It’s been a while. Between 2020 and mid-2023, I wrote pretty regular “Long Links” posts, curating links to long-form pieces that I thought were good and I had time to read all of because, unlike my readers, I was lightly employed. Well, then along came my Uncle Sam gig, then fun Open Source with Topfew and Quamina, then personal turmoil, and I’ve got really a lot of browser tabs that I thought I’d share one day. That day is today.
Which is to say that some of these are pretty old. But still worth a look I think.
True North Indexed · Let’s start with Canadian stuff; how about a poem? No, really, check out Emergency Exit, by Kayla Czaga; touched me and made me smile.
Then there’s Canada Modern, from which comes the title of this section. It’s an endless scroll of 20th-century Canadian design statements. Go take it for a spin, it’s gentle wholesome stuff
Renaissance prof · uses this has had a pretty good run since 2009; I quote: “Uses This is a collection of nerdy interviews asking people from all walks of life what they use to get the job done.” Older readers may find that my own May 2010 appearance offers a nostalgic glow.
Anyhow, a recent entry covers “Robert W Gehl, Professor (Communication and Media Studies)”, and reading it fills me with envy at Prof. Gehl’s ability to get along on the most pristine free-software diet imaginable. I mean, I know the answer: I’m addicted to Adobe graphics software and to Apple’s Keynote. No, wait, I don’t give that many conference talks any more and when I do, I rely on preloaded set of browser tabs that my audience can visit and follow along.
If it weren’t for that damn photo-editing software. Anyhow, major hat-tip in Prof. Gehl’s direction. Some of you should try to be more like him. I should too.
Now for some tech culture.
Consensus · The IETF does most of the work of nailing down the design of the Internet in sufficient detail that programmers can read the design docs and write code that interoperates. It’s all done without voting, by consensus. Consensus, you say? What does that mean? Mark Nottingham (Mnot for short) has the details. Consensus in Internet Standards doesn’t limit its discussion to the IETF. You probably don’t need to know this unless you’re planning to join a standards committee (in which case you really do) but I think many people would be interested in how Internet-standards morlocks work.
More Mnot · Check out his Centralization, Decentralization, and Internet Standards The Internet’s design is radically decentralized. Contemporary late-capitalist business structures are inherently centralized. I know which I prefer. But the tension won’t go away, and Mnot goes way deep on the nature of the problem and what we might be able to do it.
For what it’s worth, I think “The Fediverse” is a good answer to several of Mnot’s questions.
More IETF · From last year, Reflections on Ten Years Past the Snowden Revelations is a solid piece of work. Ed Snowden changed the Internet, made it safer for everyone, by giving us a picture of what adversaries did and knew. It took a lot of work. I hope Snowden gets to come home someday.
Polling Palestinians · We hear lots of stern-toned denunciations of the Middle East’s murderers — both flavors, Zionist and Palestinian — and quite a variety of voices from inside Israel. But the only Palestinians who get quoted are officials from Hamas or the PLA; neither organization has earned the privilege of your attention. So why not go out and use modern polling methodology to find out what actual Palestinians think? The project got a write-up in the New Yorker: What It Takes to Give Palestinians a Voice. And then here’s the actual poll, conducted by the “Palestinian Center for Policy and Survey Research”, of which I know nothing. Raw random data about one of the world’s hardest problems.
Music rage · Like music? Feel like a blast of pure white-hot cleansing rage? Got what you need: Same Old Song: Private Equity Is Destroying Our Music Ecosystem. I mean, stories whose titles begin “Private equity is destroying…” are getting into “There was a Tuesday in last week” territory. But this one hit me particularly hard. I mean, take the ship up and nuke the site from orbit. It’s the only way to be sure.
Movies too · Existentially threatened by late capitalism, I mean. Hollywood’s Slo-Mo Self-Sabotage has the organizational details about how the biz is eating its seed corn in the name of “efficiency”.
I’m increasingly convinced that the whole notion of streaming is irremediably broken; these articles speak to the specifics and if they’re right, we may get to try out new approaches after the streamers self-immolate.
A target for luck · I’ve mostly not been a fan of Paul Graham. Like many, I was impressed by his early essays, then saddened as he veered into a conventional right-wing flavor that was reactionary, boring, and mostly wrong. So these days, I hesitate to recommend his writing. Having said that, here’s an outtake from How To Do Great Work:
When you read biographies of people who've done great work, it's remarkable how much luck is involved. They discover what to work on as a result of a chance meeting, or by reading a book they happen to pick up. So you need to make yourself a big target for luck, and the way to do that is to be curious. Try lots of things, meet lots of people, read lots of books, ask lots of questions.
Amen. And the humility — recognition that good outcomes need more than brains and energy — is not exactly typical of the Bay-Aryan elite, and is welcome. And there’s other thought-provoking stuff in there too, but the tone will put many off; the wisdom is dispensed with an entire absence of humility, or really any supporting evidence. And that title is a little cringey. Could have been shorter, too.
“readable, writerly web layouts” · Jeffrey Zeldman asks who will design them. It’s mostly a list of links to plausible candidates for that design role. Year-old links, now, too. But still worth grazing on if you care about this stuff, which most of us probably should.
Speaking of which, consider heather buchel’s Just normal web things. I suspect that basically 100% of the people who find their way here will be muttering FUCK YEAH! at every paragraph.
Enshittification stanzas · (My oldest tabs, I think.) I’m talking about Ellis Hamburger’s Social media is doomed to die and Cat Valente’s Stop Talking to Each Other and Start Buying Things: Three Decades of Survival in the Desert of Social Media, say many of the same things that Cory is. But with more personal from-the-inside flavor. And not without streaks of optimism.
Billionaires · It’s amazing how fast this word has become shorthand for the problem that an increasing number of people believe is at the center of the most important social pathologies: The absurd level of inequality that has has grown tumorously under modern capitalism. American billionaires are a policy failure doesn’t really focus on the injustice, but rather does the numbers, presenting a compelling argument that a society having billionaires yields little to no benefit to that society, and precious little to the billionaires. It’s sobering, enlightening, stuff.
Gotta talk about AI I guess · The “T” in GPT stands for “Transformation”. From Was Linguistic A.I. Created by Accident? comes this quote:
It’s fitting that the architecture outlined in “Attention Is All You Need” is called the transformer only because Uszkoreit liked the sound of that word. (“I never really understood the name,” Gomez told me. “It sounds cool, though.”)
Which is to say, this piece casts an interesting sidelight on the LLM origin story, starting in the spring of 2017. If you’ve put any study into the field this probably won’t teach you anything you don’t know. But I knew relatively little of this early history.
Visual falsehood ·
Everyone who’s taken a serious look at the intersection of AI and photography offered by the Pixel 9 has reacted
intensely. The terms applied have ranged from “cool” to “terrifying”. I particularly like Sarah Jeong’s
No one’s ready for this,
from which a few soundbites:
“These photographs are extraordinarily convincing, and they are all extremely fucking fake.”
“…the easiest, breeziest user interface for top-tier lies…”
“…the default assumption about a photo is about to become that it’s faked…”
“A photo, in this world, stops being a supplement to fallible human recollection, but instead a mirror of it.”
We are fucked.
And that’s just the words; the picture accompanying the article are a stomach-churning stanza of visual lies.
Fortunately, I’m not convinced we’re fucked. But Google needs to get its shit together and force this AI voodoo to leave tracks, be transparent, disclose what it’s doing. We’re starting to have the tools, in particular a thing called C2PA on which I’ve had plenty to say.
Specifically, what Google needs to do is, when someone applies an AI technique to produce an image of something that didn’t happen, write a notification that this is the case into the picture’s EXIF and include that in the C2PA-signed manifest. And help create a culture where anything that doesn’t have a verifiable C2PA-signed provenance trail should be presumed a lie and neither forwarded nor reposted nor otherwise allowed to continue on its lying path.
Fade out · Here’s some beautifully performed and recorded music that has melody and integrity and grace: The Raconteurs feat. Ricky Skaggs and Ashley Monroe - Old Enough.
I wish things were a little less hectic. Because I miss having the time for Long Links.
Let’s all do the best we can with what we have.
Comment feed for ongoing:
From: Geoff Arnold (Sep 09 2024, at 08:49)
Apropos of AI images, I just posted the following on FB:
We know it is possible for a browser plugin to inspect all of the images on a web page. (Pinterest, for example.) Imagine a browser plugin which could inspect the metadata on each image and tag all AI-generated images. Ideally this would use robust, secure metadata, but while we're still waiting for that I'd settle for a heuristic or crowd-sourced solution.
[link]
From: Walter Underwood (Nov 04 2024, at 15:20)
I've only been on one IETF standards committee, but that was very well run (by you), the Atom Publishing Protocol.
I still remember the very long discussion about what should cause the published datetime to be updated on an Atom entry. We finally decided that the definition of a meaningful change belonged to the publisher, not the spec.
The experience was really useful throughout the rest of my career. It is really important to sort through what was necessary, wanted, forbidden, or just not relevant in a spec. As a search engineer, a lot of my time was spent connecting to new kinds of documents.
Atom has been quietly, wildly successful. At LexisNexis, it is the common search result format across 100+ different document collections. HighWire Press uses it extensively internally. I don't even mind that people pronounce Atom as "RSS".
[link]