kane_magus: (Default)
I'm just linking to the blog, even if the comic associated with it (being part two of the previous comic, as implied by the title) is pretty cool so far.

Mainly, I just want to post this for Tycho's words about recent AI bullshit. And by that I mean I just want to copy and paste Tycho's words here, those relevant to the whole AI thing, anyway, because I mostly agree with them, without feeling a need to elaborate or anything. Besides, I've already posted about this shit.

(Links and italics are his [well, except that I inverted the italics below].)



Behind a cut, of course. )
kane_magus: (Default)
I post this mostly not for the thing itself, which is kind of stupid, but for one of the comments under it.

One of the best descriptions I've seen in a while, not of LLM/AI itself, necessarily, but of people's reactions to/acceptance of LLM/AI, is this:

"So often, the response to AI is 'this isn't useful for [thing I do] because it lacks the human creativity and ingenuity necessary to [do the thing that I do]. But I bet it would be really useful for [thing I don't do or know anything about but wrongfully think is mindless work devoid of creativity].'"

I will elaborate on that a bit. If one says that wholesale use of AI is "okay" for creating large swaths of code (or, indeed, entire, whole-ass programs/applications), because coding is "mindless work devoid of creativity," then that's the equivalent of saying that it's okay to use AI to create large swaths of prose (or, indeed, entire, whole-ass novels), because writing is "mindless work devoid of creativity," or saying that it's okay to use AI to create larges swaths of lines/colors (or, indeed, entire, whole-ass drawings/paintings), because drawing/painting is "mindless work devoid of creativity." The former is just as ridiculous and untrue as the latter two are.

I'm not even saying that you can't use AI or AI-like tools at all for these things. If a webcomic creator, for instance, wants to use a copy/paste tool to reuse backgrounds they've already drawn once, instead of drawing the same thing by hand repeatedly in every panel, that's fine (as long as it wasn't a LLM that generated those backgrounds for them out of whole cloth in the first place). If a writer has a recurring phrase that some character says in their book that they have set up to autocomplete when they start to type it, rather than having to type it out in full every time, that's fine (as long as it wasn't a LLM that generated said character and/or their recurring phrase out of thin air in the first place). I'm just saying treat "AI used for 'coding'" with the same respect/disdain, as the case may be, as you do for "AI used for 'writing'" or "AI used for 'imagery.'" Don't simply dismiss out of hand the "AI used for 'coding'" bit, just because you may not like/understand/do coding work yourself.

As for the article/blurb itself:

No, "AI code is" not "different from AI art and writing." What I don't want is AI code slop in the games, and that comes from outsourcing programming to gen AI. The only actual difference is that AI code slop isn't as readily visible in the end result as the AI "art" and "writing" slop is, but that just makes it actually even worse and more insidious, honestly.

"In a sense [Tim Sweeney] isn't wrong," First of all, Tim Sweeney is fucking wrong, because him saying that Steam should get rid of the AI tag is asinine. I'm in favor of there being more information available to the consumer, personally, not less.

"but is there a difference between using AI for coding compared to creativity?" Second of all, as the comment to which the above comment was replying, in agreement, also said: "Coding is creativity."
kane_magus: (Default)
A post on John Scalzi's Whatever.

The gist: Someone at the bottom end of the 1% is closer to someone with no money at all than they are to someone at the top end of the 1% who has billions of dollars, yet the person at the bottom end of the 1% is still in the 1%, so too often tries to live like the guy with billions of dollars. The whole thing is simultaneously utterly asinine but also makes sense that idiots with more money than brains would try (and fail) to live like that.

I read the whole post but only watched the video, at 2X speed, until about halfway in, bailing at the start of the shill. The post itself is more humblebragging about how John Scalzi and his family mostly don't do all the stupid shit that 1%ers do.
kane_magus: (Default)
A post on The Daily Cartoonist.

If by "Humpday's Happy Diversions," they mean "this post doesn't mention Donald Trump or Jeffrey Epstein or Charlie Kirk by name even once, but only just barely," then sure, but it's still not all sunshine and rainbows in this post, even so.

...

"This cartoon was posted on line and somebody commented of AI that 'it's just a tool.'

"No, man. You're just a tool."


...

"It would be hard for any reasonably intelligent person not to know about Taylor Swift, just as it was once hard not to know about Frank Sinatra or Johnny Carson. You don't have to like her or care about her. You just have to have an IQ in three digits and a minimal amount of curiosity.

"We used to celebrate something called the Renaissance Man, a type of Jack of All Trades, admired for knowing something about a whole lot of things.

"It wasn't confined to Leonardo da Vinci, either: Samuel Johnson spoke with some awe of Elizabeth Carter, whose intellectual accomplishments are jaw-dropping, but who Doctor Johnson admired for being a down-to-Earth person, saying 'My old friend Mrs. Carter could make a pudding as well as translate Epictetus from the Greek, and work a handkerchief as well as compose a poem.'

"The real snobs these days are not the ones who can translate Epictetus but those who sneer at anybody who can.

"For my part, I am curious and well-informed enough to know we have Cabinet members who like to cosplay as police officers, though I hope to god that, like Barney Fife, they're permitted to wear the uniform but not to load their guns."
kane_magus: (Default)
A post on WIL WHEATON dot NET.



"Wil Wheaton on His Star Trek Family and His New Podcast Storytime with Wil Wheaton"

"Hi everyone, happy Tuesday! I am so excited for this week’s episode. I’m talking to the one and only Wil Wheaton! You know Wil from his roles as Wesley Crusher on Star Trek: Next Generation, Gordie in Stand By Me and his appearances on The Big Bang Theory. Wil is also a super-nerd and a prolific audiobook reader. He has a new podcast called Storytime with Wil Wheaton, where he reads a new short speculative fiction story each episode and I highly recommend it! We had such a lovely conversation— Wil is a deep thinker and one of the kindest people I know. I can’t wait for you to get to know him a little bit better! Stick around after the interview for the hindsight, where my producer Jeph and I talk about the episode, as well as some upcoming live podcast recordings, our new Patreon and oh yeah, time travel!"



Apparently the only thing you have to do to get me to watch an entire hour and a half episode of your podcast is simply to have Wil Wheaton on as the guest for that episode. Like so.

Also, it's kind of funny, because I'm sure most other geeknerds (or is the proper term "nerdgeeks"? *shrug*) like me would know Katee Sackhoff from Battlestar Galactica. Here's the thing, though... I have never seen anything of Battlestar Galactica, either the 2004 version or the original show or anything else, outside of maybe an occasional clip on Youtube or something that I don't even remember now. (I always tend to get it confused with Babylon 5, as well, which is another geeknerd show that starts with the letter "B" and of which I have yet to see a single episode. How far in the toilet is my geeknerd cred now?) Nor have I seen any of that other nerdgeek stuff like all that Star Wars stuff mentioned on her Wikipedia page there. What I have seen a fair bit of her in, though, is Longmire, because there is an over-the-air TV channel which I don't recall the name of right now that my sister watches, and it airs a metric assload of Longmire episodes, back to back, on at least a weekly if not daily basis, and I see bits and pieces of it whenever I happen to exit my room to go the kitchen or something (also stuff like Stargate SG-1 and The Closer/Major Crimes and Rizzoli & Isles and NCIS and whatever else that channel [those channels?] tend to air, but that's beside the point). Beyond that, I have no other experience with pretty much anything Katee Sackhoff has done, unfortunately. At least as of right now. So far. Yet.

Okay, so...

There's a fair bit of overlap here with what Wil said in the Mayim Bialik podcast and what he has said on his own blog, but there's a lot of stuff that's new, too. That's one of the cool things about Wil is that even if he's telling largely the same story as one he's told before, elsewhere, he's still able to put a new spin on it. It's not just the same thing, over and over.

At one point, early on, Wil says that if he could give up all of his acting success if it meant that he would instead have a normal childhood where he had parents who weren't terrible, he would do it in a heartbeat. However, later on, about halfway into the episode, when asked about if he could time travel and change something, would he do it, he says that if he could go back and change the bad things about his childhood, he would not do it, if it meant that it led to him never meeting Anne, his wife. It was an interesting contrast. His acting career he would sacrifice in a moment, if it meant he could instead have had a good childhood with loving parents, but not his wife and her kids that he later adopted as his own. By the way, his stories about his wife's children, separately, asking him to formally, officially adopt them when they were each 18 was very touching.

Oh, and the short story they were talking about is Wikihistory. I'm glad Wil mentioned it because it gave me an excuse to read it again (this will be the fourth of fifth time now). In fact, I literally paused the video, then went and read it again, before returning to the video. It's definitely as good as he says it is (though he did get some of the details about it wrong, i.e. there was no "baby Schlimmel" or whatever).

And the fish story is this one or one of several like it told by Michio Kaku.

Finally... I haven't gotten around to it as of yet, but I think I'm going to start actually making the time to go through It's Storytime with Wil Wheaton now. Perhaps even getting started right now, in fact.

"There's always going to be shitty people in the world, that's just how it is, but we can choose whether we're going to be one of them."
kane_magus: (Default)
A post on wilwheaton.net.
kane_magus: (Default)
A post on John Scalzi's website, in which he ruminates about the foreign sales of his books and also once again rightfully castigates SCROTUS IMPOTUS Trump II and his mind-numbingly idiotic tariffs.
kane_magus: (Default)
"NaNoWriMo thinks writing is merely a suggestion."

It is an egregiously clickbaity headline and deck/drop head, to be sure, and the giant image of an upset Lisa Simpson is completely extraneous, and such is usually the case with my admittedly very limited (pretty much for that very reason) exposure to The Mary Sue.

But, in this particular case at least, the article itself under all of that superfluous crust at the top was actually kind of interesting. (Even including all the Twitter embeds. At the very least, it wasn't an "article" that consisted of around 98% of nothing but Twitter embeds, which is something I've seen far too often lately on other such "news" websites.)

In any case, if you don't want to go through The Mary Sue, here are some other alternatives, (possibly) with less sensationalism. The thing on The Mary Sue is just what popped up in my Google News App, is all. (And, yes, I just took that and used it as an opportunity to criticize The Mary Sue itself in the first half or so of this post, as much as the whole NaNoWriMo thing in the second half, below.)

Anyway.

Basically, the people supposedly "in charge of" NaNoWriMo apparently seem to be of the opinion that using AI to write is okay, actually, and that you're being "abelist" and "classist" if you criticize its use in writing, because poor, disabled people apparently need all the help they can get. Or some shit. The people whom they ostensibly claim to represent seem to vehemently disagree, however, to the point that they are cutting ties with the so-called "organization" altogether.

I know people who have participated in NaNoWriMo in the past, but I've never tried it myself. It always seemed needlessly stressful to me, but good on anyone who has ever succeeded at it. I'm not sure why there ever needed to be an "official" "organization" attached to the whole thing, though. Especially one with such dubious notions. When I first heard about NaNoWriMo, decades ago, I just thought it was an unofficial activity that people were just kind of doing, of their own accord, as a "viral Internet thing," rather than something that was an actual "official" Thing™, which "needed" an "organization" of people "in charge of it" or whatever. Like, who even decided that shit? And who let them get away with it, and why?

From the end of The Mary Sue article: "The bottom line is: Writing is free. You don’t need to wait for a specific month or be involved in NaNoWriMo to do it. There are writing communities everywhere, especially online, where you can find support, validation, and inspiration to finish that story you want to tell."
kane_magus: (Default)

Link to comic.

Link to blog.

Yeah, just to be clear, I'm making a post about this Penny Arcade comic more for the associated blog (and similar things) than for the comic itself. There are even mentions of John Scalzi there.
kane_magus: (Default)
It seems like common sense to not use fake quotes from real people in promotion of your book, but since Mr. Scalzi had to write this post about it, I guess it's not, at least for some people.
kane_magus: (Default)
A post on Wil Wheaton dot net.

As of reading this post is the first time I've ever heard of this Bagman game. It looks like the kind of game that probably would have eaten quite a few of my quarters if I'd known about it in ye olden dayes.

It was also cool that when Mr. Wheaton mentioned "my friend John gave himself a Daily December last month, where he wrote about a different comfort movie every day," I knew exactly who he was talking about.
kane_magus: (Default)
A post on Whatever.
kane_magus: (Default)
"Thoughts are swarming my mind like a nest of cyber-rats."

I haven't tried Sudowrite, and I probably won't try Sudowrite, because it's rather a bit pricier than the alternatives like NovelAI or HoloAI (both of which I have tried) or whatever. I don't doubt that it's better, given that it's using GPT-3.5/4, rather than the less powerful (at least for now) models used by those others.

But ignoring those specifics, just generally speaking, my take on novels/novellas/whatever written by AI is this: as long as you are completely upfront and transparent about the thing you wrote with an AI being written by an AI, then I'm fine with it. You say, "Hey, look at this cool thing that this AI wrote for me," then I'll probably take a look at it and be, "Oh, that's neat."[1] I've certainly done that myself, after all. And honestly, I'd rather read a fair-to-middling story written by an AI or a human/AI collaboration than read piss-poor dreck written solely by a human.

However, if you show me a thing and then say, "Hey, look at this cool thing that I wrote," I don't care how amazingly good it might or might not be, if I find out that you actually didn't write it, but instead it was written mostly or entirely by an AI, then I'm going to be way more blasé about it, at the very least. More likely, though, I'll be actively antipathetic toward it. Or, rather, toward you. Just as I would be if you came up and said, "Hey, look at this thing I wrote," and it turned out that another not-you human was the one who wrote it.[2] And, let's be honest here, anything you get an AI to write for you, at least for now, is going to be the equivalent to something written by not-you humans, given that these AIs are trained on the writing of not-you humans. Even if you supply it with samples of your own writing, it's still going to be pulling from the writing of other not-you humans (unless, somehow, it's an AI that was trained solely on your own writing and nothing else).

TL;DR version: I think there's a place for AI-written stuff. And I think there's a place for human/AI collaborative efforts, as long as the human is honest about it. I don't, however, think that there's a place for humans passing off AI-written stuff as human-written stuff. Or, at least, I don't think there should be a place for it. And I'm not sure how I feel about humans directly profiting (whether it's monetarily or "just" reputational or whatever) off of the "work" of AI, regardless of whether the human is upfront about it or not, given that AIs are trained on the works of other, not-them, humans (willing/knowingly or, as is most likely the case, not so). If I had to say how I felt about it, I would say it feels a little hinky to me, at very best.

[1] - Disclaimer: I have not read The Electric Sea, and I'm not sure I'm going to. I'd guess that it's still worthy enough of an "oh, that's neat" and all, but I'm simply not all that interested in reading the whole thing myself, at least not at the moment. There it is, though, for anyone else (or even myself, later) who may want to give it a read.

[2] - Just take a look at my review of Reality Hunger: A Manifesto by David Shields if you want to see how I feel about the subject of plagiarism (or almost-plagiarism or plagiarism-in-everything-but-name). Essentially, I felt strongly enough about it that it's one of the few books on Goodreads for which I wrote an actual review as opposed to just leaving a star rating. Not to put too fine a point on it, I'm not a big fan of plagiarism, whether it's wrapped up in pretentious bullshit or not.
kane_magus: (Default)
A post on Whatever.
kane_magus: (Default)
An article on NPR about fan fiction.
kane_magus: (Default)
I saw mention of this story elsewhere but didn't feel the need to make a post about it, until now.

Honestly, anyone who writes a story using an AI and then seriously tries to submit said story[1] as their own work deserves every bit of the censure and ridicule said action will bring down upon them. (And anyone who may have trouble telling the difference is similarly worthy of censure and ridicule as well.)

[1] - Even heavily edited, and if you're just going to do that anyway, you may as well just go on ahead and write your own, original, non-AI-"enhanced" story and be done with it.
kane_magus: (Default)
(See also: "The Island" and "The Boy and His Mother and Grandfather")

Just to follow up a bit on the previous post in a sort of tangential way, I decided to just let NovelAI ramble for a while. All of these were made using the "Krake" AI (the one that's only available at "Opus" tier subscription and is, therefore, the strongest one [ostensibly]), with default settings, no Memory or Author's Notes or Lorebook entries. In these cases, I used a very simple prompt, which was this: "Let's tell a story, shall we? Let's see how far we can go before non sequitur nonsensical nonsense rears its ugly head." That was it. That was my sole input. Everything after came from the AI itself. I didn't "Retry" a single time. I stopped at the first point it tried to do a "chapter break," indicated by a "***" (or until I simply decided it had rambled for long enough, whichever came first). And I guess I'll keep doing this until this post tells me I've gone past the allotted post size, or until I get tired of messing with it. Again, whichever comes first.



Example #1: I don't know if Kanye West actually said that. Nor do I know if that's an accurate description of Fullmetal Alchemist at all, having not read/seen it myself. But I would guess... not? Also, this is the only one I just stopped, rather than letting it go to a chapter break. )



Example #2: This one is relatively short up to the first break, but it got really dark, really quick, but at least it seemed to recognize that it was all nonsensical nonsense, I guess. It also couldn't maintain a proper POV, switching back and forth from first and second. )



Example #3: Warning, this one actually went to the NSFW places I thought the previous example was heading toward but didn't. This one seemed to be the most... stream of conscious of the three. Also, this one has some dialogue, at least? Anyway, I'm going to stop here, with this one. I'm actually surprised the DW post was able to hold all this much text without balking. )



Anyway, yeah, those are just three examples of how this one particular AI can take the same simple prompt and run in vastly different directions with it. I may try this "experiment" again at some point later, just for shits and giggles. (EDIT) Actually, I won't be doing that, at least not any time soon, because I've since canceled my NovelAI subscription (yet again). Not for any specific reasons, this time, but simply because I don't think it is good enough (yet) to justify spending $25 USD per month on it. Maybe I'll give it another try in 6 months or a year or so. (/EDIT)
kane_magus: (Default)
Like, at all. (Dreamwidth posts don't count, not really.)

I've left MLP/Star Trek crossover fanfiction behind. (Like, duh, since I haven't written any of that since before moving back to NC from WA). Or, perhaps more appropriately, it left me behind, given I kinda stopped watching MLP:FiM after the start of season 5 or so, even though I always keep meaning to go back and watch the rest of it "someday." (Sure, I'll get right on that after I finish Pretty Guardian Sailor Moon and/or Sailor Moon Crystal which I've also been meaning to get around to watching "someday." In any case, the likelihood of doing that getting me to write another Sailor Moon/Castlevania fic or whatever is about the same as [re]watching MLP is to getting me to finish those MLP/Trek things. I.e. not too awfully high at all.)

But there's always that novel that I've been "working on" for the past... *counts on fingers* ...eight... no, more like almost nine by this point... ...years... ....goddamn..... now... and on which I've made next to no progress for about half that time.

Or the other novel that I've been working on occasionally thinking about for even longer than that, by like at least a whole decade (talking even pre-DigiPen here), and even wrote a shitty prologue for at one point.

Hell, I barely even write in my damn pencil-and-paper journal anymore. Maybe once or twice a month at most, these days.

Thought I'd cheer myself up a bit by listening to some music... but then that just got me thinking about Alec Holowka again... which probably isn't the best thing for my frame of mind right now. And yet, I'm still listening to it even as I type this thing here, whatever this is. Oh well, at least I haven't reached Shivers 2: Harvest of Souls levels of funk, yet... oh... wait... (Seriously, I love that song, and it's my favorite of that soundtrack, but it [and some of the others] can be pretty damn depressing, if I'm in just the wrong sort of mood, which appears to be the case tonight.)

Blargh. Should probably just go to bed already. Enough 報復性熬夜 (and self-referential links) for one night, I guess.

</whiny, effete rant mode>

Profile

kane_magus: (Default)
kane_magus

January 2026

S M T W T F S
    1 2 3
4 5678910
11121314151617
18192021222324
25262728293031

Most Popular Tags

Style Credit

Page generated Jan. 6th, 2026 03:48 pm
Powered by Dreamwidth Studios