Month: November 2024

Potential Algorithmic Bias on X During the 2024 U.S. Election ⇥ eprints.qut.edu.au

Timothy Graham and Mark Andrejevic:

This technical report presents findings from a two-phase analysis investigating potential algorithmic bias in engagement metrics on X (formerly Twitter) by examining Elon Musk’s account against a group of prominent users and subsequently comparing Republican-leaning versus Democrat-leaning accounts. The analysis reveals a structural engagement shift around mid-July 2024, suggesting platform-level changes that influenced engagement metrics for all accounts under examination. The date at which the structural break (spike) in engagement occurs coincides with Elon Musk’s formal endorsement of Donald Trump on 13th July 2024.

While this is presented in academic paper format, you should know that it is still an unpublished, non-peer-reviewed working paper. Its methodology involves just ten X accounts and, as the authors note, their analysis is limited due to the site’s opacity for researchers. Also, the authors do not once mention the assassination attempt that led to Musk’s endorsement on the very same day — a conspicuous absence, I think. None of this means it is inherently inaccurate. It does mean you should hold onto these findings very, very loosely.

It is worth reading, though, because even if I do not entirely trust its findings, it is still compelling (PDF). I am not sure what criteria were used to select the ten accounts in question, but the five Democrat-aligned accounts are all either lawmakers or political leaders in some way. The five Republican-aligned accounts, on the other hand, are all commentators and also Donald Trump Jr., and I am not sure that is a reasonable comparison. Surely it would be better to compare like-to-like.

Even so, it sure appears the date of Musk’s endorsement matches the timing of a change in political activity on X. One possibility is for the assassination attempt and endorsement to have caused more activity on his platform, and specifically among those who do not find its owner to be an odious buffoon. However, a more cynical possibility suggested by this research is of the platform taking sides, despite its new owner promising neutrality. Theoretically, we can check this for ourselves. In the name of “full transparency”, X published “the algorithm” on GitHub; indeed, it appears it was updated around the same time as these researchers found this partisan boost. But there is not a corresponding public commit — no public commits, in fact, since July 2023, as of writing — so it is impossible to know if this is related or just someone fixing a typo. “Transparency” does not work when it depends on unreliable actors.

Also, if the work of these researchers represents a true shift, I believe it will be the first time fears of an explicitly partisan influence on algorithmic recommendations have been demonstrated in the United States. Meta has avoided suggesting posts it deems political in nature — probably because they are more difficult to moderate, and partly because it is beneficial for Meta to ingratiate itself with the incoming administration. TikTok, despite public fears, has no demonstrated partisan political influence.

But X? Its users and ownership have carved out a space for explicit discrimination and — possibly — partisan bias.

A Popular A.I. Training Set Includes Data from OpenSubtitles ⇥ theatlantic.com

Alex Reisner, the Atlantic:

I can now say with absolute confidence that many AI systems have been trained on TV and film writers’ work. […]

The files within this data set are not scripts, exactly. Rather, they are subtitles taken from a website called OpenSubtitles.org. Users of the site typically extract subtitles from DVDs, Blu-ray discs, and internet streams using optical-character-recognition (OCR) software. Then they upload the results to OpenSubtitles.org, which now hosts more than 9 million subtitle files in more than 100 languages and dialects. […]

The Atlantic has built a search engine of subtitles used in training. This is in addition to — but in the same data set as — YouTube subtitles.

The files provided by websites like OpenSubtitles are, to my knowledge, not exactly legal. Courts in Australia and the Netherlands have treated them as distinct works protected by copyright. I am not arguing this is correct — fan-created subtitles are useful and can permit more translation options — but it is noteworthy for these models to be trained not only on original works without explicit permission, but also on derivative works made illegally.

Put it this way: would it be right if models used for generating movies were trained on a corpus of pirated movies, or music to be trained on someone’s LimeWire collection? It arguably does not matter whether copyright holders were paid for the single copy used in training materials, since it is a derivative created without permission in either case. But it feels a tiny bit worse to know generative models were trained using illicit subtitles instead of quasi-legitimate ones.

What Is a Skeptic? ⇥ defector.com

Albert Burneko, Defector:

What is skepticism? In my lifetime as a word-nerd, I have known “skepticism” to refer to a sort of stubborn insistence upon rigor and evidence in place of things like dogma and “common sense.” A skeptic, by those terms, is someone who questions what they are told. Crucially, a skeptic actually questions, as in seeks answers. A person who merely refuses to learn what can be known is not a skeptic, but rather an ignoramus; a person who raises questions but does not seek their answers is not a skeptic, but a bullshitter. A person who rejects empirical knowledge, who refuses the answers that exist while requesting ones more to their liking that flatter their preference for unfounded contrarian gibberish and conspiratorial paranoia, is not a skeptic. They’re the exact opposite of that: a mark. A sucker. A credulous boob.

There is nothing wrong with asking questions, even about well-founded and understood phenomena — but the asker must be honestly willing to accept answers, not using questions as a sly means of discrediting actual knowledge and expertise.

My one complaint with Burneko’s piece can be found in this sentence a little bit later:

[…] At any rate, “vaccine skeptic” certainly is nicer and less contentious than calling Kennedy a motivated bullshitter, a peddler of antiscientific garbage, the type of dogshit-brained imbecile who will stiff-arm all that can be learned from centuries of medical research and practice because he preferred what he learned from a 25-second TikTok video made by a spiral-eyed homeschool casualty who’ll be hospitalized next month with an illness that hasn’t sickened a human being since the Bronze Age. […]

To be clear, it is not this sentence itself I have a problem with. This is phenomenal. Burneko’s essays are among my favourite things I read in any given year.

No, the problem I have is that this sentence sits in the middle of a paragraph about how the New York Times uses words like “skeptic” to launder and, by extension, validate the unhinged claims of people like Robert F. Kennedy Jr., but it dismisses his knowledge as coming from a “25-second TikTok video”. That is not where he is learning these things. What is offensive is that Kennedy’s view of health and disease comes from a mix of a specific ecosystem pushing these claims, and mainstream media outlets like the Times giving them a modicum of credibility.

It is not just Kennedy; a 2011 article in the Times about Andrew Wakefield is front-loaded with descriptions of the “controversial figure” and his “concerns about the safety of the measles-mumps-rubella vaccine”. It takes a couple thousand more words to get to Brian Deer’s brutal debunking, but it is only afforded a small chunk of the article, with the author writing “it would take a book to encompass” the back-and-forth between Deer and Wakefield. To a casual reader, this article would feel like a cautious middle-ground approach even though, at the time it was published, Deer had already released the scathing results of his investigation in a series of Sunday Times articles and a documentary. Luckily, a book is now also available.

Voice-from-nowhere journalism may not be solely responsible for the beliefs of people like Kennedy, but it validates those views all the same. Media should be skeptical, in that it ought to continuously ask questions and articulate the answers and the evidence.

Canadian Right-to-Repair Legislation Receives Royal Assent ⇥ theregister.com

Brandon Vigliarolo, the Register:

Royal assent was granted to two right to repair bills last week that amend Canada’s Copyright Act to allow the circumvention of technological protection measures (TPMs) if this is done for the purposes of “maintaining or repairing a product, including any related diagnosing,” and “to make the program or a device in which it is embedded interoperable with any other computer program, device or component.”

Elizabeth Chamberlain, iFixIt:

There’s one major limitation that Canada shares with the US: neither country allows for the trafficking of repair tools. While Canadians can now legally bypass TPMs to fix their own devices, they can’t legally sell or share tools designed for that purpose. This means Canadian consumers and repair pros still face technical and legal hurdles to access the necessary repair tools, much like in the US.

This win for Canadians is still huge — it’s the first time federal law anywhere has tackled digital locks in favor of repair. But the restriction on tools limits who can benefit, which is why the repair fight continues.

This legislation has been a long time coming. I thought I had written about C–244 earlier this year, but it turns out it was last November. Still, progress, and with unanimous agreement.

Slop on Spotify ⇥ theverge.com

Elizabeth Lopatto, the Verge:

[…] But so had a lot of artists with single-word names, such as Swans, Asia, Standards, and Gong. A new album would appear on an artist’s Spotify page, bearing their name but no similarity to their music. Sometimes, as with the fake HEALTH albums, they would disappear after a few days. Other times, they would linger indefinitely, even against the artist’s will.

[…]

It looks like Standards, Annie, HEALTH, Swans, and a number of other notable one-word artists were targeted directly. Spotify confirmed that the onslaught of AI garbage was delivered from one source, the licensor Ameritz Music. Ameritz Music did not respond to a request for comment.

The great ensloppening of the internet continues — except, in the case of Spotify, this is a repeat problem.

The Onion Buys InfoWars ⇥ theonion.com

Ben Collins today promised “the funniest news you’ve ever heard in your entire life” and, boy, did he deliver.

“Bryce P. Tetraeder”, “CEO” of Global Tetrahedron, as published the Onion:

Founded in 1999 on the heels of the Satanic “panic” and growing steadily ever since, InfoWars has distinguished itself as an invaluable tool for brainwashing and controlling the masses. With a shrewd mix of delusional paranoia and dubious anti-aging nutrition hacks, they strive to make life both scarier and longer for everyone, a commendable goal. They are a true unicorn, capable of simultaneously inspiring public support for billionaires and stoking outrage at an inept federal state that can assassinate JFK but can’t even put a man on the Moon.

The CEO may be fake, but this is real: the Onion bought InfoWars with the assistance of the families of Sandy Hook victims. The relaunched site will be supported by Everytown for Gun Safety. What a perfect, full-circle kind of outcome to dilute the influence of one of the worst figures in media.

Congratulations to Collins on being the proud owner of InfoWars’ assets, legally speaking. Given the hosts’ predilection for heavy drinking and indoor smoking, I bet the studio reeks.

Update: Some no-fun judge might be a real jackass about this whole thing and do the second-least-funny thing this year.

Misguided Apple Intelligence Ads ⇥ tidbits.com

Adam Engst, TidBits:

In the first ad, Apple Intelligence enables a goof-off who wastes time and annoys his colleagues to surprise his boss with an unexpectedly well-written email. It’s not clear that the boss is impressed; he just can’t believe the guy would have written a professional message.

[…]

The second ad channels a similar suggestion — that Apple Intelligence is a crutch for the thoughtless. […]

Michael Tsai:

It’s really quite a different message than a bicycle for the mind.

These ads come across either as unimaginative as the people they represent, or as a Freudian slip, depending on your perspective.

The first is a little better than the second because it at least hints at something I bet many of us dread: writing work email. But why not a version which elevates someone who cares? The armchair director in me wants this to be an employee who is clearly trying hard, writing a frustrated email to someone who is not, and needing to adjust the tone of a pretty mean email.

The second ad is beyond helping. If someone had handed me their own phone with a photo slideshow at any point in the past five years, I would have assumed they did not make it themselves. I do not know anybody in real life who has ever done so.

I know there are many A.I. skeptics out there — those who think the whole thing is a bust. But even if that describes you, try setting that aside and put on your best marketing smile: even you can probably imagine a handful of ways to show features like these in ways that do not make people look lazy or forgetful. How about someone struggling to find the words for something, using Writing Tools for inspiration, and then making edits to fit their personality? Or someone searching through their photo library with vague terms for a specific picture — say, a special dinner with a particular dish they want to make again? Or someone finding memories of an apartment they are leaving as they move to another city? I am sure someone on the marketing team pitched ideas like these and they were shot down for one reason or another, but they all feel more palatable to me than what I see here.

Update: I live my life by the adage never read the comments but, in this case, it would have been useful. “Joe Mac User” on TidBits points to two other ads, one of which is pretty similar to my thoughts of how to improve the first of the ads Engst linked to. Maybe that makes me biased, but it is easily the least inappropriate of these four.

Elon Musk-Themed A.I. Slop ⇥ 404media.co

Jason Koebler, 404 Media:

I have reported on AI-generated spam for a year now, and have watched as different trends come and go. In August, when I wrote about this community of people, bizarre Jesus content, surreal landscapes and dream homes, and birthday celebration posts were performing very well on Facebook. At the time, one Facebook AI spammer told me that they intended to begin spamming Facebook with “American news.” It is clear that pro-Elon Musk inspiration porn is the new strategy, or “meta” for these spammers, and creating AI spam that specifically targets people in the United States is part of the new strategy. The YouTube page for one of the Indian influencers who teaches people how to do this is full of videos for “USA CHANNELS” and US-focused spam. This strategy is clearly trickling down to Facebook at a large scale that is impossible to quantify or systematically study because Meta has killed CrowdTangle, a research tool that showed how content spread on the platform.

It seems to me this would not be nearly so popular if not for two phenomena: the transition of Facebook into a recommendations-focused product, and the idolization of tycoons. Yes, the invention of semi-realistic image generators is a necessary component, but the slop would not be nearly as successful if it did not have those two factors. Facebook is now a mix of the stuff people signed up for — like pictures from their friends, and group discussions with people in their neighbourhood — and suggestions of all kinds of crap Facebook thinks they might want to see. Also, ads, which are functionally similar in that they are things people did not ask for and must endure in order to see the things they care about.

Facebook promotes posts based in part on the activity they generate — comments, likes, and views — which is unsurprising. These things can be indicators of a noteworthy post. But Goodhart’s law suggests signals like these can become targets, therefore making them useless as metrics. Even so, they are used by Facebook to pollute users’ feeds with unrelated posts and juke engagement statistics. I do not think these posts would be nearly as widespread if they were not recommended to users. If people had to seek them out, the odd one would be shared by one of your more gullible friends.

Koebler says these posts are likely being propelled into users’ feeds by other real people, not automated traffic. That also suggests to me some level of CEO-as-celebrity idolization. This does not work for every business magnate — how many viral posts have you seen invoking Thomas Peterffy, or Gina Rinehart, or even Phil Knight? These are all billionaires, but none has a public cult of personality in the same way as does Musk, or Bill Gates, or Jeff Bezos, or even, to some extent, Warren Buffett. If we correctly recognized the adversarial relationship these tycoons have with the rest of the world, I also think this slop would struggle to gain traction.

Microsoft Thinks the Problem With Its A.I. Stuff Is the Name ⇥ techradar.com

Eric Hal Schwartz, TechRadar:

Microsoft has made Copilot the name and style of its AI assistant and other AI services for more than a year, but it seems the company might have a rebranding project underway with a transparent origin. In references shared on X from the appprivacy.adml file, it looks like AI-powered features in Windows 11 will be collected under the umbrella name “Windows Intelligence.” While positioning AI centrally in the operating system is certainly not a surprise, the name is either a deliberate attempt to leverage Apple’s “Apple Intelligence” branding or the biggest coincidence since Mark Zuckerberg started a company with the same name as Harvard’s digital yearbook.

I do not really care about the similarity to Apple’s branding. Far funnier, to me, is how Microsoft seems incapable of sticking with a name for anything newer than Windows. I think “Copilot” is a nice, friendly name; I think “Windows Intelligence” sounds inherently oxymoronic.

How Amazon Threatens Canada’s Post Office ⇥ disconnect.blog

Paris Marx:

That leaves us with an important question to consider. Not just what we want the future of Canada Post to be, but also what kind of society we want to live in. We should want to take advantage of the post office’s unique, nationwide infrastructure to provide more and better services to Canadians instead of dismantling something that we may never be able to rebuild. But even more than, the government should see Amazon’s low-wage, non-union model as a threat not just to Canada Post, but to Canadian workers across the board, and intervene to rein it in.

The Canadian Union of Postal Workers today gave notice of an impending strike after a year of contract negotiations. It would be horrible for our national postal system to be hollowed out by a private — and foreign — corporation with no obligation to service all Canadians.

Old iPhone Videos Hidden in Plain Sight ⇥ ben-mini.github.io

Ben Wallace:

While Send to YouTube can be thoroughly analyzed as a milestone on the “frenemy” timeline between Apple and Google, I want to explore a pleasant consequence of this moment. Apple uses the ‘IMG_XXXX’ naming convention for all images and videos captured on iOS devices, where XXXX is a unique sequence number. The first image you take is named “IMG_0001”, the second is “IMG_0002” and so on. During the Send to YouTube era of 2009 and 2012, the title of one’s YouTube video was defaulted to this naming convention. Unwitting content creators would then upload their videos on a public site with a barely-searchable name. To this day, there are millions of these videos.

Like sharing Strava routes, this feels like a throwback to a different time. I found some videos shot on the fifth-generation iPod Nano, too, which used the same naming scheme.

Apple’s ‘Best Albums’ List Available in $450 Coffee Table Book Form ⇥ assouline.com

I am not sure what amuses most about this book, as there is so much to choose from. The $450 price tag, perhaps, for what appears to be the same short essays featured in Apple Music. Maybe it is the lack of anything released before 1959. Perhaps it is in celebrating the hollowness of Apple’s ranking.

For me, though, it is that this book, which features the covers of the hundred best albums of all time, contains — according to this product page — exactly 97 illustrations. Which three album covers are missing, I wonder?

(Via Christina Warren.)

Apple Quietly Introduced iPhone ‘Inactivity Reboot’ ⇥ 404media.co

Joseph Cox, 404 Media:

On Thursday, 404 Media reported that law enforcement officials were freaking out that iPhones which had been stored for examination were mysteriously rebooting themselves. At the time the cause was unclear, with the officials only able to speculate why they were being locked out of the devices. Now a day later, the potential reason why is coming into view.

“Apple indeed added a feature called ‘inactivity reboot’ in iOS 18.1.,” Dr.-Ing. Jiska Classen, a research group leader at the Hasso Plattner Institute, tweeted after 404 Media published on Thursday along with screenshots that they presented as the relevant pieces of code.

The way this was explained in the original article does not appear to be accurate:

[…] The law enforcement officials’ hypothesis is that “the iPhone devices with iOS 18.0 brought into the lab, if conditions were available, communicated with the other iPhone devices that were powered on in the vault in AFU. That communication sent a signal to devices to reboot after so much time had transpired since device activity or being off network.” They believe this could apply to iOS 18.0 devices that are not just entered as evidence, but also personal devices belonging to forensic examiners.

None of this appears to be true. It only seems as though iPhones reboot automatically after inactivity, making them harder to crack. It seems the cops believed iPhones were secretly communicating with each other because some of them were running older iOS versions, forgetting the explanation that satisfies Hanlon’s razor: iOS is kind of buggy.

It is impossible to differentiate between improving the security of user data on an iPhone that has been stolen, and locking out police as a phone sits in an evidence locker. The former is worth pursing, and sorry about the latter.

Sponsor: Magic Lasso Adblock: Incredibly Private and Secure Safari Web Browsing ⇥ magiclasso.co

Online privacy isn’t just something you should be hoping for – it’s something you should expect. You should ensure your browsing history stays private and is not harvested by ad networks.

Magic Lasso Adblock: No ads, no trackers, no annoyances, no worries

By blocking ad trackers, Magic Lasso Adblock stops you being followed by ads around the web.

As an efficient, high performance and native Safari ad blocker, Magic Lasso blocks all intrusive ads, trackers and annoyances on your iPhone, iPad, and Mac. And it’s been designed from the ground up to protect your privacy.

Users rely on Magic Lasso Adblock to:

  • Remove ad trackers, annoyances and background crypto-mining scripts

  • Browse common websites 2.0× faster

  • Block all YouTube ads, including pre-roll video ads

  • Double battery life during heavy web browsing

  • Lower data usage when on the go

With over 5,000 five star reviews; it’s simply the best ad blocker for your iPhone, iPad, and Mac.

And unlike some other ad blockers, Magic Lasso Adblock respects your privacy, doesn’t accept payment from advertisers and is 100% supported by its community of users.

So, join over 350,000 users and download Magic Lasso Adblock today.

Wonder Has Raised Over a Billion Dollars to Be a Maker of Generic Food ⇥ grubstreet.com

Matthew Schneier, Grub Street:

What even is Wonder? Founded in 2018, it is, according to its own marketing copy, “a new kind of food hall.” More of a Potemkin food hall, really. Under its green shingle, Wonder comprises some 30 “restaurants,” which are really more like sub-brands. […]

[…] The company partners with chefs or restaurateurs for an up-front fee and equity, and its team of “culinary engineers” works with the chefs for months to develop a scalable, deliverable menu. New ideas are then piloted at a Wonder location — Downtown Brooklyn and Westfield, New Jersey, are both pilot stores — after which the food is rolled out to many more. Once items make it to the larger menu, they’re prepped in a centralized commercial kitchen in New Jersey and sent daily, mostly as kits, to Wonder’s stores, where everything is finished to order. […]

So it is a bunch of styles of food assembled — not made or cooked — in the same kitchen by the same people. Like on an airplane, famously the best cuisine you are able to get when you are stuck ten kilometres in the sky, and by no other metric.

Rebecca Deczynski, Inc, in March:

Wonder, [Marc] Lore’s New York City-based food delivery startup, which currently has 11 brick-and-mortar locations that serve fast-cooked meals by chefs including Bobby Flay and Michael Symon, has completed a $700-million funding round, the company announced today.

Wonder has raised, according to Deczynski, $1.5 billion. To be clear, neither Flay nor Symon is making your dinner.

I get how this allows a group of people to each get the kind of food they want, all in one order. But we already have restaurants that do that, and they usually suck. If you have ever dined at a place offering sushi, wings, fettuccini Alfredo, and seafood, you already know none of those things will be as good as a cheap and unfussy cuisine-specific neighbourhood joint.

This is an evolution of the ghost kitchen concept. Like those, I can see this sort of thing fragmenting communities as supporting a local restaurant is replaced with this mediocre and inexpensive — for now — venture capital-funded alternative.

The MacBook Pro Nano-Texture Display Is Matte From the Inside ⇥ petapixel.com

Jaron Schneider, of PetaPixel, reviewed the new MacBook Pro:

The only other major change to the design of this laptop is the choice to add Apple’s nano-texture display which significantly reduces glare. The MacBook Pro has been very prone to glare over the years and would have to rely solely on pure brightness to overcome it. Now it has another tool in its arsenal although the implementation is slightly different than nano-texture has been on previous Apple devices like the Pro Display XDR and iPad Pro. While those two devices have a layer of glass which is then etched with the nano-texture, the MacBook Pro doesn’t use that same glass cover. The nano-texture is, therefore, instead embedded on the inside of the display. The effect is the same, or similar enough, and Apple includes its special polishing cloth too, but it’s not strictly “necessary” to use to clean the MacBook Pro display (although it is recommended).

I have not seen it mentioned anywhere else that the texture layer is on the inside of the display. It feels like it would be more appropriate for a portable product as it would be easier to clean. But the iPad Pro is available with a matte display, too, and its texture is on the outside. If an iPhone model becomes available with a nano-texture display option, I would bet its texture is on the inside, like the MacBook Pro.

Also, notable to me is the vast price gap in nano-texture options. Choosing that option on the iPad Pro costs $100 more than the standard glass in U.S. pricing; on the MacBook Pro, it is $150; on the iMac, $200. On the Studio Display, it is $300. And, still — five years later — it remains a $1,000 upgrade on the Pro Display XDR.

The Fight

You might want to skip this one.

From the perspective of this outsider, the results of this year’s U.S. presidential election are stunning. I feel terrible for those within the U.S. who will endure another four years of having longtime institutions ripped apart by a criminal administration and its enablers in the legislative and judicial branches. This is true of just about everybody, but the brunt of the pain inflicted will — again — be directed toward the LGBTQ community, immigrants, visible minorities, and women.

As the world’s sole superpower, however, the effects of U.S. lawmaking will be felt everywhere. The incoming administration’s actions will, at best, disregard consequence. Again: at best. The rest of the world will attempt to govern itself around the whims of an unstable sex abuser, his dangerously feckless cabinet, and a host of grovelling billionaires whispering in his ear.

While the oligarchs and authoritarians of the world will have influence over what happens next in the U.S., us normal people will not. The best we can do is prevent a similar catastrophe befalling our communities. Democracies around the world have elected a raft of far-right ideologues and strongmen — in Austria, Belgium, France, Indonesia, Italy, and the Netherlands. Nationalist ideologies in Europe are now the “establishment”.

Here at home, Canada’s Conservative Party leader is more popular than his rivals and he is itching for an election. Though not our farthest-right party, his policies are of the slash-and-burn variety; his party uniformly voted against those new privacy laws.

Closer still, our provincial government is enacting massive reforms aligned with some of the most conservative U.S. states. At their recent conference, they embraced carbon dioxide as a token principle. Like many conservative governments, they are targeting people who are transgender with restrictive legislation opposed by medical professionals. These policies got the attention of Amnesty International when they were announced.

A predictable response from centrist parties is that they will move rightward to present themselves as a moderating alternative to the more hardline conservatives. I am not a political scientist, but it does not seem that growing the size of the tent will be inviting to a electorate increasingly comfortable with far-right ideas. There are thankfully still places where democracies in recent elections have not embraced a nationalist agenda, and where elections are not between shades of conservatism. Our politicians would do well to learn from them.

We each get to choose our societal role. At the moment, for those of us who do not align with these dominant forces, it can feel pretty small. This is not an airport book; I am not ending this thing on a hopeful note and a list of to-dos. I am scared of what this U.S. election means for decades to come. I am just as worried about policies close to home, and those are the ones I can try to do something about.

I am not giving up. But I am overwhelmed by how far democratic countries around the world have regressed, and how much further they are likely to go.

Canadian Government Bans TikTok’s Canadian Operations, but Not Its App ⇥ cbc.ca

Catharine Tunney, CBC News:

Citing national security concerns, the federal government has ordered TikTok to close its Canadian operations — but users will still be able to access the popular app.

My position is that TikTok should not be banned; instead, governments should focus on comprehensive privacy legislation to protect users from all avenues of data exploitation. So it is kind of a good thing the Canadian government is not prohibiting the app or users’ access — except the government’s position appears to be entirely contradictory. It is very worried about user privacy:

Former CSIS director David Vigneault told CBC News it’s “very clear” from the app’s design that data gleaned from its users “is available to the government of China” and its large-scale data harvesting goals.

But laws drawn up in 2022 which would restrict these practices have been stuck in committee since May. So there is an ostensibly dangerous app posing a risk to Canadians, and the government’s response is to let people keep using it while shutting down the company’s offices? The Standing Committee on Industry and Technology had better get moving.

The Powerful Density of Hypertextual Writing ⇥ kottke.org

Please forgive me for quoting this New York Times editorial board piece in its entirety:

You already know Donald Trump. He is unfit to lead. Watch him. Listen to those who know him best. He tried to subvert an election and remains a threat to democracy. He helped overturn Roe, with terrible consequences. Mr. Trump’s corruption and lawlessness go beyond elections: It’s his whole ethos. He lies without limit. If he’s re-elected, the G.O.P. won’t restrain him. Mr. Trump will use the government to go after opponents. He will pursue a cruel policy of mass deportations. He will wreak havoc on the poor, the middle class and employers. Another Trump term will damage the climate, shatter alliances and strengthen autocrats. Americans should demand better. Vote.

This is arresting, and not just because of how direct it is. It makes the best of the medium of the web in a way that would simply be impossible in print: by stacking link upon link.

Jason Kottke:

A simple list of headlines would have done the same basic job, but by presenting it this way, the Times editorial board is simultaneously able to deliver a strong opinion; each of those links is like a fist pounding on the desk for emphasis. Lies, threat, corruption, cruel, autocrats — bam! bam! bam! bam! bam! Here! Are! The! Fucking! Receipts!

John Gruber:

The way I’ve long thought about it is that traditional writing — like for print — feels two-dimensional. Writing for the web adds a third dimension. It’s not an equal dimension, though. It doesn’t turn writing from a flat plane into a full three-dimensional cube. It’s still primarily about the same two dimensions as old-fashioned writing. What hypertext links provide is an extra layer of depth. Just the fact that the links are there — even if you, the reader, don’t follow them — makes a sentence read slightly differently. It adds meaning in a way that is unique to the web as a medium for prose.

Both these pieces are so good, I just had to point to them and add my own stance: link often, and link generously. Writing on the web is not like print, where too many citations can feel interruptive. On the web, it is just part of the visual vocabulary. It encourages a more expansive tapestry where references can be used for more than just acknowledging source material. One can also point to definitions, tangential pages, or jokes. The hyperlink is among the singularly magical elements of the web.

The Times is among the worst offenders for not crediting others’ past reporting by linking to it. You will notice every one of the links in its paragraph is to another Times story, which makes sense in this context. (It would be perhaps slightly more powerful if each was to a different publication to capture the breadth of this uniquely odious man, but then the Times runs the risk of pointing readers to something outside its known editorial context.) In other reporting, there is simply no excuse for the Times to not link to someone else’s preceding work.

Link often, link generously.

We Are Cursed to Have More Connector Standards and, Also, Articles About Connector Standards

In short.

In long:

Ten years ago, the USB Implementers Forum finalized the specification for USB-C 1.0, and the world rejoiced, for it would free us from the burden of remembering which was the correct orientation of the plug relative to the socket. And lo, it was good.

And then we all actually got around to using USB-C devices and realized this whole thing is a little bit messy. While there was now a universal connector, the capabilities of the cable can range from those which support only power with maybe a trickle of data, all the way up to others which carry data at USB4 speeds. But that is not all. It might also support various Thunderbolt standards — 3, 4, and now 5 — and DisplayPort. That is neat. Again, this is all done using the same connector size and shape, and with cables that look practically interchangeable.

Which brings us to Ian Bogost, writing in the Atlantic — a requisite destination for intellectualized lukewarm takes — about his cable woes:

I am unfortunately old enough to remember when the first form of USB was announced and then launched. The problem this was meant to solve was the same one as today’s: “A rat’s nest of cords, cables and wires,” as The New York Times described the situation in 1998. Individual gadgets demanded specific plugs: serial, parallel, PS/2, SCSI, ADB, and others. USB longed to standardize and simplify matters — and it did, for a time.

But then it evolved: USB 1.1, USB 2.0, USB 3.0, USB4, and then, irrationally, USB4 2.0. Some of these cords and their corresponding ports looked identical, but had different capabilities for transferring data and powering devices. I can only gesture to the depth of absurdity that was soon attained without boring you to tears or lapsing into my own despair. […]

Reader — and I mean this with respect — I am only too willing to bore you to tears with another article about USB-C. Bogost is right, though. The original USB standard unified the many different ports one was expected to use for peripherals. It basically succeeded for at least two of them: the keyboard and mouse. Both require minimal data, so they work fine regardless of whether the port supports USB 1.1 or USB 3.1. Such standardization also came with loads more benefits, too, like reducing setup and configuration once necessary for even basic peripherals.

Where things got complicated is when data transfer speeds actually matter. USB 1.1 — the first version most people actually used — topped out at 12 Mbits per second; USB 2.0 could do 480 Mbits per second. Even so, the ports and cables looked identical. If you plugged an external hard drive into your computer using the wrong cable, you would notice because it would crawl.

This begat more specs allowing for higher speeds, requiring new cables and — sometimes — new connectors. And it was kind of a mess. So the USB-IF got together and created USB-C, which at least solves some of these problems. It is a more elegant connector and, so far, it has been flexible enough to support a wide range of uses.

That is kind of the problem with it, though: the connector can do everything, but there is no easy way to see what capabilities are supported by either the port or the cable. Put another way, if you connect a Thunderbolt 5 hard drive using the same cable as you use to charge new Magic Mouse and Keyboard, you will notice, just as you did twenty years ago.

Bogost, after describing his array of gadgets connected by USB-A, USB-C, and micro-HDMI:

This chaos was supposed to end, with USB-C as our savior. The European Union even passed a law to make that port the charging standard by the end of this year. […]

Hope persists that someday, eventually, this hell can be escaped — and that, given sufficient standardization, regulatory intervention, and consumer demand, a winner will emerge in the battle of the plugs. But the dream of having a universal cable is always and forever doomed, because cables, like humankind itself, are subject to the curse of time, the most brutal standard of them all. At any given moment, people use devices they bought last week alongside those they’ve owned for years; they use the old plugs in rental cars or airport-gate-lounge seats; they buy new gadgets with even better capabilities that demand new and different (if similar-looking) cables. […]

If the ultimate goal is a single cable and connector that can do everything from charge your bike light to connect a RAID array — do we still have RAID arrays? — I think that is foolish.

But I do not think that is the expectation. For one thing, note Bogost’s correctly chosen phrasing of what the E.U.’s standard entails. All devices have unified around a single charging standard, which does not demand any specialized cable. I use a Thunderbolt cable to sync my iPhone and charge my third-party keyboard, because I cannot be tamed.1 The same is true of my laptop and also my wife’s, the headphones I am wearing right now, a Bluetooth speaker we have kicking around, our Nintendo Switch, and my bicycle tire pump. Having one cable for all this stuff rules.

If you need higher speeds, though, I would bet you know that. If the difference between Thunderbolt 4 and Thunderbolt 5 matters to you, you are a different person than most. And, I would wager, you are probably happy that you can connect a fancy Thunderbolt drive to any old USB-C port and still read its contents, even if it is not as fast. That kind of compatibility is great.

Lookalike connectors are nothing new, however. P.C. users probably remember the days of PS/2 ports for the keyboard and mouse, which had the same plugs but were not interchangeable. 3.5mm circular ports were used for audio out, audio in, microphone — separate from audio in, for some reason — and individual speakers. This was such a mess that Microsoft and Intel decided PC ports needed colour-coding (PDF). Even proprietary connectors have this problem, as Apple demonstrated with some Lightning accessories.

We are doomed to repeat this so long as the same connectors and cables describe a wide range of capabilities. But solving that should never be the expectation. We should be glad to unify around standards for at least basic functions like charging and usable data transfer. USB-C faced an uphill battle because we probably had — and still have — devices which use other connectors. While my tire pump uses USB-C, my bike light charges using some flavour of mini-USB port. I do not know which. I have one cable that works and I dare not lose it.

Every newer standard is going to face an increasingly steep hill. USB-C now has a supranational government body mandating its use for wired charging in many devices which, for all its benefits, is also a hurdle if and when someone wants to build some device in which it would be difficult to accommodate a USB-C port. That I am struggling to think of a concrete example is perhaps an indicator of the specificity of such a product and, also, that I am not in the position of dreaming up such products.

But even without that regulatory oversight, any new standard will have to supplant a growing array of USB-C devices. We may not get another attempt at this kind of universality for a long time yet. It is a good thing USB-C is quite an elegant connector, and such a seemingly flexible set of standards.

  1. I still use a Lightning Magic Trackpad which means I used to charge it and sync my iPhone with the same cable, albeit more slowly. Apparently, the new USB-C Magic Trackpad is incompatible with my 2017 iMac, though I am not entirely sure why. Bluetooth, maybe? Standards! ↥︎