Comment Yes (Score 1) 247
- Parsing data: I love being able to upload a CSV of some random data to [insert favorite AI tool here] and have it return specific results. Like "Take this CSV and return MySQL insert statements with columns x,y, and z, ordered descending by the "Date Added" column". Another example was when our local county fair decided to publish an _alphabetical_ list of their scheduled performances in PDF form. There was no easy way to plan a day to visit the fair and decide which performances we'd like to watch. I used ChatGPT to distill several PDFs down into a chronological text list of performances, grouped by date and sorted by time. That would have taken me a LOT more time to do manually or with a hand-built script.
- I've been playing with the "cursor" IDE (https://www.cursor.com) for some of my projects, and it's almost scary how well the autocomplete works. I'm old-school and am used to crafting HTML code for a web form, then copying the HTML to another script in a comment block to do the actual form processing. The Cursor IDE actually _mostly_ recommended all the code needed to process said form post data and insert it into my database. It even removed the comment block when it noticed I had addressed all the form fields. I'm still blown away by the time saved by it!
But like others have said, _always_ verify the outputs of the LLM you are using, because it's not 100% accurate.