Basically a deer with a human face. Despite probably being some sort of magical nature spirit, his interests are primarily in technology and politics and science fiction.

Spent many years on Reddit before joining the Threadiverse as well.

  • 0 Posts
  • 1.22K Comments
Joined 2 years ago
cake
Cake day: March 3rd, 2024

help-circle

  • Okay, wow. I’ve garnered plenty of downvotes on the Fediverse by not auto-hating many of Microsoft’s new features and updates, I’m sure I’ve been labelled a “Microsoft shill” or somesuch in some folks’ user notes. But this is just ridiculous.

    The single most important rule Microsoft should have is “thou shalt not brick thy customers’ computers with a routine update.” Sure, it’s not the most common set of triggering conditions in the world, but the problem is immediate and obvious upon booting up. How do they not have a test plan that would catch this?







  • I just recently came across Jules, an autonomous coding agent by Google, and was impressed with its free access level. I dumped a whole directory full of haphazard Python scripts I’ve been accumulating over the years into it and asked it to refactor them into less of a mess and it did a remarkable job turning the folder into somewhat of an actual application. You can hook it up to your Github account too, if you want, and it’ll submit its changes as a new branch.

    Just bear in mind that it’ll make mistakes, I did have to do a thorough debugging run to make sure everything still worked the same. But the amount of grunt work it saved me was huge.


  • Lately I’ve been using ChatGPT in rotation with a bunch of other LLMs, since I don’t want to habitually use just one and miss out on developments by others. I’ve found that I am gravitating towards using ChatGPT for language-related stuff - “what does slang term X mean”, “could you write up a speech for a character to say”, “create a detailed description of a magic item”, stuff like that. I also sometimes ask it to generate images, though not so much now that the GPT image tool is available through Bing’s image creation interface. I mostly use local AI image tools nowadays, they’re much more controllable, but ChatGPT’s images are often a great starting point.






  • Thanks for giving it a fair shake, and not just defaulting to “AI bad!”, though. I figured it was unlikely that you’d like these specific songs, since the whole point is that they are specifically for me - often a result of an in-joke that exists solely inside my head (the reason I made so many songs about the RS-232 standard, for example, is based on a gag from an episode of the British TV show “Spitting Image” that I saw maybe 30 years ago at this point).

    But that’s the point, in the end. These songs are “mine”, they don’t need to be anyone else’s. I just hope that a few of them at least elicited a chuckle, if only at the fact that they existed. :)



  • Alright, here’s what’s probably my most diverse and “fun” playlist: Funny How Science Works. It’s a collection of songs about science and technology. Each song often starts out technical and educational and then veers off into insanity. For example Tectonic Truth, which is about plate tectonics, and Along the Lines of Blaschko, which is about Blaschko lines, a pattern of cell differentiation found in females but not in males as a result of X-chromosome inactivation (I think this is probably the only song in existence about Blaschko’s lines). Others are a bit less deranged, for example Million Player Co-Op is just about how awesome ants are. And speaking of ants, Little Pilot is about Dicrocoelium dendriticum, the liver lancet fluke, a parasitic flatworm that uses ants as an intermediate host.

    Criterion Three is about how the IAU reclassified Pluto as a dwarf planet in 2006 and in particular it’s about how much of a hypocrite S. Alan Stern was about it. This is perhaps one of the pettiest and most specific diss tracks I’m aware of. Speaking of diss tracks, The Pro-K is sung by a prokaryotic bacterium dissing eukaryotes for how unreasonably complicated our cellular structure is.

    If you’re leaning anti-AI then you might enjoy my AI Apocalypse playlist, which is a bunch of songs along the general theme of AIs taking over the world. I’m not personally concerned about the stereotypical Skynet bombs-and-Terminators scenario, that’s Hollywood, but I do have some concern about the emergence of a Super-Persuader AI that’s better at convincing people to do stuff than humans are. So a number of the songs are along those lines.

    Many of the songs in these playlists are ones where I wrote pages of detailed prompts and tinkered with the results, since they involve personal philosophies and technical scientific subjects that I wanted to make sure were right. Here’s an example of a song that was generated with just a simple prompt to see what would come out; Insert Romance Here, which I generated by simply telling the AI “generate the most absolutely generic possible romantic song.”

    Here’s a playlist of songs that’s entirely about the RS-232 standard, a standard introduced in 1960 for serial transmission of data.

    A lot of the songs in these lists have probably only been listened to by me, nobody else has ever heard them. It’s a little annoying that Producer.ai, the service I used to generate all of these, doesn’t support attaching an “author’s note” to each song; many of these only really make any sense with a little bit of context about what was in my head when I generated them. But hopefully most of the ones I’ve made public are self-explanatory to some degree.


  • It’s a difficult question to answer, “why do you like the things that you like?”

    There is surprise that can be had at the generation phase. Sometimes I’ve got a particular detailed thing in mind that I’d like to hear about, I can end up writing pages of prompts, hand-tweaking the lyrics extensively, and regenerating bits until it’s just right. Other times I’ll throw out a very general vibe and just see what the AI comes up with. Some of my favourites are songs that definitely surprised me.

    I’m actually not very familiar with the specific ways of describing the sound of different styles of songs - pop-rock, folk ballad, driving bass lines, all those sorts of descriptive terms are lost on me. So I usually let the AI come up with something for that based on a more general vibe, like I tell it a song should be “upbeat” and “informative” and it’ll come up with something. I get to hear a lot of variety that way.

    Would you like me to link specific examples? Honestly I was expecting more of an “AI sucks and you suck” kind of response but I guess there’s enough responses to this post that I haven’t been noticed in among them all, so I could show you a playlist or a few specific tracks that show the kind of music that I enjoy that I don’t think I’d have ever been able to find in significant quantities from “traditional” music sources.


  • I care a fair bit about music, but in a way that’s likely very unpopular around these parts.

    I’m a fan of AI-generated music. I’ve got about 600 songs I’ve generated over the years and almost all of them are about things and are in styles that I personally find relevant to my interests and life.

    Some of your other questions are rendered moot by this. There’s no albums or artists for these songs. I’ve loosely grouped them into a few giant playlists based on the sort of mood they fit. I listen to them while doing things like walking my dog. I don’t use particularly fancy earbuds or other hardware, though I made sure to get decent speakers for my computer for while I work on them to make sure they’re high quality at their source.

    Since most of the songs are personally meaningful to me, if only because they scratched a very particular itch I was having the day I generated them, I care about them. I don’t expect most of them would be of interest to other folk. I mostly don’t share them since they’re for me and anyone else can easily generate their own custom stuff, though I’ve made some public when they’ve struck me as particularly amusing or interesting.