Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Magic Loops – Combine LLMs and code to create simple automations

magicloops.dev

Post reply on HN
Print view View original

11–20 of 53 posts

Re: Show HN: Magic Loops – Combine LLMs and code to create simple automations

#11

Post by notahacker »

I do like the way your UI strips out some of the ambiguity in how your initial prompt is interpreted by an LLM and allows for editing and parameterisation and IFTTT outputs. Especially since exactly how LLMs parse complex instructions is hard to test. There's definitely a sweet spot between "type instruction and hope for magical understanding" and "write the program yourself"

(though as it still involves LLMs I'd probably want to use it for stuff that sounds less mission critical than some of your examples!)

Re: Show HN: Magic Loops – Combine LLMs and code to create simple automations

#12

Post by jondwillis »

notahacker wrote: post #11

I do like the way your UI strips out some of the ambiguity in how your initial prompt is interpreted by an LLM and allows for editing and parameterisation and IFTTT outputs. Especially since exactly how LLMs parse complex instructions is hard to test. There's definitely a sweet spot between "type instruction and hope for magical understanding" and "write the program yourself" (though as it still involves LLMs I'd pro…

+1, I love how the blocks provide transparency. I have been mulling over the UX for a somewhat similar app that I have been working on, and I may borrow some of how this works.

Re: Show HN: Magic Loops – Combine LLMs and code to create simple automations

#14

Post by jumploops »

notahacker wrote: post #11

I do like the way your UI strips out some of the ambiguity in how your initial prompt is interpreted by an LLM and allows for editing and parameterisation and IFTTT outputs. Especially since exactly how LLMs parse complex instructions is hard to test. There's definitely a sweet spot between "type instruction and hope for magical understanding" and "write the program yourself" (though as it still involves LLMs I'd pro…

Completely agree -- LLMs are super powerful, but code is predictable. We found the best loops are ones that are almost entirely code (built with the LLM's help of course!) and only use LLM blocks for very specific and repeatable tasks, if at all.

We see a future where the code can be "self-healing" as well, with user approval of course.

Re: Show HN: Magic Loops – Combine LLMs and code to create simple automations

#15

Post by jumploops »

jondwillis wrote: post #12
notahacker wrote: post #11

I do like the way your UI strips out some of the ambiguity in how your initial prompt is interpreted by an LLM and allows for editing and parameterisation and IFTTT outputs. Especially since exactly how LLMs parse complex instructions is hard to test. There's definitely a sweet spot between "type instruction and hope for magical understanding" and "write the program yourself" (though as it still involves LLMs I'd pro…

+1, I love how the blocks provide transparency. I have been mulling over the UX for a somewhat similar app that I have been working on, and I may borrow some of how this works.

Happy to chat about some of our decisions (or lack thereof!)

Re: Show HN: Magic Loops – Combine LLMs and code to create simple automations

#20

Post by MihaiBulic »

obiefernandez wrote: post #17

What are you using to scrape?

We’re using Apify to scrape with their Web Content Scraper actor. It’s nice because they handle proxying for us which decreases the likelihood of us getting blocked. Their “to markdown” functionality is also a lot better than libs you find on npm. They are expensive tho
Post reply on HN