3 ms·
I like using LLMs for building non-critical tools that make me more productive. Things like shell scripts, Github actions, or one-off tools for visualising some
by yen223 2y ago
I like using LLMs for building non-critical tools that make me more productive. Things like shell scripts, Github actions, or one-off tools for visualising some problem space. The kind of thing where code quality doesn't really matter, but which will save you a ton of time in the long run
- edanm 2y agoThis is an enormously fruitful part of using LLMs for me too. As a programmer, there's always a million small scripts I can think of to improve my life, but it's often not worth spending time actually building them. But with LLMs, I can now usually build a script, even in a domain I don't know much about, in minutes. I don't think this would work as easily for non-programmers yet, because the scripts still aren't perfect and the context of actually running them is sometimes non-trivial, but for me it works very well. (I say scripts, but sometimes even tiny one-off things like "build me this Excel formula" is helpful.)
- bawis 2y ago"even in a domain I don't know much about, in minutes." How is this done exactly ? can you provide an easy example for me (not a programmer but I use python/R from time to time)
- Breza 2y agoHere's an example. I had a ton of XML files in an external drive and I needed to delete all the files that didn't contain a certain phrase. I had Claude write a bash script to move all the non-matching files to a new folder, spot checked a few to confirm that it had worked properly, and deleted them. My bash skills are decent but it still saved me time.
- emgeee 2y agoI've found them incredibly useful for writing Dockerfiles or other bits of infra config like K8s yaml
- vdvsvwvwvwvwv 2y agoEven just using your ai assistant in your .zshrc / whatever to add useful aliases etc. is a good win.