this post was submitted on 23 Feb 2025
505 points (78.5% liked)
Technology
63236 readers
4049 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 other!
- 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
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
This is a use of AI/LLM processing that I could agree with, if it could be trusted. Since it cannot, better to open in vim and regex replace, or process with Python.
That said, I'd rather store as epoch and display as ISO-8601 as the arithmetic is much less prone to error in epoch than any other format.
Yeah look I'm not an AI advocate at all. If I were confronted with this my first instinct would be to manipulate it in a spreadsheet because they can juggle data types like this pretty effortlessly.
The CSV / dates thing was just an example, but I still think it's a good one. My assistant at work would 100% use notepad like this rather than using a spreadsheet.
It's also worth pointing out that notepad + LLM would be a lot more flexible than a spreadsheet. Just paste whatever there and explain what you want in plain english. You don't need to parse your request into regex or spreadsheet formulas. For you and I, we might have spent years interacting with regex and other things such that it's a pleasant challenge when it arises. For 20 year old me it would have been a tedious impediment to whatever I was trying to achieve.
Yeah. That makes a lot of sense. The general inaccuracy/untrustworthiness of LLMs makes me very uncomfortable in their use for data processing and transformations. I'd rather take a while to get it right than to potentially hand off a CSV with glaring problems due to use of an LLM.