Coyote Tracks

While I don’t think Apple Podcasts itself has a “trim silence” feature, it’s popular in a bunch of podcast players. Overcast probably kicked it off with its “Smart Speed” function, but it’s in Pocket Casts, Castro, and I’m sure a bunch of other ones. The idea is that the player “listens” for silences in the podcast and (hopefully subtly) shortens them, so you’re saving a little time even when you’re playing the podcast at normal speed. I suspect many users always run with Trim Silence enabled. I did, until I realized that occasionally Overcast interpreted very, very soft sounds as trimmable, subtly screwing up the timing of music. No problem: you can disable it for specific podcasts, right? Well, if you do that, you’ll learn that podcasts that I think of as “NPR style,” e.g., ones with a certain kind of production value like 99% Invisible and Pop Culture Happy Hour, actually sound pretty good without Trim Silence enabled. In fact, they sometimes sound better. They’re edited precisely—even though PCHH itself is, despite being an actual NPR production, very much the “several dudes talking” kind of show unique to the podcast form. (With the asterisk that PCHH’s rotating panel is extremely diverse across just about any demographic line you care to name.) Now, this isn’t a unique observation; I remember chatter a year or two ago on social media that I follow from people discovering this (and making dramatic denunciations of the option). But what I didn’t expect is what happened when I listened to podcasts from folks who are part of what one might cheekily dub the 5by5 Diaspora without trim silence enabled. (Brief aside: 5by5 was a network a decade and change ago run by Dan Benjamin, with shows featuring hosts that pod-listening Apple nerds know: Merlin Mann, John Siracusa, Myke Hurley, Marco Arment, John Gruber, and more. It was huge for a brief time, then fell apart in just as brief a time, with next to all its hosts either jumping to the nascent Relay or going independent. Someday I’d love to know the story of what happened there; reading between the lines, it sounds as if Benjamin was either difficult to work with, didn’t give hosts generous business deals, or both.) My impression of many of those conversational shows was that even though they were relatively lightly edited, they were still edited tightly enough to take out awkward pauses. But listening to a few shows that I otherwise love without the Smart Speed or Trim Silence function at work, though, makes me question that belief. Are their creators are, consciously or not, assuming that everyone’s listening with Trim Silence enabled? This could be a style choice, to be sure, the idea that the conversation’s going to sound more natural if you keep in the two or three seconds while a host gathers their thoughts between sentences or phrases. In a sense, it does. But I don’t think it’s the kind of “natural” you actually want in a podcast any more than you want it in a scripted show, where any long silence is deliberately written into the script. When I listen to these podcasts as recorded, I think: this could have used a little more editing. Is this bad? No, not in the grand scheme of things. I am re-evaluating my podcast listens, in part because I don’t have the free time to keep up any more, and in part because I’m burning out on tech news in general and on AI AI AI AI AI AI in particular. Even so, I’m not going to stop listening to podcasts that I still like because of too-long pauses; I’ll just keep Trim Silence enabled for them. To support my writing, consider a tip on Ko-fi.com. (originally published at Coyote Tracks)

It’s a match made in…frankly, I’m not sure where. But I’ve been experimenting with using org-journal as a replacement for Day One, and while it’s pretty good, there are two things that I immediately missed: The way Day One records my location and the weather for every entry. Being able to quickly insert photos that I’ve taken into an entry. Emacs can’t do that directly, but there are Shortcut actions that do everything I need there, right? So there’s got to be a way to get Emacs to talk to Shortcuts. Well, there is a Shortcuts package for Emacs, but a bit of poking at it revealed that it seems to be just one way communication: it lets Emacs fire off a Shortcut, but doesn’t let the Shortcut return anything back to Emacs. Not so useful. But you know what? There’s a shortcuts command line utility which can not only run a Shortcut, but get results returned to it. As an example of how to use it, here’s the photo picker Shortcut I created (click on it for a bigger image): The Photo Picker Shortcut The first action lets you select photos. Then, it resizes each one and saves them as JPEG files to ~Documents/journal/images/. Lastly—and this is the part that needs to get back to Emacs—it turns each filename into an Org-style link and outputs it. So, what does Emacs do to call this (and receive it?) It is an extremely simple bit of Lisp code: (defun wm-journal-photos () "Select photos, export, and insert as org links." (interactive) (insert (shell-command-to-string "shortcuts run \"Photos to Org link\" | cat"))) The shell-command-to-string function does what it says: runs a command and gets its output as a string. That string just gets inserted at point. The location and weather Shortcut is even simpler: it just calls the “Get weather at Current Location” action, then creates a text block with LOCATION and WEATHER as Org-mode properties. (Why, you might ask, does shortcuts run have to be piped through cat to work? You got me. It’s not an Emacs quirk, for the record, it’s an Apple quirk.) To support my writing, consider a tip on Ko-fi.com. (originally published at Coyote Tracks)

Did you know em dashes are a sign of AI writing? Since when? Well, that’s what all the LinkedIn influencers say! Humans don’t use them in writing, right? They use hyphens. Of course we use em dashes—they’re basic punctuation marks! We use em dashes as a pause for emphasis—like I did in the first sentence in this paragraph—or to set off a quasi-parenthetical phrase, like I just did in this sentence. A hyphen is used to form compound words, or to break words by syllable at the end of lines. They’re not the same. Nobody better tell the influencers about en dashes; they’ll spontaneously combust. So where’d this idea come from? You got me, but I’d guess it stems from the nature of AI-generated text. It always shows a certain level of polish: no spelling or grammatical mistakes, complete sentences, a clear organization and structure. LLMs get this right in their “first drafts” because the majority of text they were trained on is itself polished—and it turns out that this level of polish covers the correct use of em dashes. Is there a better way to detect AI-generated text than looking for em dashes, then, smart guy? All generative AI is, by literal definition, statistically median, and this isn’t something that can be overcome by creative prompt engineering. Flair and personality in human writing comes from individualistic, quirky choices born of a writer’s likes, dislikes, and life experiences. LLMs don’t have any of those. So LLM-generated text tends to read like pleasantly anodyne business memos, or book reports by a precocious ninth-grader. So if the text you’re playing “robot or not” with is a business memo, a high school paper, or a blog post by the kind of LinkedIn influencer who thinks em dashes are a sign of AI writing, it might be a challenge. In other circumstances, it shouldn’t be. But it apparently is! The problem—and, dear reader, there is no gentle way to put this—a lot of ostensibly smart people neither write particularly well, nor read enough to get a sense of what good writing is like. Most of the writing they deal with—memos, emails, text messages, Amazon and Yelp reviews—is slapdash, unpolished. So if they come across anything in those contexts that looks refined yet relatively generic, they suspect ChatGPT. Apparently, we’ve so come to expect a lack of writing style from humans—and I mean style in the sense of Strunk & White, not in the sense of actual style—that we’re suspicious of it when we see it. If you don’t want your writing to be mistaken for AI, what should you do? Use “fuck” more often. To support my writing, consider a tip on Ko-fi.com. (originally published at Coyote Tracks)

You can also follow me on Mastodon or Bluesky. (I mean, you could follow me on both, but since I mostly post the same things to both it’d be repetitive.)

It’s so easy, watching on the news, to scream at people who don’t evacuate ahead of oncoming storms. I get it. I scream at them, too. It’s often a fitting reaction. And yet. Evacuation is a process. It’s not “throw a handful of things in an overnight bag and floor it.” You still need to put in the time, money, and effort of preparing your house (or condo, or apartment) for the storm, to maximize the chance of having anything to come back to. You need to have somewhere to go: if you don’t have a friend somewhere out of harm’s way you can stay with, the farther you’re going to have to drive to find a hotel room or an open shelter, and the worse the traffic jam will be getting there. You can solve those problems if you evacuate earlier—before the mandatory orders come down—but only if you have the resources to do so. As I’m typing this, on a beautiful-looking Tuesday morning with a partly cloudy blue sky outside my office window, I’m more or less in the path of Hurricane Milton. If it sticks with projections, it’ll pass within 60 miles of here. Over the past day or so, I’ve watched many people who need to panic do so—and many people who maybe don’t need to panic do so, too, often because people are shouting at them on social media to panic. I don’t doubt this is well-intentioned, but I doubt it’s helpful. There’s a straightforward metric to use to decide if you should evacuate: If you are in an evacuation zone that’s been told to evacuate, evacuate. If you aren’t, don’t. Sure, there are exceptions—for instance, if you need power for medical devices, get to a special needs shelter. But evacuating when you don’t need to is at best unnecessary and at worst makes things fractionally worse for people who do have to evacuate by putting one more vehicle on the road and occupying one more hotel room. I’ll be surprised if we don’t lose power and wired internet by Thursday morning, but despite being in the “Greater Tampa Bay Area,” I’m over 30 miles inland on relatively high ground. (“High” is always relative in Florida, of course, but we’re 70 feet over sea level.) There are flooding concerns if the conditions are right, because they’re inescapable here—the state is full of lakes and rivers—but for now they’re not likely. See you on the other side. Footnote: wildfire threat in California vs. hurricane threat in Florida I feel—tentatively—like it’s easier to provision a home (or subdivision) with fire breaks than it is to protect against storm surges, hurricane force winds, and floods. Beyond that, this makes three out of three years I’ve been back in Florida with the area I lived in under the threat of a major hurricane, while there were zero out of twenty years in California with the area I lived in under the threat of burning down. Also, the last two years have taught me to stop treating “it’s okay, it’s a dry heat” as a joke. I’d take a Sacramento August over a Florida one in a heartbeat. (originally published at Coyote Tracks)

On LinkedIn, the social network of mandatory happiness, software engineer Michael Sayman writes about his new product: Introducing SocialAI, a private social network where you receive millions of AI-generated comments offering feedback, advice & reflections on each post you make. I will admit my first thought about this was: this is a prank, right? This is a fake product that somebody who hates generative AI and the “techbro culture” would come up with to mock it. My second thought was: oh no, it’s real. SocialAI is more than just another project for me – it’s the culmination of everything I’ve been thinking about, obsessing over, and dreaming of for years. I’ve always wanted to create something that not only showcases what’s possible with tech but also helps people in a real, tangible way. SocialAI is designed to help people feel heard, and to give them a space for reflection, support, and feedback that acts like a close-knit community. This app is a little piece of me – my frustrations, my ambitions, my hopes, and everything I believe in. It’s a response to all those times I’ve felt isolated, or like I needed a sounding board but didn’t have one. I know this app won’t solve all of life’s problems, but I hope it can be a small tool for others to reflect, to grow, and to feel seen. And my next thought was—forgive me, but does Mr. Sayman know how genuinely sad an application that creates artificial parasocial relationships for you sounds? I don’t love the phrase “go touch some grass,” but Jesus Hasenpfeffer Christ, if there’s any sign you need to delete Visual Studio Code from your laptop and go take up goat yoga, isn’t this it? As far as I can tell, his intention here is sincere; his new company, “Friendly Apps,” has a mission “to foster connections that prioritize mental health and holistic well-being”. But if you’re thinking “the only kind of person who would think a flood of AI-generated ‘support and feedback’ would be therapeutic would be, I don’t know, a Gen-Z product manager at Facebook or Google,” I have unsurprising news for you. To be sure, Sayman is a prodigy, joining Facebook when he was 18 after already published successful iOS apps. His priority there, though, was helping them go after teenagers, developing an ill-fated Snapchat competitor that lasted a year. One could make the blunt but not unfair case that Facebook/Instagram has contributed more to the need for therapy than any other company, and that perhaps one should have thought of that before, you know, becoming a product manager for them. Maybe I’m being cynical. Maybe SocialAI will be fruitful in combating loneliness and isolation. But maybe you should hang out with a non-robotic friend or two, whether online or off, instead. (originally published at Coyote Tracks)

It’s as much of an age-old question as anything of the internet age is, but I think about it every few years. Slapping up a web page full of personal thoughts in 1999 didn’t carry the same weight that it does in 2024; employers, even ones literally in the internet space, didn’t necessarily do a web search on your name to see what came up. Now, though, I don’t know. Now everyone does that kind of search. And, as someone who’s been conducting an unsuccessful, albeit woefully low-key, job hunt in 2024, it’s hard not to ask if what I have online at this point is working against me with more skittish employers. Even before I wrote about conservatives being snowflakes, it wasn’t exactly hard to suss out my left-wing politics, and I’ve been vocally skeptical of generative AI. In a moment where tech companies madly chase AI in either a sincere belief it’s going to change everything or mere FOMO that it might, vocal AI skepticism might be a career-limiting move—and in a moment where alt-right techbros cow companies into walking back DEI efforts, merely noting my pronouns at all, let alone noting them as “he/they”, might also be one. (“What kind of woke beta cuck uses pronouns?” Anyone who speaks English, you stupid motherfucker.) (Also, cursing in my blog might be a career-limiting move.) And yet, I don’t know. Do I want to frantically dash around the internet, tearing down anything that might give a prospective employer the heebie-jeebies about me? Is that even possible? While I’ve never reached “internet famous” status, I’ve posted lots of stuff in lots of places. When I was more actively doing tech blogging at Coyote Tracks, it got linked to by other blogs often enough that I used to joke that while you’re probably not reading me, somebody you’re reading probably is. There is an alternate universe in which I figured out how to monetize that. And, well, while I don’t advertise being a furry, I’ve been writing furry fiction for decades and have been guest of honor at more than one furry con. One does not have to possess mad internet sleuthing skills to put two and two together. Frankly, I’m always amazed when I find people I grew up with, especially members of the Original Internet Nerd cohort, have next to no footprint online. The one whose LinkedIn includes “Futurist at the Center for 21st Century Teaching Excellence” you’d kind of expect to have, I don’t know, a low-effort Substack, if not a YouTube channel, right? If you can tell someone your job title is “futurist” and keep a straight face, I’m pretty sure they’re legally required to give you a TED Talk. But apart from that LinkedIn, there’s basically nothing out there about them, and that’s true for the majority of other folks I knew in that place at that time. Does that help with job hunts? Maybe. On the other hand, maybe it hurts. It’s my understanding the former futurist’s LinkedIn is long out of date, and they struggled to find work for years before moving into non-tech fields. I can’t say that actually being, you know, present on the internet might have helped if they’d wanted to stay in tech, but I wonder. I made the shift to tech writing largely because of my loudmouth tech blog, when Joey Zwicker at RethinkDB came across it and thought, “we should talk to them about our new tech writing position.” Of course, I don’t know that they did want to stay in tech. If they didn’t, I couldn’t blame them. In any case, I don’t think erasing my digital footprint is either possible or worth it. Anyone who reads what I write and decides they can’t work with someone like me is probably right. (originally published at Coyote Tracks)

There’s been plenty of words written already explaining why calling Trump, J. D. Vance, and the rest of the MAGA branch of the Republican Party—which is to say, the Republican Party—weirdos seems so effective. These conservatives have spent years, so the thought goes, asserting their worldview is the standard, the default, the normal worldview, and everything that deviates from it is wrong, deviant, and yes, weird. Calling them weird flips that on its head. While that’s undoubtedly true, I think the real reason it’s effective is that conservatives are—and have been for years—snowflakes. This isn’t new to the MAGA wing; it was the real animating factor of the Tea Party, not economics. (It’s never really been about economics; if it was, we’d be having some real engagement over why, if Republicans are better at running the economy, the economy has, for the last seventy years or so, consistently done better under Democratic administrations.) “But,” someone might say, “surely liberals are the real snowflakes, the ones who want their feelings to be protected from facts?” To which I say: no, imagined interlocutor, that’s transparent bullshit. I mean, come on. Is it liberals and leftists who’ve been screaming for years that merely seeing queer people in television shows and movies more often now is literal warfare? No, it is not. Have you met anyone truly offended by seeing Christmas decorations or having a cashier say “Merry Christmas?” No, but you’ve met dozens who claim to get the vapors when a cashier, in a store full of Christmas decor, says “Happy Holidays”. I had a conservative (ex-)friend strenuously argue that yes, in fact, a state giving gay people the right to marry somehow devalued his good conservative Christian marriage. His arguments were terrible, because all the arguments for that position are terrible. No matter what kind of intellectual varnish one tries to give it, the heart is the same as every other culture war argument: No one should be allowed to make my family and I aware of things I don’t personally approve of. Books show queer people in a positive light? Get ’em out of the libraries, because kids might read ’em, and the mere knowledge that gay, lesbian, and transgender people exist is corrupting. (Everyone knows there were no gay or trans people before Bill Clinton was elected.) Boycott stores that put up rainbow flags, or hire women and blacks. (Sure, you’re not against hiring minorities, you’re just against visibly making an effort to do so. Gotcha.) Don’t teach our children history that could conceivably show their ancestors in a bad light—that might make them slightly uncomfortable. What kind of monster makes children engage with difficult topics? The anti-MAGA narrative has always had an undercurrent of these people would be laughable if they weren’t so dangerous. But the “weird” argument sticks because it acknowledges one can be both dangerous and laughable. Any guy irresponsibly blasting away with an assault rifle is dangerous. One irresponsibly blasting away with an assault rifle at cases of Bud Light because he was made aware that the beer company sells to trans people, too? I mean, come on. How do you get more snowflake than that? Actually, forget I asked that. Based on MAGA reactions to the Harris campaign so far, we’re gonna have some real blizzards. Buckle up. (originally published at Coyote Tracks)

Bluesky (@coyotetracks.org)

Winelight