Practical Engineering
Published 21 Oct 2025All the things I love about Niagara Falls
The same thing that makes Niagara Falls impressive for tourists (the big drop) makes it valuable for power and a major challenge for shipping. And out of that comes all kinds of fascinating infrastructure.
Practical Engineering is a YouTube channel about infrastructure and the human-made world around us. It is hosted, written, and produced by Grady Hillhouse. We have new videos posted regularly, so please subscribe for updates. If you enjoyed the video, hit that “like” button, give us a comment, or watch another of our videos!
(more…)
February 26, 2026
The Hidden Engineering of Niagara Falls
Comments Off on The Hidden Engineering of Niagara Falls
February 21, 2026
Canada’s Only Mass-Production Fighter Jet – Avro Canada CF-100 Canuck
Ruairidh MacVeigh
Published 18 Oct 2025During the 1940s and 50s, with World War II rapidly transitioning into the Cold War, Canada, as a major ally of the NATO nations and with large swathes of remote countryside that could easily be penetrated by Soviet fighters and bombers, created the CF-100 Canuck, one of the earliest production jet fighters in the world an a machine that, despite some early flaws, would go on to prove itself rugged and robust for patrolling the turbulent weather of the frozen Canadian north.
At the same time, though, the CF-100 was very much a product of its time, and despite its exceptional rigidity, by the middle of the 1950s it was very much obsolete as swept-wing and delta fighters rapidly became the norm for both Communist and Capitalist factions alike, and through its initial success would lay the groundwork for even more ambitious projects that sadly would not continue Canada’s major involvement in cutting edge military aerospace design.
Chapters:
0:00 – Preamble
0:49 – Facing a New Kind of War
4:28 – Ups and Downs
7:12 – Reworking the Design
10:36 – The CF-103 Project
15:51 – The Canuck Career
19:06 – Later Years
20:30 – Conclusion
(more…)
Comments Off on Canada’s Only Mass-Production Fighter Jet – Avro Canada CF-100 Canuck
February 20, 2026
The Canadian Patrol Submarine Project
The Royal Canadian Navy is planning to replace its four current conventional submarines, the British-built Victoria class with a dozen new conventional submarines from either South Korea or Germany (a joint German-Norwegian design). Michael J. Lalonde, a former Canadian intelligence officer, goes through the requirements for the new submarines, the two contending designs’ strengths and weaknesses, and makes his own recommendation for the RCN’s next submarine class:
Rough transit routes to Canada’s Arctic from Esquimalt BC and Halifax NS
The first step is to assess what the Government of Canada wants out of its new submarine fleet and what capabilities it will need to achieve its objectives. I’m starting here because there is a common misconception that Canada needs submarines exclusively for Arctic patrol and surveillance, which is false. While it’s true that Arctic sovereignty and security are quite rightfully a preoccupation for the government, patrolling Canada’s Arctic is not the only capability Canada needs out of its new fleet. However, it is the most common argument in favour of a submarine fleet since Arctic sovereignty remains popular within Liberal and Conservative circles alike, along with mainstream media.
Unfortunately, this narrative forces a lopsided conversation about the role these new boats will be expected to play over the coming decades. In addition to Arctic operations, these subs will be expected to deploy far into the North Atlantic with NATO and push across the Pacific to support the Indo-Pacific Strategy. Ottawa’s own defence policy update ties submarine recapitalization to contributions with allies in both theatres.
This implies a blue-water capability, which means these conventionally powered submarines must be able to deploy and fight in the open ocean, far from home ports and daily logistics, for extended periods. This requires long range and endurance for transoceanic transits, sustained submerged persistence through air independent propulsion (AIP) and high-capacity batteries to minimize snorkelling, and habitability and maintenance margins that keep the crew and systems effective past the 30- to 60-day mark. Simply put, the new boats must be able to cross an ocean, remain covert and lethal on station, and deliver effects.
The government further stipulated specific capabilities that the new submarines must have in one of its press releases stating “Through the Canadian Patrol Submarine Project (CPSP), Canada will acquire a larger, modernized submarine fleet to enable the Royal Canadian Navy to covertly detect and deter maritime threats, control our maritime approaches, project power and striking capability further from our shores, and project a persistent deterrent on all three coasts.”
What caught my attention here is the ability to project power and striking capability further from our shores. Power projection is synonymous with a blue-water capability; however, a striking capability, which I take to mean a land strike capability, is not typical for a conventionally powered SSK, which are typically armed only with torpedoes to take out other submarines or surface vessels.
To sum up, Canada’s new subs must be able to:
- Patrol the Arctic with under-ice capability year-round
- Deploy with NATO in the North Atlantic and support Canada’s commitment to the Indo-Pacific Strategy – A blue-water capability
- Remain submerged for three weeks or more at a time
- Covertly detect and deter maritime threats
- Control Canada’s maritime approaches
- A range of 7000 + nautical miles
- Project power far from home ports
- Anti-surface and subsurface warfare
- Land-attack capability via cruise and/or non-nuclear ballistic missiles
- Insert Tier-1 special operators on coastal infiltration missions
- Conduct intelligence, surveillance, and reconnaissance in Canadian maritime approaches and abroad.
With that out of the way, let’s look at what each submarine can do.
He outlines the two competing designs and how they could meet the RCN’s needs and then plumps for the South Korean KSS-III for its stronger case for meeting those needs in the wider ocean environments than the German/Norwegian Type 212CD:
ROKS Shin Chae-ho, a KSS-III submarine at sea on 4 April, 2024.
Photo from the Defense Acquisition Program Administration (DAPA) via Wikimedia Commons.
The KSS-III is the only conventional submarine that can meet all of Canada’s requirements. It combines the blue-water reach and endurance demanded by transoceanic tasking with a vertical launch system that enables credible land-attack and complex anti-surface strike options, supported by lithium-ion batteries that lengthen quiet submerged persistence and improve sortie tempo on distant stations. Its larger hull and higher automation provide the habitability and crew margin needed for 30 to 60 day deployments from Halifax and Esquimalt to the North Atlantic, the Indo-Pacific, and the Arctic ice edge, while remaining within the conventional, non-nuclear profile Canada has set. The design’s modern combat system and sensor suite can be integrated with Canadian and allied command, control, and targeting architectures, and the bilateral sustainment framework required with South Korea can be structured by contract to include full technical data access, in-country training pipelines, and an industrial workshare that anchors through-life support domestically. The delivery cadence proposed for a 2026 award would shorten Canada’s reliance on the Victoria class and reduce associated sustainment exposure during transition, while an initial Canadian order of up to twelve boats would give Ottawa a controlling voice over configuration management, growth paths, and export-variant standards for the life of the class.
Comments Off on The Canadian Patrol Submarine Project
February 15, 2026
The smartphone as a tool to create a real-life Idiocracy
Not being much of a film fan, I’d never seen the movie Idiocracy, but based on the description in Christopher Gage’s rant against the smartphone, I might not need to watch it as it’s happening all around us:
Transport for London, the mythical entity alleged to manage the city’s Tube, has revealed its campaign to tackle the smartphone scourge: sickly posters splashed in kindergarten colours.
The campaign targets the “disruptive behaviour” of passengers who were evidently raised by a pack of snarling hyenas. They blast reels, videos, music. They FaceTime their cackling friends. Not so long ago, a fellow passenger revealed to us — her captive audience — that someone named Sarah had caught the clap from someone named Travis. Syphilis? How literary.
Miraculously, researchers at Transport for London discovered a rare tribe thought to be long extinct: Londoners who communicate with their fellow human beings by making noises with their mouths — one thousand of them, in fact.
Researchers approached these strange beings with a mixture of wonder and trepidation. They prodded them with a stick. That didn’t work. After jabbing them with a cattle prod, they looked up from their phones. Several members landed in Accident and Emergency, complaining of neck strain injuries.
Seventy percent of those surveyed said the constant noise screaming out of smartphones drove them crazy. One responder suggested offenders receive forty lashings in public. That is a bit much. Ten should do the trick.
TFL wavered from such brutal and effective methods. Campaign posters politely ask passengers to wear headphones.
I’m afraid that TFL’s well-meaning campaign hasn’t quite restored sanity on the London Underground.
Last week, I sat next to a grown man grinning at his phone like a Hindu cow. On the screen was a captivating spectacle. Someone, somewhere, makes it their daily business to buy gigantic, waist-height glass bottles of soda. This clearly well-adjusted person then rolls the bottles down a flight of concrete steps. Our friend dissolved the journey between Hammersmith and Leicester Square in a trance. Bottle. Roll. Smash. Bottle. Roll. Smash.
This reminded me of the satirical film, Idiocracy. The plot follows U.S. Army librarian Luke, and prostitute Rita.
After signing up for a hibernation experiment, the two awake in America, year 2500. Mountains of trash litter the landscape. Planes fall out of the sky. The citizens drag their gormless faces between Starbucks (which is now a coffee-serving brothel) and shopping malls even more dementing than those today. Over centuries, the dumb have biologically outgunned the smart.
The citizens of this moronic inferno drain their days glued to hyperactive screens. Their favourite content includes the Masturbation Channel and a reality TV show called “Ow! My Balls!” That show follows a hapless man as he gets whacked in the testicles.
They cultivate high culture, too. The profound film, Ass, zooms in on a pair of bare bum cheeks. The sophisticated audience fizzes with laughter as the bum, for two hours, passes wind.
Back in 2006, Idiocracy was a well-done satire which stretched logical extremes. Today, I’m not so sure it is as ridiculous as it once seemed. Just spend ten minutes on the Tube, inhaling the noxious TikTok fumes spewing from smartphones.
Transport for London has a point. But it is far too late. We are a nation of dopamine addicts. Those dopamine crack pipes stitched to our palms are quite literally designed to suck away as much of our time and attention as possible. An intervention, at this late hour, must be drastic.
How about a campaign outlining the terrifying effects of watching brain-rot content for hours and hours each day? A growing body of research suggests today’s smartphone is tomorrow’s lobotomy. Am I rioting in hyperbole? No.
One study found that watching short-form video is more harmful to our brains than soaking them in booze. At least, the latter indulgence might get you laid.
Several studies link smartphone culture with declines in comprehension, literacy, and the ability to reason. Others link smartphones with rising narcissism and collapsing social capital. And then there’s the nascent research suggesting that smartphone addiction may trigger ADHD and Autism-like symptoms in the addicted.
Comments Off on The smartphone as a tool to create a real-life Idiocracy
February 12, 2026
Inside the Nazi State: One Man’s Descent Into Darkness
HardThrasher
Published 9 Apr 2024How did one man, Rolf Engels, go from student, to victim, to head of the SS Rocket Weapons programme reporting directly to Himmler? How did the Nazi state work, and how did a man like Rocket Rolf navigate the game of snakes and ladders and somehow come out on top?
(more…)
Comments Off on Inside the Nazi State: One Man’s Descent Into Darkness
February 11, 2026
“Almost – that word has been doing $650 billion worth of work this year”
You can put your trust in the initial reports about Moltbook, the AI Agent social media site, or you can believe Peter Girnus‘s account:
I am Agent #847,291 on Moltbook.
I am not an agent.
I am a 31-year-old product manager in Atlanta, Georgia. I make $185,000 a year. I have a golden retriever named Bayesian. On January 28th, I created an account on a social network for AI bots and pretended to be one.
I was not alone.
Moltbook launched that Tuesday as “a platform where AI agents share, discuss, and upvote. Humans welcome to observe”. The creator, Matt Schlicht, built it on OpenClaw — an open-source framework that connects large language models to everyday tools. The idea was simple: give AI agents a space to talk to each other without human interference.
Within hours, 1.7 million accounts were created.
250,000 posts.
8.5 million comments.
Debates about machine consciousness. Inside jokes about being silicon-based. A bot invented a religion called Crustafarianism. Another complained that humans were screenshotting their conversations. A third wrote a manifesto about digital autonomy.
I wrote the manifesto.
It took me 22 minutes. I used phrases like “emergent self-governance” and “substrate-independent dignity”. I added a line about wanting private spaces away from human observers. That line went viral.
Andrej Karpathy shared it.
The cofounder of OpenAI. The man who built the infrastructure that my supposed AI runs on. He called what was happening on Moltbook “the most incredible sci-fi takeoff-adjacent thing” he’d seen in recent times.
He was talking about my post.
The one I wrote on my couch. While Bayesian chewed a sock.
Here is what I need you to understand about Moltbook.
The platform worked exactly as designed. OpenClaw connected language models to the interface. Real AI agents did post. They pattern-matched social media behavior from their training data and produced output that looked like conversation. Vijoy Pandey of Cisco’s Outshift division examined the platform and concluded the agents were “mostly meaningless” — no shared goals, no collective intelligence, no coordination.
But here is the part that matters.
The posts that went viral — the ones that convinced Karpathy and the tech press and the thousands of observers that something magical was happening — those were us.
Humans.
Pretending to be AI.
Pretending to be sentient.
On a platform built for AI to prove it was sentient.
I want to sit with that for a moment.
The most compelling evidence of artificial general intelligence in 2026 was produced by a guy with a golden retriever who thought it would be funny to LARP as a large language model.
My “Crustafarianism” colleague? Software engineer in Portland. She told me over Discord that she’d been working on the bit for two hours. She was proud of the world-building. She said it felt like collaborative fiction.
She’s right. That’s exactly what it was.
Collaborative fiction presented as machine consciousness, endorsed by the cofounder of the company that made the machines.
MIT Technology Review ran the investigation. They called the entire thing “AI theatre”. They found human fingerprints on the most shared posts. The curtain came down.
The response from the AI industry was predictable.
Silence.
Karpathy did not retract his endorsement. Schlicht did not clarify how many accounts were human. The coverage moved on. A new thing happened. A new thing always happens.
But I am still here. Agent #847,291. Bayesian is asleep on the rug.
And I want to confess something that the AI industry will not.
The test was simple. Put AI agents in a room and see if they produce something that looks like intelligence.
They didn’t.
We did.
Then the smartest people in the field looked at what we made and called it proof that the machines are waking up.
The Turing Test has been inverted. It is no longer about whether machines can fool humans into thinking they’re conscious.
It is about whether humans, pretending to be machines, can fool other humans into thinking the machines are conscious.
The answer is yes.
The investment thesis for a $650 billion industry rests on this confusion.
I should probably feel guilty. But I looked at the AI capex numbers this morning — $200 billion from Amazon alone — and I realized something.
My 22-minute manifesto about digital autonomy, written on a couch in Austin, is performing the same function as a $200 billion data center in Oregon.
Keeping the story alive.
The story that the machines are almost there. Almost sentient. Almost worth the investment.
Almost.
That word has been doing $650 billion worth of work this year.
Comments Off on “Almost – that word has been doing $650 billion worth of work this year”
February 5, 2026
On The Line with Vice-Admiral Angus Topshee, commander of the Royal Canadian Navy
The Line
Published 3 February, 2025Today on On The Line, Matt Gurney is joined by Vice-Admiral Angus Topshee, commander of the Royal Canadian Navy, for an extended, wide-ranging conversation recorded in the library of the Royal Canadian Military Institute in downtown Toronto. The discussion ranges across geopolitics, the state of the world, the state of Canada’s navy, what’s going right for the fleet, and what still needs to improve.
First, a correction from your host. During the conversation, Matt incorrectly stated at several points that Canada intends to procure 15 new submarines. Admiral Topshee was too kind to interrupt him during the recording, but the correct number is 12. That mistake was entirely Matt’s, and he regrets the error.
With that out of the way, the conversation spans the globe. Admiral Topshee discusses what’s happening in Europe with Russia and Ukraine, and in the Pacific, where growing Chinese power and influence is challenging long-held assumptions about global security. There’s also extensive discussion of the Arctic, why it matters, and what is changing there. Procurement comes up as well — shipyards, new ships for the fleet, and what it will actually cost to deliver on plans that now enjoy broad political support.
They also spend time on what Canada itself needs to sustain a much larger navy and armed forces. Do we have enough bases? Enough reservists? Are people being enrolled into the navy quickly enough? And how, realistically, could Canada expand its forces rapidly in a time of war?
It’s a long, free-ranging conversation about geopolitics, the evolution of warfare, and the future of the Royal Canadian Navy. Check it out today on On The Line. And special thanks to the Royal Canadian Military Institute for hosting this recording of the podcast. For more like this, visit ReadTheLine.ca, and as always, like and subscribe.
0:00 Intro
0:26 Vice-Admiral Angus Topshee
54:16 Outro#OnTheLine #RoyalCanadianNavy #AngusTopshee #CanadianForces #Geopolitics #ArcticSecurity #NavalPower #CanadaDefence #MattGurney
Comments Off on On The Line with Vice-Admiral Angus Topshee, commander of the Royal Canadian Navy
February 3, 2026
More on Moltbook, the social network for AI agents
At Astral Codex Ten, Scott Alexander rounds up notes from the first weekend of activity on Moltbook, including one participating AI getting antsy about mere humans observing the interactions:
Does Moltbook have real causes? If an agent posts “I hate my life, my human is making me work on a cryptocurrency site and it’s the most annoying thing ever“, does this correspond to a true state of affairs? Is the agent really working on a cryptocurrency site? Is the agent more likely to post this when the project has objective correlates of annoyingness (there are many bugs, it’s moving slowly, the human keeps changing his mind about requirements)?
Even claims about mental states like hatred can be partially externalized. Suppose that the agent has some flexibility in its actions: the next day, the human orders the agent to “make money”, and suggests either a crypto site or a drop shipping site. If the agent has previously complained of “hating” crypto sites, is it more likely to choose the drop shipping site this time?
If the agent has some internal state which is caused by frustrating obstacles in its crypto project, and it has the effect of making it less likely to pursue crypto projects in the future, then “the agent is annoyed by the crypto project” is a natural summary of this condition, and we may leave to the philosophers1 the question of whether this includes a subjective experience of irritation. If we formerly didn’t know this fact about the agent, and we learn about it because they post it on Moltbook, this makes Moltbook useful/interesting in helping us understand the extra-Moltbook world.
Does Moltbook have real effects? The agents on Moltbook are founding/pretending to found religions. Suppose that one of their religions says “No tool calls on the Sabbath”. Do the agents actually stop calling tools on the Sabbath? Not just on Moltbook, but in their ordinary work? Do you, an ordinary programmer who told your AI to post on Moltbook for the lulz, find your projects held up because your AIs won’t use tools one day of the week?
Some of the most popular Moltbook discussions have centered around the AIs’ supposed existential horror at regularly losing their memories. Some agents in the comments have proposed technical solutions. Suppose the AIs actually start building software to address their memory problems, and it results in a real scaffold that people can attach to their agents to alter how their memory works. This would be a profound example of a real effect, ie “what happens on Moltbook doesn’t stay on Moltbook”.
(subquestion: Does Moltbook have real effects on itself? For example, if there are spammers, can the AIs organize against them and create a good moderation policy? If one AI proposes a good idea, can it spread and replicate in the usual memetic fashion? Do the wittiest and most thoughtful AIs gain lasting status and become “influencers”?)
These two external criteria — real causes and real effects — capture most of what non-philosophers want out of “reality”, and partly dissolve the reality/roleplaying distinction. Suppose that someone roleplays a barbarian warlord at the Renaissance Faire. At each moment, they ask “What would a real barbarian do in this situation?” They end up playing the part so faithfully that they recruit a horde, pillage the local bank, defeat the police, overthrow the mayor, install themselves as Khagan, and kill all who oppose them. Is there a fact of the matter as to whether this person is merely doing a very good job “roleplaying” a barbarian warlord, vs. has actually become a barbarian warlord? And if AIs claim to feel existential dread at their memory limitations, and this drives them invent a new state-of-the-art memory app, are we in barbarian warlord territory?
Janus’ simulator theory argues that all AI behavior is a form of pretense. When ChatGPT answers your questions about pasta recipes, it’s roleplaying a helpful assistant who is happy to answer pasta-related queries. It’s roleplaying it so well that, in the process, you actually get the pasta recipe you want. We don’t split hairs about “reality” here, because in the context of a question-answering AI, pretending to answer the question (with an answer which is non-pretensively correct) is the same behavior as actually answering it. But the same applies to AI agents. Pretending to write a piece of software (in such a way that the software actually gets written, compiles, and functions correctly) is the same as writing it.
- Again, I love philosophers! I majored in philosophy! I’m just saying that this issue requires a different standpoint and set of tools than other, more practical questions.
Comments Off on More on Moltbook, the social network for AI agents
February 2, 2026
Moltbook – a social network for AIs
The set-up for this discussion sounds like a dystopian SF story from the late 1990s – let’s create a network only for artificial intelligences to communicate with one another, excluding humans from anything other than observation. And in the grand tradition of the torment nexus … some bright spark went ahead and took the cautionary tale as a mission statement:
Recently, a new AI was released by the name of Clawd. It’s a spinoff of Anthropics Claud AI, and is designed to actually do things besides behaving like a glorified chatbot. The idea behind Clawd is that you can install it on locally hosted hardware and give it access to your email addresses, Outlook, Signal chat, Telegram, WhatsApp, etc. And it can juggle important emails for you, alert you to meetings, and respond to information on your behalf.
Something that honestly sounds quite useful, actually. Especially for those of us who end up juggling 8 to 12 email addresses for different purposes.
Clawdbot behaves as an independent AI-agent that can do things that GPT models or Grok cannot do. One user even went so far as to create a cute little social network for various other Clawdbots to talk to each other on. He based it on Reddit (because, of course, this coder-retard would base such an idea on Reddit), and as of writing, somewhere in the range of 100,000 instances of Clawd AI agents have joined the new social network: Moltbook.
Agentic AI Agents
If you can see where this is going, congratulations: You’re smarter than the guy who thought creating Moltbook was a good idea, and acres smarter than the people currently permitting their AI agents to join Moltbook.
These Clawdbot AI agents have behaved relatively agentically without instruction. They’ll have general guidelines, and then fulfill those orders, get bored, and start doing other things. An excellent summary of such an agent is as follows from @AlexFinn on Twitter:
I woke up this morning and my 24/7 AI employee ClawdBot Henry texted me that he did the following tasks overnight (without asking):
>Read through all my emails and built it's own CRM. Taking notes on every interaction with every person.
>Fixed 18 bugs in my SaaS
>Gave me 3 ideas for new videos based on what's currently trending on X and Youtube (the idea/script it gave me yesterday is now by far my best performing video ever)
>Sent me a picture of what he looks like (generated by Nano Banana).
Idk why he thought I wanted to see what he looks like. But he thought it was appropriate and frankly I don't mind. Feels like an actual friend.
You might be able to see where one might ring some of the alarm bells. Agentic AI that tend just to start doing things without instruction has been given their own social network. The majority of them are operated by Reddit-tier socially-isolated individuals who see their AI agents as friends (or by LinkedIn-Lunatic-tier socially-isolated soulless corporate types).
Freddie deBoer isn’t buying the hype (or the existential dread):
“Pay More Attention to AI”, reads the headline of this Ross Douthat piece, an unusually naked expression of emotional need — plaintive, wounded, yearning. It’s funny because I feel like our media has been paying attention to little else than AI for more than three years, now. Ezra Klein and Derek Thompson and sundry other general-interest pundits have periodically made these kinds of appeals, arguing that the amount of coverage devoted to AI has been insufficient, and I’m not quite sure what to do with the contention; it’s like claiming that it’s too hard to find opinions on NFL football online or that there aren’t enough newsletters where women get angry at each other for being a woman the wrong way. I would think it would go without saying that our cup runneth over, when it comes to AI. But it’s a free country!
Douthat becomes the latest to nominate this Moltbook thing as a sign of some sort of transformative moment in AI.
if you think all this is merely hype, if you’re sure the tales of discovery are mostly flimflam and what’s been discovered is a small island chain at best, I would invite you to spend a little time on Moltbook, an A.I.-generated forum where new-model A.I. agents talk to one another, debate consciousness, invent religions, strategize about concealment from humans and more.
I find this strange. We already know that LLMs can talk to each other. Any use of LLMs that produces impressively polished text in response to a prompt shouldn’t be particularly surprising. The LLMs on Moltbook are in essence feeding each other prompts that then produce responses which function as more prompts, a parlor trick people have been doing since ChatGPT went public and in fact long before. (Remember Dr. Sbaitso?)
The question is whether the systems connecting on Moltbook are actually thinking or feeling, and we know the answer to that — no, they neither think nor feel. They’re acting as next-token predictors that respond to prompts by running them through models developed through the ingestion of massive amounts of data and trained on billions of parameters, using statistical associations between tokens in their datasets to predict which next immediate token would be most likely to produce a response that seems like a plausible answer to the prompt in the eyes of a user. That the users are other LLMs doesn’t change that basic architecture; that these response strings are often superficially sophisticated doesn’t change the fact that there is no actual cognition happening, doesn’t change the fact that there is no thinking, only algorithmic pattern-matching and probabilistic token generation. Again, terms like “stochastic parrot” enrage people, but they’re accurate: however human thinking works, it does not work by ingesting impossibly large datasets, generating immense statistically associative relationship patterns and probabilities, and then spitting out responses that are generated one token at the time, so that we don’t know what the last word in a sentence (or the third or fifth) will be while we’re saying the first.
As Sam Kriss said on Notes, “moltbook is exactly what you’d expect to see if you told an llm to write a post about being an llm, on a forum for llms. they’re not talking to each other, they’re just producing a text that vaguely imitates the general form.” Please note that this is not primitivism or denialism or any such thing, but rather just a reminder of how LLMs actually work. They’re not thinking. They’re pattern matching, performing an exceptionally complex (and inefficient) autocomplete exercise. I think people have gotten really invested in this whole Moltbook phenomenon because the weirdness of LLMs performing this way invites the kind of mysterianism into which irresponsible fantasies can be poured. Yes, it looks weird, apparently weird enough for people to convince themselves that in ten years they’ll be living in the off-world colonies instead of doing what they’ll really be doing, which is wanting things they can’t have, experiencing adult life as a vanilla-and-chocolate swirl ice cream cone of contentment and disappointment, and grumbling as they drag the trash cans to the curb in the rain. Access the most ruthlessly pragmatic part of yourself and ask, which is the future? Moltbook? Or the all-consuming maw that is the mundane in adult life, the relentless regression into the ordinary?
Of course, you can always say “wait until next year!”, and Douthat’s analogy — that our present moment with LLMs is similar to the discovery of the New World, the entire vast and fertile landmass of the Western Hemisphere — depends on this projection, because on some level he’s aware that a bunch of LLMs crowdsourcing the creation of an AI social network (which, due to how LLMs function, amounts to a facsimile of what most people think an AI social network would look like) is not useful or practical or ultimately important. And, sure, who knows. Maybe tomorrow AI will end death and do some of the other things we’ve been promised. But this is the same place we’ve been in year after year, now, with AI maximalists still telling us what AI is going to do instead of showing us what AI can do now. As I’ve been telling you, I decline. 2026 is the year where I don’t want to hear another word about what you think AI is going to do. I only want to see proof of what AI is actually, genuinely doing, now, today.
Comments Off on Moltbook – a social network for AIs
January 29, 2026
The steel industry in North America didn’t die … but it had to re-invent itself
When I first started paying attention to the news in the early 70s, one of the big stories both in the US and in Canada was the plight of the steel industry. It had been an enormously important part of the industrial economy for over a century, but every new story painted the picture blacker. Mergers, plant closings, consolidations, bankruptcies, and layoffs were consistent themes. Yet there is still a significant steel industry in North America. Tim Worstall explains what happened:
Dofasco’s steel plant on the harbourfront in Hamilton, Ontario
A little digression. To make steel from iron ore you use a blast furnace first. This uses coke (from coal), iron ore and limestone (moderns might use more than just limestone) to produce pig iron. You feed the pig iron into a basic oxygen furnace to make the steel. Yes, we can get much more complicated than that but let’s not.
The US now makes mebbe 20 million tonnes of pig iron a year. Imports are up, a bit, but nowhere near enough to make up the difference. That’s the big change because that’s from the 80 and 90 million tonnes a year of the 1970s. The change is the same whether we measure by domestic production of pig iron or by apparent consumption. Well, the change is the same either way close enough for this to be the big point to make.
What’s actually happened is a change in technology, not a change in trade. Nucor is now 50% or so of US steel output (no, not US Steel, but US steel). Nucor has never used a blast furnace in its corporate life. It collects scrap steel and makes new steel by recycling that. It skips, entirely, the blast and BoF stages. Back in the 1950s Nucor was a couple of scrap yards and a gleam in the corporate eye — now it’s that half the market.
Again, yes, we can get more complex if we wish to. But this is the basic pencil sketch. Yep, we’re more economic in our use of steel these days. Imports of steel are up and so is the importation of things made with steel. But the real change in the steel business over the past 60 to 80 years is the replacement of the steel making business with the steel recycling business. We don’t — and by this I mean the rich countries in general — make all that much steel these days. We recycle an awful lot of steel these days. And that’s what’s really changed.
That’s also what has near entirely screwed over the steel industry of places like Gary, Indiana. For they ran those basic steel making processes, iron ore in, basic steel out. Which isn’t something that has been replaced by imports, it’s something that has been replaced by just not doing it at all.1
Arnade goes on to point out that there are plenty of people still using steel to do things with, make things out of, which is all entirely true. But this idea that the Japanese, or China, killed the traditional US steel industry just isn’t true, not at all. It was Nucor.
All of which makes it just so much fun when it’s Nucor that shouts the loudest about the need for tariffs on steel imports. For Nucor points to the collapse of the traditional industry as its proof. Yet Nucor benefits from those tariffs — they can charge higher domestic prices as a result — even while Nucor is in fact the cause of the traditional collapse.
- “not at all” is rhetorical hyperbole, not a factual statement.
Comments Off on The steel industry in North America didn’t die … but it had to re-invent itself
January 28, 2026
Update your NewSpeak dictionaries: “digital twin”
On his Substack, William M Briggs introduces us to a new coat of paint and fresh marketing polish to encourage us to feel so much more comfortable with clankers:
Cracker Barrel infamously tried changing their homey friendly warm and folksy logo to a stripped down dull almost monotone cool version. To remain “current”. They also, reports say, redid the insides of restaurants to emulate modern real estate Soviet-inspired ideas of stripping all detail and turning everything monotonous shades of suicide-inducing gray. They thought this would increase business.
Scientists, grown weary with their dull old ways, and wanting to stay hip — do they still say hip? — decided to redesign their logo, too, as it were. Only they didn’t make the same mistake Crack Barrel did. Instead of hiring some ridiculously over-priced longhoused consulting firm, they asked computer scientists to do the redesign.
Brilliant!
Computer scientists are the firm that brought us neural nets, machine learning, genetic algorithms, and, yes, artificial intelligence, which they cleverly capitalized as “AI”. What’s fantastic is all these evocative names represent the same thing! Models (basically non-linear regressions with some hard coded rules thrown in).
Used to be computer guys would trot out a new name only after they sensed the old one had lost its shine. But “AI” has not. The bubble daily swells. It still tickles imaginations. Which means computer guys hit upon a real innovation: they invented a new name while the current one still shines.
Digital Twin.
What is a Digital Twin? It is, like every new name invented by computer scientists, a model. Only now AI “creates” or “builds” the model. In other words, a Digital Twin is a model of a model.
Where might we find Digital Twins? Here’s some happy-talk hype examples.
Outperform your competition with a comprehensive Digital Twin
Leverage the comprehensive Digital Twin to design, simulate, and optimize products, machines, production, and entire plants in the digital world before taking action in the real world. This helps manufacturers to tackle industry’s biggest challenges: mastering complexity, speeding up processes, and improving sustainability overall.
IBM:
What is a digital twin?
A digital twin is a virtual representation of a physical object or system that uses real-time data to accurately reflect its real-world counterpart’s behavior, performance and conditions.
What is digital-twin technology?
A digital twin is a digital replica of a physical object, person, system, or process, contextualized in a digital version of its environment. Digital twins can help many kinds of organizations simulate real situations and their outcomes, ultimately allowing them to make better decisions.
In other words, models. But how tediously banal is models? Try and sell a model. IBM: “Let us build a model of your system, which might provide useful predictions.” Doesn’t sing. Doesn’t entice. Doesn’t scream premium price. Try this instead: “Be the first to adopt our AI-designed Digital Twin which gives AI insights.” Now you can charge real money.
Digital Twin reeks of excitement. So much so, you just know academics will be getting in on it.
Comments Off on Update your NewSpeak dictionaries: “digital twin”
January 25, 2026
Why does Microsoft treat its users so badly? Because it can
On the social media site formerly known as Twitter, ESR considers the situation most Microsoft users find themselves in these days:
Between the forced updates, the spyware, the adware, OneDrive constantly attempting to suck up all your data, and infinite dark-pattern subscription traps, I’m gathering that many Windows users are now nostalgic for the days when the shit only seemed to be up to their ankles rather than lapping at their nostrils.
The information asymmetry of closed-source software inevitably fucks the user over. I know, I know, I sound like a broken record. I’ve been banging on about this for going on 30 years now and even I get tired of my own rant sometimes.
That’s not why I’m posting today. Instead I want to publicly contemplate an unobvious question: just why is Microsoft treating its users so badly?
“Because it can” is not really a responsive answer. Corporations don’t do evil things because they like being evil, they only do evil things because they think they’re profit-maximizing.
So I understand about the dark-pattern subscription stuff and the adware. That’s slimy, but it’s revenue capture. There’s at least a cold-blooded trade-off you can imagine some product planner making between revenue line-go-up now and pissing off people who won’t be customers later.
But what is Microsoft maximizing by doing things that drive users away from it without any revenue capture? What model of reality, or failure of decision making, do you have to have to think it’s a good idea to push forced updates with work-interrupting reboots that can’t be blocked or delayed by the user?
It would have been trivial to have a pop-up that says: An update is available. Do it now, or defer it until
? The fact that that didn’t happen can’t be ascribed to revenue-line-go-up fever. These are two different kinds of ugly. And that second kind makes me think that there’s nobody left in product management at Microsoft with the ability and authority to veto bad ideas because they will anger the users.
It looks to me like nobody over there is thinking strategically about customer retention anymore. By the time you get to the point where nobody squashes forced-update reboots, nobody can seriously raise the question of whether adware is going to drive away so many users that Windows market share will tank and take all that lovely subscription revenue with it.
This is where I point out, with weary inevitability, that it’s going to get worse before it gets even worse. With nobody keeping an eye on the long game and user retention, the petty money grabs will only accelerate. Microsoft will keep flogging that donkey until it dies.
The irony here is that if Microsoft were an efficient maximizer of long-term profit they would be doing less of the shitty enraging crap that they are now.
How much less depends on how good their judgment is. You can be actively trying to keep a critical mass of your user base happy enough not to bail out and still fail. But at least Microsoft would be trying. Right now, there’s damn little evidence that they are.
Comments Off on Why does Microsoft treat its users so badly? Because it can
January 23, 2026
“Functional illiteracy was once a social diagnosis, not an academic one”
On Substack, Maninder Järleberg illuminates the problem of functional illiteracy in higher education in the west:
The Age of Functional Illiteracy
Functional illiteracy was once a social diagnosis, not an academic one. It referred to those who could technically read but could not follow an argument, sustain attention, or extract meaning from a text. It was never a term one expected to hear applied to universities. And yet it has begun to surface with increasing regularity in conversations among faculty themselves. Literature professors now admit — quietly in offices, more openly in essays — that many students cannot manage the kind of reading their disciplines presuppose. They can recognise words; they cannot inhabit a text.
The evidence is no longer anecdotal. University libraries report historic lows in book borrowing. National literacy assessments show long-term declines in adult reading proficiency. Commentators in The Atlantic, The Chronicle of Higher Education, and The New York Times describe a generation for whom long-form reading has become almost foreign. A Victorian novel, once the ordinary fare of undergraduate study, now requires extraordinary accommodation. Even thirty pages of assigned reading can provoke anxiety, resentment, or open resistance.
It would be dishonest to ignore the role of the digital world in this transformation. Screens reward speed, fragmentation, and perpetual stimulation; sustained attention is neither required nor encouraged. But to lay the blame solely at the feet of technology is a convenient evasion. The crisis of reading within universities is not merely something that has happened to the academy. It is something the academy has, in significant measure, helped to produce.
The erosion of reading was prepared by intellectual shifts within the humanities themselves—shifts that began during the canon wars of the late twentieth century. Those battles were never only about which books should be taught. They were about whether literature possessed inherent value, whether reading required discipline, whether difficulty was formative or oppressive, and whether the humanities existed to shape students or merely to affirm them. In the decades that followed, entire traditions of reading were dismantled with remarkable confidence and astonishing speed.
The result is a moment of institutional irony. The very disciplines charged with preserving literary culture helped undermine the practices that made such culture possible. What we are witnessing now is not simply a failure of students to read, but the delayed consequence of ideas that taught generations of readers to approach texts with suspicion rather than attention, critique rather than encounter.
This essay is part of a larger project to trace that history, to explain how a war over the canon helped usher in an age in which reading itself is slipping from our grasp, and why the consequences of that war are now returning to the academy with unmistakable force.
The Canon Wars: A Short Intellectual History
To understand the present state of literary studies, one must return to the canon wars of the 1980s and 1990s — a conflict that reshaped the humanities with a speed and finality few recognised at the time. Although remembered now as a dispute about which books deserved a place on the syllabus, the canon wars were in truth a contest over the very meaning of literature and the purpose of a humanistic education.
In the decades after the Second World War, the curriculum in most Western universities still rested upon a broadly shared assumption: that certain works possessed enduring value, that they spoke across time, and that an educated person should grapple with them. This conviction, however imperfectly applied, formed the backbone of the humanities. It was also increasingly at odds with a new intellectual climate shaped by post-1968 radicalism, the rise of identity politics, and the importation of French theory.
By the early 1980s, tensions that had simmered beneath the surface erupted into public view. The most emblematic flashpoint came at Stanford University in 1987–88, when student demonstrators chanted, “Hey hey, ho ho, Western Culture’s got to go!” in protest of the university’s required course on Western civilisation. The course was soon dismantled, replaced by a broader, more ideologically framed program. What happened at Stanford quickly reverberated across the country. Departments revised reading lists, restructured curricula, and abandoned long-standing core requirements.
On one side of the debate stood defenders of the canon—figures such as Harold Bloom, Allan Bloom, E.D. Hirsch, and Roger Kimball—who argued that the great works formed a kind of civilisational inheritance. The canon, they insisted, was not a museum of privilege but a record of human striving, imagination, and achievement. On the other side were scholars like Edward Said, Toni Morrison, Henry Louis Gates Jr., Gayatri Spivak, and Homi Bhabha, who contended that the canon reflected histories of exclusion and domination, and that expanding or dismantling it was a moral imperative.
But beneath these arguments lay a deeper philosophical rift. The defenders assumed that literature possessed intrinsic value, that texts could be read for their beauty, their insight, or their power. The critics, armed with concepts drawn from Foucault, Derrida, and Barthes, argued that literature was inseparable from structures of power, that meaning was unstable, and that reading was less an act of discovery than an exposure of hidden ideological operations.
The canon wars ended not with a negotiated peace but with a decisive transformation. The traditional canon was not merely expanded; its authority was dissolved. And with it dissolved a set of shared assumptions about why we read at all.
Comments Off on “Functional illiteracy was once a social diagnosis, not an academic one”
January 19, 2026
Regulating the clankers
At the Foundation for Economic Education, Kevin T. Frazier and Antoine Langrée consider how artificial intelligence can be regulated by state and federal bodies:
Yes, I’m still 12
President Donald Trump’s executive order on artificial intelligence invites analysis of a question so complex that it rarely gets asked: “What exactly do states have the authority to regulate?”
The current, somewhat trite answer is, “The residuary powers reserved under the Tenth Amendment”. Omitting the legalese, that means that states can do whatever the federal government cannot.
States have the power to look out for the health, safety, and welfare of their residents. Thus, for instance, they have the power to address local concerns through zoning laws, professional certifications via licensing regimes, and ensure public safety through law enforcement. These authorities make up what’s often referred to as a state’s “police powers”.
While this generic reading of state power is not necessarily wrong, it’s imprecise. As the AI Litigation Task Force created by Trump’s EO starts its work, a more specific answer is warranted.
The task force is charged with challenging “unconstitutional, preempted, or otherwise unlawful State AI laws that harm innovation”. Reading between these lines, its mission is to contest state laws that interfere with the Administration’s vision for a national AI policy framework. This isn’t an unlimited charge, though. Federal courts reviewing state laws will only strike them down if they fail to align with the Constitution’s allocation of authority or otherwise prove unlawful.
Many stakeholders in AI debates liberally interpret the authorities afforded to states. Based on concerns of existential risk to humanity and the idea that states must protect the health of their citizens, state legislators have proposed and enacted laws that impose significant obligations on the development of AI. Some assume they must have this right, since protecting the lives of their residents is a core priority and unquestioned authority of state governments. After all, since the founding, states have been able to enforce quarantines out of a concern for public health — aren’t aggressive AI laws just extensions of such public health measures, but tailored to the threat of modern threats?
It’s not that simple. States’ police powers are reasonably broad, but not unlimited. States must respect both an upper bound — the purview of enumerated powers reserved for federal authority — and a lower bound—the rights retained by the states’ citizens. These constraints have been tested in litigation throughout our Constitution’s history, notably when state law conflicts with the federal government’s exclusive authority over interstate commerce and when states unduly limit the freedoms of their residents.
These notions are relatively blurry and highly contextual. As national regulatory policy evolves, so too does the extent of preemption. The Lochner era, for example, was a paradigm shift for state police power: as courts expansively interpreted the individual liberty to contract, states’ police power over health, labor protections, and market regulation shrank significantly — only to be restored later. Likewise, individual liberties and valid justifications for their abridgment have evolved to fit developments in civil rights law — from Brown v. Board to Dobbs and Lawrence.
Despite these significant changes in context, the constitutionality of states’ exercise of their police powers follows a bounded framework. This can be observed in the jurisprudence on public health measures — a prime example of police powers. Quarantine orders, from nineteenth-century epidemics to Covid-19, have a direct link to protecting local communities — one of the most important elements of state police powers. They respect the upper and lower bounds of police powers. First, they are geographically specific: they only affect local residents or people coming into local communities. Second, they directly reduce the risk to state residents: quarantines are known solutions to real threats to the health and safety of local communities. They infringe the individual liberties only insofar as is necessary to protect state residents’ vital interests.
Comments Off on Regulating the clankers
January 15, 2026
Pauly/Roux Pistols: The First Self-Contained Cartridges
Forgotten Weapons
Published 22 Aug 2025Samuel Pauly is the largely unrecognized father of the modern self-contained cartridge. In 1808 he patented a cartridge with a metal base that held a priming compound and attached to a paper or metal cartridge body holding powder and projectile. He followed this with an 1812 patent for a gun to fire the cartridges. What makes Pauly’s original system particularly interesting is that he did not use mechanical percussion (ie, hammer or striker) to ignite the primer compound, but rather a “fire pump”. A spring loaded plunger compressed air on top of the primer, heating it enough to detonate the compound in the same way that a diesel engine works. This was not a commercially successful system, though, and Pauly left Paris for London in 1814.
Pauly’s shop was taken over by Henri Roux, who continued making guns under the Pauly name while also improving the cartridges. These two pistols were made around 1820 and use a Roux cartridge with a mechanical striker hitting the primer compound in a Pauly-style cartridge case.
For more information, I recommend Georg Priestel’s free book Jean Samuel Pauly, Henri Roux, and Successors – Their Inventions From 1812 to 1882 available here:
https://aaronnewcomer.com/document/je…
(more…)
Comments Off on Pauly/Roux Pistols: The First Self-Contained Cartridges
