I wanna see someone do Hocus Pocus, with extra points for the insane facial expressions from the version performed live on Midnight Special:
2,095 karma · joined April 6, 2023
Email me here! eric@epiccoleman.com
Videos: https://www.youtube.com/@epiccoleman
Also been playing a bit of Blood Dungeon, which is pretty fun. I've also got a Factorio save that has mildly petered out after landing on Vulcanus (which I have kinda taken advantage of to break the addiction lol, but I'm sure I'll get back in there sometime over the winter).
I've been kinda deliberately taking a bit of a break from programming "side projects" after spending a humongous amount of free time and energy on a conference talk I gave a few weeks ago. It was titled "Human Creativity in the Age of AI" (which I later learned was an unfortunately generic title, lol), and I built this whole big musical and visual contraption that drove the presentation, including a "guitar solo generator" which recuts phrases of my own playing into a new solo on the fly.
It was a really rewarding project overall but definitely made me remember why I hadn't done public speaking for quite some time before - I am way too perfectionistic about it and it just becomes a massive time sink.
This one was worth it though, got a great reception and I'm really proud of the talk on the whole.
In the 2-3 weeks after the talk I spent a ton of time editing and laying out a really nice video of it, which I am also quite proud of. I didn't know anything about Davinci Resolve and I was able to get a really nice jump start with some help from Codex. I'm sure I don't know 98% of what a software package this big can do, but I did learn enough to be dangerous, and also enough to cut together some nice short clips of the talk as well.
It's been pretty rewarding to get a handle on a powerful bit of creativity kit like Resolve. It's incredibly capable and I'm really delighted with the results I was able to get. Obviously I am a rank amateur but to my untrained eye, my videos look pretty damn good. (And I could go on for another few paragraphs about how much work it took to get good results from the relatively shitty source video and audio I was working with).
I'm planning on doing some writeups because there was so much interesting tooling that came together as I worked on the talk and the videos, I could mine this one project for a year's worth of blog posts if I wanted to!
Anyway: links.
OrcaSlicer calibration guide - seriously great: https://www.orcaslicer.com/wiki/guides/calibration_guide
Blood Dungeon: https://store.steampowered.com/app/2431080/Blood_Dungeon/
I assume a link to Factorio is unnecessary here, ha.
My talk "landing page": https://epiccoleman.com/tripulator
Talk video: https://youtu.be/kD5qYiLIhJc
One of the shorts - the guitar solo part. I'm really stoked on the "presentation" bells and whistles in this video (though I think they're a bit mismatched to the regular intent of YouTube Shorts - analytics indicate that the intro part I slaved over is probably too long and people swipe away before the good stuff, but oh well).
There's a bit of a hump you have to jump through when you're first getting used to having separate modes for input and editing / navigation, but that's honestly the worst part.
After that, you just need to notice things that you need to do often, and then go "I bet vim makes that easy." You pick up a little trick and use it. You probably don't remember it the next time, but you remember there was a way, and you look it up again and reinforce it. You eventually remember the things that were useful to you, and forget the "tricky stuff" that wasn't.
Over time, you start getting your head around how it works, the whole verb / subject thing, and then you start learning to speak the language. You stop thinking "ok, i need to go to normal mode, then type c, then i, then t" and start thinking "change inside tag".
Vim's not really about "editing super fast." It's more like a different mental model of editing text that really, genuinely, is better than the default mode on offer in your standard text editor. It's more keyboard driven and composable and just _capable_. Text editing operations that are annoying in standard editors are trivial in vim.
I think these little games are neat, but ultimately wrongheaded. The best way to learn vim is to use vim to do vim things. I've been using vim (or, more often, vim editing modes in VSCode or IDE du jour) for around a decade, and I bet I don't use 95% of what it can do. But that 5% I do know is extremely useful.
Also though: it's a nerdy text editor, and it's fine if you don't like it or want to learn to use it.
But - there's no reason this couldn't actually be a good thing. I can totally imagine a platform[0] where the focus is less on creation and more on curation. I sub to some person or community who's going to share good links, and I see only that, instead of the slop firehose from The Algorithm.
(and yes, social media was already more like this, in the early days).
[0] and platform implies more than I mean, could just be some nice wrapper around RSS or activitypub or whatever
Developing the talk was ton of fun, and I learned a lot in the process. (Cutting the video was interesting too, turns out AI is extremely helpful for someone just getting their head around Resolve).
There's some extra info and a link to the repo at my site, too: https://epiccoleman.com/tripulator.
I'd love to hear you feedback, thoughts, discussion, whatever!
Again, not claiming it was right, or good in the long run - just that Carlin points out the context around the decision makes it seem less insane than it would by today's standards (hence "Logical Insanity").
but there is something at least a little interesting about the notion that people working in the nuclear space they thought they might be saving the world - because in some sense, they... did?
I've always felt like the fact that no nukes have ever been dropped (in war) after 1945 was a pretty huge success story for the human race. We invented this insanely powerful weapon and, despite the typical pattern of humanity, we managed to solve one of our biggest "coordination problems" effectively enough to not use another one for 80 years (and counting).
I don't know that I'd extend this analogy to the AI labs, I'm not necessarily implying that they're all right about "saving the world," I just think it's worth noting that there was a contingent of people out there trying to save the world from a real threat and who seem to have - at least for now - succeeded.
Take a look at https://github.com/epiccoleman/tharmas/blob/main/schooner/do... for an example if you like.
Apparently I haven't pushed, shelfmark used to be `calibre-automated-book-downloader`
And yet: I have not written assembly since school except for one Advent of Code challenge, and I expect it would take me quite a while to pick up a productive level of facility with it.
I found this Youtube playlist which includes a bunch of people's different takes.
https://www.youtube.com/playlist?list=PLHxPId0U7JNhALhwS-3kD...
(after reading it I am slightly disappointed - the piece is still great but my memory of it was longer!)
you could easily argue that there are plenty of things our government has done that matter much more in the grand scheme of things, but the rampant twitter / x meme posts and absolute horseshit like this "arcade" site disgust me. it's not funny or cute, it's despicable.
Still, I'm happy they're doing this, it seems like a cool product and I bet it would be almost as much fun as a platform to hack around on as it will be for gaming.
In the case of codex, at least, I felt some friction about just how much of my data I was willing to share, and my level of comfort meant that most of the advice was fairly generic and based on rough numbers rather than true analysis of statements, transactions, etc.
If you can find a way to tell a good story around data privacy I think that might be a good method to stand apart from the pack (both in terms of competing products and in terms of "just a guy who puts his tax return into Codex").
It's the kind of thing that sounds insanely boring on the surface, but there's a lot of interesting problems to solve, lots of coordination between various parties, lots of "deep state"[0] politicking - and so it becomes a surprisingly cool area to learn about and work in, and really matters in a way that many software projects do not.
Best of luck - it's a cool problem space.
[0]: I use "deep state" here somewhat ironically, but also just in the literal definitional sense: "people who are not elected but who work for the government and are surprisingly important". 0% intent to use it in the often insulting manner of popular discourse.
I guess the potential failure mode is ... no one ever showing up. I've got a standing slot (during work hours) at my current company where it's just "AI Jam" - just come in, eat lunch, and we'll hack on something, ramble about the news, etc. It's been moderately successful but there are definitely weeks where you just end up in a call with one or two guys with nothing to show.
That's not so bad though - last week we just talked about video games. Still fun.
So anyway - best of luck. Seems like a cool idea!
I think it's a bit unfair to draw the line here. Some of the most memorable music in my life has been from video games and I'd hate to lose the human touch here just as much as I'd hate to lose the human touch at live shows.
DOOM 2016, FTL, the Donkey Kong Country games, Mega Man and the Mega Man X series - and of course anything Koji Kondo's done - just to mention a few off the top of my head. I've never really been the type to listen to game music outside the context of a game (unless I've got a particularly bad earworm), but man, I do appreciate good game music.