Oct 2024
821 points (98.8% liked)
Technology
58875 readers
3580 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
Don't overestimate LLMs, it can't code and never will be. It can create templates convincingly enough and do boilerplate parts that are nonsense only sometimes, but those aren't the fun parts of the coding process anyway. In my experience, LLM isn't helping at all and I spend more time fixing it's nonsense than I would do if I don't use it at all, so I don't
As someone without a computer science background and who started learning Python for data science shortly before LLMs became mainstream, I gotta say it's been pretty useful for the learning process. I don't mean I just use it to write scripts for me but rather it can be a useful sorta of guide the way a scripted advisor mihht be in a game. Seems to me that one of the good sides of LLMs is that they can make technically dofficult fields more accessible as long as you understand its limits and know what it can and cant do._ i would never use it for any sort of subjective issue but I find it great for logical tasks. And this is not to say that's its perfect for that either but it has increased my efficiency for certain work tasks tremendously.
As someone with degrees and decades of experience, I urge you not use it for that. It's a cleverly disguised randomness machine, it will give you incorrect information that will be indistinguishable from truth because "truth" is never the criteria that it can use, but be convincing is. It will seed those untruths into you and unlearning bad practices that you picked up at the beginning might take years and cost you a career. And since you're just starting, you have no idea how to pick up bullshit from truth as long as the final result seem to work, and that's the works way to hide the bullshit from you.
The field is already very accessible for everyone who wants to learn it, the amount of guides, examples, teaching courses, very useful youtube videos with thick Indian accent is already enormous, and most of them are at least trying to self-correct, while LLM actively doesn't, in fact it's trying to do the opposite.
Best case scenario you're learning inefficiently, worst case scenario you aren't learning at all