pathdependent

not adhering to the prompt guide is def a valid strong criticism. resampling i think less so for the demo just because fewer people look at k samples per model, so just taking literally the first one has the fewest of my own biases injected into it

hrm. yea you're right. the page on fal used to produce it was linked with the image, but maybe i made a mistake and sloppily saved wrong one. ill have to reroll to check

Shameless plug:

This (very) thin wrapper lets you write your PlantUML code in Jupyter cells.

https://github.com/jbn/IPlantUML

To play with some ideas, nothing is more accessible than NetLogo: https://ccl.northwestern.edu/netlogo/

I'll be posting some things to HN soon. My writing style is very accessible, to anyone. I always shoot for an "explain it to your grandmother"-level of writing.

Also, my entire dissertation is Python/Cython/Jupyter with a prebuilt docker container, so, it's also low-friction to play with. But, complicated by dissertation rules for sharing pre-publication (i.e. "is this definitely your work, because it's on github and you accepted PRs?").

This is my field, too. I am a Ph.D. candidate in computational social science, working on a computational model of social belief systems.

I'm quite happy to see this on HN. Like many other disciplines, our field is blossoming with the rise of cheap, high performance computing!

If your involved in this field and on HN, please reach out to me at @generativist on twitter. I'm in the process of setting up a community for CSS work, and will be soliciting feedback soon!

This is a great point that I wish I wrote about.

I think this is actually the source of the "steep learning curve" problem. Once you've used it enough, it is an almost orthogonal part of your brain that does thats the VIM command-ing. It becomes more like, "to walk, put one foot in front of the other." You don't actually process that thought when walking.

But, until that point, it absolutely IS the same part of your brain. Until that point, there IS a PAINFUL contention between work brain and editor brain.

Buried in my footnote is the real gist is: learn your editor, deeply.

IntelliJ is a great tool, too.

Full quote, for context:

   Being conservative, let’s say it makes you 5% more productive.

It's subjective. I'd like to see a before and after study, but empirically measuring code productivity is -- a hard problem.

I admit, 5% is a subjective prior / fudge-factor. But, statistical white-lies communicate a promise which extends the invitation: try it, see if you're more productive.

I'd love to see some objective stats though!

Yes.

The had-I-learned-it-first concept is something I didn't mention, but do think is important. And, often it works in an inhibiting way. With an IDEA-like rich IDEs -- which I think is a great tool, esp with VIM bindings! -- you get such a boost from things like good contextual auto-complete, that you never feel the need to learn it more deeply.

Thanks!

I put VIM in the title because 1) it's got a recognizable identity to pull in readers and, 2) I use VIM. But, in a weird and almost recursive way, editor wars are kinda like language wars. If it gets you that 5% (or more) boost, great! But, the differences between them often washes out.