Category: Life

Posted on
by

Last weekend we went to WorldCon / LAConV for a day. It was our first convention since 2019 (for, you know, reasons), and Katie finally got to build a rose goblin and cosplay as October Daye, something she’s been planning to for years. (Basically since we were gifted the jacket.)

Spike is described in the books as a cross between a rose bush and a cat. Three craft store grape-vine wreaths, soaked for reshaping, give it structure. Bougainvillea vines are woven through it to add thorns, though she ended up trimming them so it wouldn’t be hazardous if someone grabbed it, or if someone else’s costume brushed up against it. (We both managed to stab ourselves on the thorns while she was making it.) She cut a bunch of artificial leaves to look like rose leaves, and sewed them onto the vines.

For the eyes, a pair of LEDs are hooked up to a battery (the switch hidden under leaves) and placed inside mini ping-pong balls, held in place inside the head by wire.

Originally she was planning to put it on a skateboard and pull it along behind her, holding it on using magnets and washers, but the magnets turned out not to be strong enough. So the day before the con, she scrapped the skateboard and built a handle into Spike’s back so she could carry it around. It turned out to be better anyway, since the skateboard wouldn’t have gone on an escalator!

Posted on
by

…and sometimes, unreality is intentional and benign.

Today, Ars Technica posted an article with a header image featuring a keyboard modified to add an “Ars Subscribe” button. But the keyboard looked weird: the visible keys looked like this:

OPGU↵
LS.,
OC/[Ars Subscribe]

The extra letters, especially the two O keys, immediately drew accusations of AI slop generation.

And yet the image was credited to Aurich Lawson / Getty.

I figured Lawson had grabbed a stock image without noticing it was slop and then added the Subscribe button. So I did some reverse image searches. And it turned out that the apparent source photo was uploaded to iStockPhoto in 2016 (Getty owns iStock now), long before the advent of modern “AI” image generation.

That left the question of why the 2016 image had weird keys, though.

Then @lw replied on Mastodon that it matches a Turkish keyboard layout with the diacritics missing. I looked back at my search results and noticed another image that looked the same, only with the extra keys blanked out. The same account uploaded this one in 2018, and this time iStock had tagged it as having been uploaded from Turkey.

So, mystery (mostly) solved! It’s a real keyboard layout. Almost certainly a real photo, deliberately edited by hand to remove the extra markings. Why? Probably the original uploader figured it would be less distracting to a global audience. Who could have predicted that a decade later it would end up looking more suspicious because of it?

The discussion picked up in the article’s comment thread, where narteb found an even older version from 2014, still with the missing diacritics, but with a different action button and a darker (presumably original) skin tone on the finger.

A Proxy for Quality

It’s no surprise that readers were so quick to assume the photo was flat-out generated slop based on the apparent mismatch (not counting the obviously intentional subscribe button). “AI”-generated images have flooded the zone, and while there used to be some obvious tells — badly drawn hands, for instance — they’re getting to be a lot subtler.

Now we all second-guess everything we see, because to do otherwise would leave us open to being manipulated by falsehoods. And it’s not just knowing whether the image shows something real, but to know whether the image represents real human intention. The *ahem* key aspect of this image is the big red subscribe button, which is obviously not real, but it’s intentionally not real.

No one was under any illusion that the photo was an actual, unmodified keyboard. But when people assumed it was generated slop, they also assumed that was an indicator of laziness. When you see an article accompanied by an obviously junk image, you wonder, what else did they skimp on? Did they generate the article too? Did they research it properly? Did they fact-check it? This particular article was just a “here’s our new forum” announcement, but what about the other stuff that they publish?

Trust

Earlier this year, Ars retracted an article that contained fabricated, LLM-generated quotes. They fired the reporter and spent time developing an AI usage policy. This kind of thing burns trust, and trust is a news publication’s most important asset.

And it doesn’t even have to be a real mistake, or a real bias, to do damage. Conservatives spent decades painting NPR and PBS as some sort of bizarre “Marxist” strawman parody of what they actually broadcast, forcing them to dig out of a hole with large parts of their potential audience, until they’d lost enough trust that it became politically expedient to defund the networks.

Gut reactions and snap decisions have a purpose. But they’re still only heuristics. When we have time, it’s good to second-guess our impressions. Sometimes they’re wrong. Sometimes they’re right. But we have to look at those impressions clearly, not just spend the effort justifying the conclusion we’ve already reached. That’s something the human brain is also really good at, and when we’re cherry-picking evidence for the wrong conclusion, that makes things even worse.

I saw at least one commenter on the thread following up and insisting it couldn’t possibly be a Turkish keyboard, because those would have the diacritics on them, and this one doesn’t. It reminded me of the character in Mystery Men rejecting the idea that the Clark Kent expy couldn’t possibly be the Superman expy on the basis that he wears glasses, and not-Superman clearly has better eyesight than that.

Blaugust 2026: My Posts

Posted on
by

It’s been gloomy all day – the kind of August day that’s hot, humid and still, sapping all your energy because nothing *actually* cools you off except real air conditioning (or standing in front of the refrigerator), not even taking a cold shower in the afternoon. At least the nights are still cooling off, even if the days are muggy.

Close to sunset, the clouds finally started to break up, and I began to think I might get to see tonight’s lunar eclipse after all.

Darkening sky with a bunch of small clouds, lit from behind/below.

That’s looking west. The clouds to the east were still a bit more solid, but low to the horizon, and by the time the moon cleared the rooftops and the eclipse was well underway, the clouds had broken up enough to see!

At first glance it looks like a half-moon...but the line between light and shadow is curved the wrong way, its endpoints aren't aligned with the center, and there's none of the texture you'd see along it in a normal moon phase

I took the sunset clouds shot with my Fairphone, which did a lot better than I expected, though I only managed to get misshapen blobs when shooting the moon. It’s been hectic enough that I didn’t try to set up the good camera. The point-and-shoot Canon with real optics does okay with the moon, as you can see here. (The teen’s iPhone, however, was able to get at least a couple of shots where the maria are visible. Not as good as the Canon, but better than any of my phones have managed.)

We went back inside, waited a while, went back outside, took some more photos, talked with the neighbors who were also looking at the moon, went back inside, rinse, repeat. Lunar eclipses are a lot more accessible than solar ones. They’re slower, since the Earth’s shadow is around 3 times the size of the moon at that distance. They’re visible from anywhere with a view of the moon, not just a narrow track. You don’t need eye protection. And unlike solar eclipses, a partial lunar eclipse is almost as cool as a total one. A lot of the time you can just walk out into your front yard.

The moon hangs in the black sky above the top of a telephone pole, one wire (and something hanging from it) passing in front of it. Most of the moon is a dim red color, on which the maria are clearly visible. A bright white arc at the top looks like a shallow beanie. The brightness completely washes out any details, and makes the outer edge of the arc bulge outside of the circle.

If you’re reading this close to when I’m posting it, the eclipse isn’t over yet. There’s still an hour or so before the moon completely passes out of Earth’s shadow. If you have clear skies and the moon is up where you are, go take a look!

Blaugust 2026: My Posts

Posted on
by

Over the last few years, right-wingers have finally started to realize that what they’re missing is empathy. Several decades’ worth of “I didn’t realize how bad this policy was until it hurt someone I know personally!” statements had already made that clear to the rest of us. “I didn’t think the leopards would eat my face!” as a meme goes back a decade now. But instead of realizing they should develop some, they’ve decided empathy must be bad, much like they previously decided that being aware of systematic injustice must be bad, and turned “woke” into a pejorative.

*sigh*

No, “Western Civilization” is not doomed by “suicidal empathy.” That people like Musk and Saal think “not treating people like garbage just because they aren’t like you” is an existential threat does not make it so. As Myers says, that tells you more about them than it does about society.

And if some aspects of our culture are threatened by empathy, such as the belief that certain types of people deserve to dominate others and other types of people deserve to be abused, then we should throw out those aspects, not cling to them even harder.

There are people who exploit vulnerabilities in society. But those people act a lot more like Elon Musk and a lot less like some random non-gender-conforming person who’s just trying to use the bathroom without getting beaten by a mob for picking the “wrong” one.

If our civilization is destroying itself, it’s not the people asking for empathy and inclusion who are doing it. It’s the reactionaries who insist on “taking back” their ability to treat others like garbage without consequence. They’re the ones tearing us apart.

Blaugust 2026: My Posts

Posted on
by

Sure, I like electronic gadgets. But I dislike planned obsolescence more.

It’s one thing when the newer model really is that much better. “All About the Pentiums” exaggerated it, but computer hardware in the 1990s really was improving by leaps and bounds. (And the software was devouring those performance gains, both in demanding more resources, and being less efficient as programmers were told to optimize for their own time, not for disk space or RAM.) That eventually settled out around the time smartphones and cloud computing started taking off, which is why I was able to run a 2012 model CPU all the way through 2024. A faster PC would have been nice, but I didn’t need one yet.

As I recall the tech industry freaked out when demand for new PCs dropped. There was probably a rash of “Millennials are killing computers!” articles.

But smartphones were going through the same kind of rapid improvement. You could easily justify buying a new phone after only two years: it was a lot faster, had a better display, had a better camera and more storage, and so on. And the carriers were happy to hand you a “free” new device to keep you on their service. Heck, if you really wanted (and could afford) to buy a new phone every year, you could make a case for it.

In that environment, you could kind of accept a company only offering 2-3 years of support for their phones.

Eventually that settled out too, and it takes longer for new phones to really be worth the upgrade. Of course the industry has tried to keep the pace going. If the hardware still works, and it’s still able to run the apps people want to use, there’s no good reason for the company to stop supporting it after only three. Especially with something whose main purpose is to be constantly connected to the Internet.

I have an Android phone from 2020 and a tablet from 2019. The tablet still works fine. (It’s great for digital comics, or PDFs that are pre-formatted for 8.5×11″ pages.) The phone worked until two days ago. (Technically it still does, but I don’t want to overload the already-swollen battery.) But they’re both unsupported. Not because the hardware has failed (up until Monday). Not, like the Poke3, because they can’t handle today’s apps. Simply because Google and Samsung don’t want to provide updates anymore. Not even bugfixes. Not even security fixes.

It’s maddening.

Especially with prices on new electronics skyrocketing due to the AI datacenter-driven RAMpocalypse.

There are many things I dislike about Microsoft, but I appreciate that they finally realized that a large chunk of their customer base can’t or won’t upgrade their hardware just to install Windows 11, no matter how they try to convince us, and has extended Windows 10 support twice with minimal hoops to jump through.

The Fairphone 4 (2021) that I pressed into service to replace the dying Pixel 5 (2020) is more or less comparable in terms of speed. I still need to experiment with the camera. But it’s still supported through 2028, meaning it’s safer to use online. Which, again, is what cell phones are for.

Blaugust 2026: My Posts

Posted on
by

I had a migraine this morning that really threw me for a loop. Then I spent the afternoon and evening dealing with a different kind of headache: getting my phone number transferred to a new(ish) phone.

We picked up a Fairphone 4 a few years ago as sort of a hedge in case the electronics industry and/or tariffs and/or international trade went bonkers and one of our phones died. (We weren’t expecting all three.) The teenager used it for a bit before we bought him a phone that was closer to what he actually wanted to use. So the Fairphone returned to the back bench, so to speak.

A cell phone held in a hand, showing the California DMV website. A bright magenta line is running vertically down the entire screen, including the status bar at the top and the onscreen buttons at the bottom.Today, a bright vertical magenta line appeared on my Pixel 5’s screen. I’d noticed it was running hot lately, and when I pulled it out of the case, sure enough it had started bulging.

I know it’s old. I’ve been trying to get as much use out of it as I can, which, well, I guess I have.

(Edit: added photo)

I do not like spicy pillows, Sam I Am.

So I reset the Fairphone and started setting up the most important apps: authenticator, password manager, stuff like that. And I set about unlocking the SIM protection on my phone number and got ready to call customer service.

I knew from experience that the Fairphone 4 works fine on T-Mobile’s US network, but their apps don’t know how to activate the e-sim. So I was prepared to talk someone through.

But because we’d been swapping back and forth between phones, the sim was still registered as attached to the teen’s number, even though the number is actually pointing to his current phone. It took over two hours (at least 45 minutes of that on hold) and three levels of escalation to first figure out that’s what the problem was, and then find someone with the access to fix it.

They did finally get it all sorted out, but it was not the way I wanted to spend my afternoon.

(And when I let the old phone cool off, the screen went back to normal, though it’s still bulging.)

Edge-on view of the phone showing that the glass screen is splitting away from the plastic case.

And to top it off, Nextcloud’s photo uploading has stopped working reliably, so I have to go through the last couple months of photos and make sure I have everything copied over. I went through some of them tonight, and need to get the rest…soon.

For now, I’ve turned the old phone off completely overnight.

Once I’ve got the photos transferred, I should figure out whether it’s worth repurposing the device with a new battery or just running it without one. And then dispose of the fire hazard.

Blaugust 2026: My Posts

Posted on
by

I recently added “Your LLM was trained on my blog” to the intro on my home page. I’m sort of being facetious, but it’s also true: my old site (hyperborea.org) has been in the Common Crawl data set for years, this blog has been online since 2002, and the major LLMs have used Common Crawl’s data in their training.

Way back when I switched to WordPress in 2003, one of the things I found annoying about it was that it would auto-convert punctuation from ASCII to “fancy” characters. Curly quotes, curly apostrophes, em-dashes instead of double-hyphens, ellipses…These were things that bugged the heck out of me when I saw them in broken emails and Word documents. Over the next year, seeing the difference it made in legibility, I came to appreciate real typography. Since then I’ve been happy to let WordPress, and later ClassicPress and Eleventy convert my plain typing.

Side note: I still have trouble using some apps with code snippets (ex: Apple Notes and Stickies), ’cause I haven’t figured out how to stop them from converting ASCII punctuation that means something in the code to Unicode punctuation that means something more specific in printed human language. Also, notice in that sentence that ClassicPress was smart enough to convert a straight up and down apostrophe used on a contraction to a curly apostrophe, and not to a curly single quote like it does with ‘here.’

Anyway, WordPress powers a huge amount of the web. I’ve seen stats bandied about such as 42% of all websites, 60% of content management systems, >50% of the top 10,000 websites (by some measure or another). The exact numbers aren’t important, it’s just important to know that a whole lot of the web uses WordPress to convert what people write into HTML pages for the web.

Now let’s get back to LLMs:

  • People think they can identify LLM-generated text because it tends to include em-dashes.
  • LLMs were trained on writing by actual people on the web.
  • A whole lot of that writing by actual people is presented by WordPress.
  • WordPress has auto-converted typography including em-dashes since 2003.
  • That means a large chunk of LLMs’ training data includes text by real people that uses real em-dashes.

“AI” doesn’t use em-dashes because real people don’t, it uses em-dashes because real people do [note: lots of swearing] (via).

The only place I can imagine it would be even a semi-reliable indicator would be in social media posts, on platforms where that kind of auto-conversion doesn’t happen. Even then, I can easily imagine someone writing a draft in Word, or copying a sentence or two from their WordPress post, or just hitting option+shift+hyphen on a Mac (or whatever the key combo is on Windows).

»All pages site-wide with this tag