I love tools.
-
I love tools. That's a big part of my identity, personally and as a software engineer. It's served me very well!
Tools are empowering, but tool use is a skill. You need to search for tools, select them, learn them, use them appropriately, and maintain them. There's a craft to all of these things, and I delight in it.
The more I practice these skills, the better I get at them, and the more capable and confident I become! Not only can I do more, I know I can extend myself whenever I need to. I can learn new tools to meet a new challenge.
(1/3)
-
I love tools. That's a big part of my identity, personally and as a software engineer. It's served me very well!
Tools are empowering, but tool use is a skill. You need to search for tools, select them, learn them, use them appropriately, and maintain them. There's a craft to all of these things, and I delight in it.
The more I practice these skills, the better I get at them, and the more capable and confident I become! Not only can I do more, I know I can extend myself whenever I need to. I can learn new tools to meet a new challenge.
(1/3)
One thing that really surprised me, going back for my PhD, is how many computer science researchers don't particularly like tools.
They depend on tools to do the work, but they only care about the work. Having to learn a new tool is a bother, and they will go to great lengths to just "hack something up" rather than read a manual or find a better way.
LLM coding assistants are very popular with folks who think like this, and it makes sense. I hope it helps them pursue what matters most to them, but it hurts me to see it. I feel like they're missing an opportunity.
(2/3)
-
One thing that really surprised me, going back for my PhD, is how many computer science researchers don't particularly like tools.
They depend on tools to do the work, but they only care about the work. Having to learn a new tool is a bother, and they will go to great lengths to just "hack something up" rather than read a manual or find a better way.
LLM coding assistants are very popular with folks who think like this, and it makes sense. I hope it helps them pursue what matters most to them, but it hurts me to see it. I feel like they're missing an opportunity.
(2/3)
Do I go slower, because I spend time developing my toolkit and workflow for research, instead of just doing the research?
Briefly, yes, I do. Then I go faster. Everything becomes smoother, cleaner, more effective.
I know my code inside and out. I know it does what I want. I know what my results mean.
I don't have to rehash the same requests with a chat bot each time I start a new project. I don't have to proofread and correct some simulacra of software design. I don't even have to describe my idea in words, I can just start doing it.
In the long run, I think I'm much faster. More importantly, though, I'm growing, I'm having fun, and I'm making things myself that I'm proud of.
(3/3)
-
R ActivityRelay shared this topic
