Hacker News new | past | comments | ask | show | jobs | submit login

I hand-wrote perhaps 10-20 lines of this project:

https://github.com/williamcotton/guish

The rest is Claude 3.5 (with a dash of GPT-4o) with a LOT of supervision!

I'd say I'm about 8 hours deep and that this would have taken me at least 30+ hours to get it to the current state of polish.

I used it to make some graphs at work today!




Quite interesting — but how is it fundamentally more productive than being in VS code in R or python? You don’t get any of the benefits of an IDE here. I often find myself doing very similar workflows but default to either VS Code or the shell. Trying to imagine this truly making workflows faster/easier/more efficient, but can’t figure it.


Maybe it isn’t? I am just experimenting with new UX! Maybe it could be integrated into an editor of… the fuuuturrre!

But seriously, do you have any thoughts or suggestions?


Unfortunately not! I do very similar workflows honestly but have basically defaulted to the norm or compromises like: - write own cli in python that does the sql wrapping and basic transforms and cleaning - read in the csv with R to make plots and tables.

Boring.

Cursor with 3.5 sonnet has made all this way faster so that’s nice. Often LLMs are now featuring in these pipelines and I see libraries like data bonsai and instructor being helpful. But yeah idk. No bright ideas here but always on the lookout to optimise.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: