Shellsharks Blogroll - BlogFlock

Finished reading The Primal Hunter - Molly White's activity feed

2026-01-04T00:55:56.000Z

Finished reading:

The Primal Hunter series, book 1.

The Primal Hunter

by Zogarth.

Published 2022. 716 pages.

Started December 6, 2025; completed January 3, 2026.

Turn a list into a web feed - James' Coffee Blog

2026-01-03T18:47:42.000Z

A while ago I made HTML Element of the Day, a site that selects a HTML element every day to be designated the “HTML element of the day.”

When I built this project, I decided that I wanted to offer both a web page and a web feed to which people could subscribe. Rather than build an RSS feed for the project, I decided to mark up the page with microformats. microformats are classes you add to your HTML that have semantic meaning.

HTML Element of the Day is published as a h-feed. Every entry on the website is a h-entry. Here is a rough outline of the markup used:

<ul class="h-feed">
  <li class="h-entry" id="2026-01-03">
    <a class="u-url" href="#2026-01-03">
      <date class="p-published" value="2026-01-03">2026-01-03</date>
    </a>
    <p class="e-content">The HTML element of the day is &lt;input&gt;.</p></li>
  ...
</ul>

This markup lets me declare a list of entries. Each entry has a URL – a HTML fragment on the page in this case – a published date, and some content. You can read the h-feed and h-entry documentation to see what properties are supported.

A web reader that understands microformats like Artemis or Monocle can work with the HTML page as it is and parse its contents to show a list of entries in a user’s web reader:

As a fallback, this page has a rel=alternate tag that uses Granary to convert the h-feed into other another format. Here is an example that takes the page and converts its feed contents to RSS:

<link rel="alternate" href="https://granary.io/url?input=html&output=rss&url=https://random.jamesg.blog/html.php" type="application/rss+xml">

This rel=alternate tag ensures any reader that understands RSS can read the page.

Tantek uses this pattern for his daylists page which lists the names of daylists that have been generated for him on Spotify.

In both the HTML Element of the Day and Tantek’s daylists example, h-feed is being used to turn a HTML list into a web feed. Indeed, there is nothing to say that a web feed must be a list of blog posts or notes.

Personal and community web search engines - James' Coffee Blog

2026-01-03T18:33:02.000Z

Ever since I made the first version of IndieWeb Search, a search engine that let you search personal websites, I have been thinking about the idea of “community search engines.”

A community search engine indexes materials related to a specific community. The information sources are hand-picked to reflect the interests of the community. This may include blogs, feeds, social media firehoses, and more. The goal of a community search engine is not to index the web – rather, to index a slice of web pages and other resources that are relevant to the group.

The successor to IndieWeb Search [1], IndieWeb Discover, follows feeds published by several dozen personal websites that write about or have written about the indie web. IndieWeb Discover, rather than sorting results by relevance, sorts results in reverse chronological order, allowing you to discover new blog posts on a given topic. It is, in a way, a community search engine, but also a community discovery engine – a place where people can go to discover blog posts on a given topic they may not otherwise have found.

Here is an example of an IndieWeb Discover page for the query “indieweb carnival”:

IndieWeb Discover is an experiment, one that I have had running in the background for a while but haven’t really announced or done anything with. I have found interesting blog posts through it, so I thought I would finally take the time to share the project on this blog.

Personal web search engine

In addition to a community-oriented search engine, a personal search engine could exist that indexes web pages of interest to you.

My version of a personal search engine is the search provided by Artemis, the web reader I use and maintain. This search feature returns web pages published in the web feeds followed by a user. In my case, Artemis shows results from sites I follow like blogs and art news and more. With my web reader’s search feature, I often find blog posts from authors I love. My search results will not be relevant to you – being mainly personal blogs I enjoy – but your own search engine could be relevant specifically for you.

Here is an example of a search result for the query “Taylor Swift” in my personal Artemis search feature:

The topic of information retrieval fascinates me perhaps because of the consequence of the fundamental question it concerns: How can and do we find information? Following on from this: How can we find information more effectively? (And, indeed, what does ”effectively” mean?)

I am going to keep thinking about this topic and I'd love to read more about it! If you can recommend any blog posts – or feel inspired and share your thoughts on your website – please send me an email!

[1]: The original IndieWeb Search used a complex technology stack, relying on Elasticsearch. This made the tool difficult to maintain. The successor, IndieWeb Discover, relies on a Sqlite database. While not offering full text search with keyword relevance-based ranking that IndieWeb Search offered, Discover has been fun to use!

How Do You Read My Content? - Kev Quirk

2026-01-03T13:35:00.000Z

I'm trying to get an idea on how people consume the waffle I put out, it should only take 5 seconds to respond, and I'd be very grateful.

It’s well publicised that I don’t run any kind of analytics on this site. For me, engagement is far more important. But I’m trying to better understand how you fine people consume the waffle I spit out into the world.

The only reason I want to do this is that I think it will be interesting to know. I could temporarily add tracking to the site, but that feels icky to me; I’d rather have something that’s opt in. So I’ve created a really simple form that you can fill in. It only has 1 question, so should take no more than a few seconds to complete.

If you’re a regular reader, I’d be very grateful if you could take a few seconds out of your day to cast a vote please. The form is embedded below, but it may not embed properly in some places (like on the RSS feed), so just in case here’s a direct link to the form too.

Thanks for reading this post via RSS. RSS is great, and you're great for using it. ❤️

You can reply to this post by email, or leave a comment.

Where is Bitcoin? - Terence Eden’s Blog

2026-01-03T12:34:09.000Z

Happy Birthday Bitcoin! At the risk of awakening long-dormant beasts, it looks like Bitcoin has failed for day-to-day transactions. So I've a simple question to ask - can you meaningfully spend any cryptocurrency in your city centre?

A few months ago, my wife and I went on a 30 day Interrail holiday across Europe. 10 countries, over a dozen cities, making over a hundred payments.

I looked in the window of every cafe, restaurant, corner shop, pub, bar, museum, and tourist-trap that we went into. All of them accepted a wide range of credit cards. Some politely informed us that they were either cash-only or no-cash. A few gleefully accepted foreign currencies at ruinous exchange rates.

None advertised accepting crypto.

Due to the language barrier and the risk of looking like a bellend, I didn't actually ask anyone if I could pay with crypto. But every bill we were handed told me which (fiat) currencies I could pay in. The tip-jars let me tap my card or drop my (metal) coins. There were signs warning that I couldn't use American Express.

But, again, no mention of crypto.

I literally only saw one sign that cryptocurrencies existed. Tucked away in Zagreb train station, next to the left-luggage lockers, was this:

I've never used a crypto ATM before. I've no idea if it's safe or whether it'd give me a good deal. I also didn't know whether a guy with a crowbar would follow me out of the station. But, hey, YOLO!

There was no queue of people waiting to use it, so I thought I'd investigate.

Delightfully, it wasn't just Bitcoin that it offered:

Including two subtly different stable coins. Which is nice, I guess.

I clicked through a screen confirming that I wasn't a Politically Exposed Person and got hit by this screen:

I've never had a cash ATM warn me about those kind of scams. Progress!

And yet, after wandering Zagreb for a few days, I didn't find anywhere to spend whatever crypto I could've purchased from there. I went to laundrettes, museums, galleries, bougie little eateries, and had dark kitchen deliveries. None advertised their acceptance of crypto.

Perhaps I'm not cool enough to go to the hip underground club where all the cocktails are only purchasable with an obscure shitcoin. Maybe there's an exclusive casino where all the beautiful people spend their evenings playing baccarat chemin de fer for tokens?

Or perhaps I'm just unobservant.

In the last few years I've wandered through the twisty passages of the Marrakech Medina and pounded the mean streets of Halifax, Nova Scotia. I've drunk in the bars of Kuala Lumpur and slurped food-court noodles in Hong Kong. From Sydney to London, I'm not exaggerating when I say that no one has ever offered to let me settle my bill with a cryptocurrency.

I've paid by mobile, in-app, using loyalty card points, instant bank transfer, and fairground tickets. No crypto.

In London, my home city, there are maybe 50 places which say they accept Bitcoin. London's population is about 10 million. It is one of the most visited cities in the world. There is, effectively, nowhere to buy a coffee using the blockchain.

Perhaps crypto is more like gold? I don't know any shop which would let me pay with an ingot or doubloon. Yet plenty of people have gold as part of their investment portfolio. Although gold is a physically and chemically useful metal whereas the utility of crypto is yet to be demonstrated.

It's the same with shares. I can't go to a pub and buy a round with GME:NYQ. I need to convert it into something more liquid before I can enjoy something more liquid.

So crypto can be an opaque speculative investment. Maybe it is the next tulips, maybe it is the next Ford Model T. But it is obviously so far from infiltrating normal payments that it can be safely ignored.

I've been writing about Bitcoin since 2011. I even studied for the Certified Blockchain Professional qualification and wrote about it for my MSc a couple of years ago.

I'm just so tired of all the unjustified hype. I despair about the amount of energy that's been wasted - both electrical and mental.

It has been seventeen long years since Dr Craig Wright published the Bitcoin paper. How much longer are people going to keep pretending that this is a transformative payment technology ready to be embraced by the masses?

Think I'm wrong. Don't bother arguing in the comments. Instead, you can put your money where your mouth is by sending Bitcoin to me at bc1q60rt4rrnky7syxdw6234zqnmy84ah2gphra5ev 😂

1929 - By Andrew Ross Sorkin - Westenberg

2026-01-03T10:00:19.000Z

💡

ISBN: 9780593296967
Date finished: 2026-01-02
How strongly I recommend it: 8/10

A terrifying // granular look at the hubris that melted the global economy. Sorkin doesn't paint the bankers as cartoon villains, so much as delusional optimists who got high on their own supply of cheap credit. Read this to understand why "this time is different" is the most dangerous phrase in human history. Essential for anyone who thinks markets are rational. For more: Galbraith's.

Notes

"How did you go bankrupt?" "Two ways. Gradually, then suddenly." That is how confidence, the lifeblood of the economy, disappears.

The greatest product of the 1920s was not the car or the radio, it was credit. It drew the wealth of tomorrow into the present. Problems arise when we get greedy and take too much.

Lengthy, uninterrupted booms produce a collective delusion where optimism becomes a drug. People lose the ability to calculate risk because the future looks like a land of ever-expanding opportunity.

Human history becomes more and more a race between education and catastrophe.

J. Pierpont Morgan believed character was more important than collateral. A man with character but no money could get a loan; a man with property but no character could not. That (he claimed, and possibly even believed) was the rule of business. But who determines character? And when you're misled - what then?

In 1929, Charles Mitchell, head of the largest bank in America, publicly defied the Federal Reserve's order to curb speculation. He argued his obligation to avert a crisis was paramount to any Fed warning and single-handedly pumped $25 million of credit into the market to keep the party going.

Groucho Marx lost a fortune because he listened to a broker who claimed Wall Street was no longer "localised" but a "worldwide market" where prices would rise forever. When the crash hit, the broker simply said, "I guess I made a mistake."

By 1929, women made up 20 per cent of market speculators. Brokerages installed special lounges for them, while critics claimed they were temperamentally unsuited for trading and were "going crazy and mortgaging their homes."

Jesse Livermore's trading rule: Profits take care of themselves, but losses never do. You must insure yourself against big losses by taking the first small loss immediately.

Livermore believed the market is a mirror of the underlying psychology. He distrusted inside information and professors, arguing that if they haven't traded on margin, they know nothing about speculation.

Winston Churchill was a terrible investor who treated the stock market like a casino. He had to be bailed out by financiers and book deals to avoid bankruptcy.

Herbert Hoover changed the name for the economic slowdown from a "panic" to a "depression" to assuage the public. It was a terrible rebranding: panics are short; depressions can last a damned long time.

No specific event produced the debacle. A sufficient number of people simply believed the hour to sell had come, starting a wave of fear that the machinery of the exchange couldn't withstand.

Charles Mitchell was so confident in his bank's stock that he personally borrowed $12 million to buy shares during the crash to prop up the price. He ended up millions in debt and indicted for tax evasion.

Many people "achieve self-approval by cheap and superficial methods, concocting clever excuses to make what we want to do seem admirable so that we can condone our actions and like ourselves."

The "White Knight" of Wall Street, Richard Whitney, was actually pilfering millions from the Stock Exchange's gratuity fund and the New York Yacht Club to cover his bad debts. The Morgan partners knew and covered it up to avoid a scandal.

The crash wasn't caused by the withdrawal of Federal Reserve credit, but it also wasn't helped by the avalanche of "bootleg" credit: loans from corporations and foreign entities that the Fed couldn't control.

Carter Glass, the father of banking reform, initially opposed federal deposit insurance because he believed it rewarded reckless banks and created moral hazard. He had to be browbeaten into accepting it.

After losing his fortune, automobile pioneer William Durant opened his old bank book and declared he was still the richest man in America, "in friends." He died broke.

The story of 1929 is about human nature, not regulation or rates. AKA we will always find new ways to believe the good times can last forever, to mistake hope for certainty.

The antidote to irrational exuberance is not regulation or skepticism, but humility. We must know that no system is foolproof and no generation is exempt.

Notable links: January 2, 2026 - Werd I/O

2026-01-02T14:21:28.000Z

Every Friday, I share a handful of pieces that caught my eye at the intersection of technology, media, and society.

We're mostly looking ahead this week: human intuition and creativity are predicted to once again become front and center in tech, and Meta looks like it's betting on being the arbiter of what is real. What could go wrong? Also, in completely unrelated news, some thoughts on what happens when news and information organizations capitulate to power.

Did I miss something important? Send me an email to let me know.

The next big thing in 2026 will be...

There’s a lot of interesting food for thought in this roundup of predictions for 2026 from investors, founders, and technology leaders. AI is dominant here — most people predict it will continue to be transformative at all levels of tech, which feels like a safe bet — but there’s another theme that I think is worth drawing out.

ClassDojo CEO Sam Chaudhary suggests:

“The atomic unit of execution inside companies collapses to a one-person squad; a ‘full-stack operator.’ As AI reduces the cost of execution to near zero, the right person can out-build a 50-person org. The bottleneck becomes finding full-stack operators with curiosity, taste, and judgment.”

To emphasize: the important thing in this prediction is finding people who can operate across the stack; their curiosity, taste, and judgment become pivotal.

Rebecca Kaden, General Partner at USV, suggests:

“The next big thing in 2026 will be a fresh boom of consumer platforms. Boredom with older platforms + curiosity at what new tools can create + continued desire for connection and the new will usher this in. After a long stretch of it being too hard to interrupt incumbents, the cracks are showing and the newer, the weirder the better as we crave new consumer digital experiences and connection.”

And I think most importantly, Homebrew Partner Hunter Walk adds:

“The next big thing in 2026 will be Kindness. Apolitical, nondenominational, online and offline kindness. Mainstream tech population is exhausted by the escalations of the last decade. Begins investing time and energy in more community, civics, and charity. Leaves the most toxic social spaces to trolls and bots.”

In other words, it’s going to be the squishy human stuff that is going to become central to tech. In many ways it really always has — tech has always been a people business at its heart — but putting it at the center of the discourse is a change. And in a world where AI is becoming more and more prevalent, particularly in transforming how code is written and deployed, it makes sense that it would be.

I do want to add a note about “taste”: I’ve always hated that term for cultural judgments. Taste is subjective, not just from person to person but from culture to culture. When we talk about it, and particularly when we evaluate it, I think there’s a real danger that we do so from one (culturally dominant) demographic and context. I’d like to argue that the only sense of taste that really matters is if you make something that resonates for someone else. If we try and make someone’s work fit into a broader, shared “taste” aesthetic, we run the risk of making it bland at best, and whitewashed at worst.

But I’m pro-human. To the extent that these predictions put human experience, communities, and values first and foremost, I’m excited for the future they describe.

2025: The year in LLMs

Simon Willison has become an expert in LLMs, their technical capabilities, and how they’re transforming software development in particular. So his end-of-year roundup is a must-read if you want to understand the space.

For what it’s worth, I agree with him here:

“The most impactful event of 2025 happened in February, with the quiet release of Claude Code.”

It has the potential to transform all of tech. I also think we’re going to see a real split in the tech industry (and everywhere code is written) between people who are outcome-driven and are excited to get to the part where they can test their work with users faster, and people who are process-driven and get their meaning from the engineering itself and are upset about having that taken away. (This is not to say that there aren’t many issues with AI aside from these things, of course.)

It wouldn’t surprise me to see more artisanal teams, startups, and small businesses spring up to give that second set a home. But I think we’ll also see more startups and projects created by the first set, too.

It really is a different way of writing software:

“Claude Code for web is what I call an asynchronous coding agent—a system you can prompt and forget, and it will work away on the problem and file a Pull Request once it’s done.

[…] I love the asynchronous coding agent category. They’re a great answer to the security challenges of running arbitrary code execution on a personal laptop and it’s really fun being able to fire off multiple tasks at once—often from my phone—and get decent results a few minutes later.”

There’s a lot more here besides. If you want to get up to speed, or see what you missed, it’s a great place to start.

What is Instagram’s Adam Mosseri really saying in his year-end memo?

Om Malik has some smart observations about Instagram’s messaging at the end of a year that was dominated by AI-generated content:

“AI is flooding the system, and feeds are filling with fakes. Visual cues are no longer reliable. Platforms will verify identities, trace media provenance, and rank by credibility and originality, not just engagement.

It starts by verifying who is behind an account, embedding provenance in media, and rewarding trust signals. Over time, Meta may tighten control and aim to be an identity broker for everyone. Instagrams want you to be prepared for this new era of tighter control over identity, authenticity, and content provenance.“

This isn’t just about AI and advisers preferring to be associated with real content: we’re starting to see age verification laws take hold in various jurisdictions, and there are likely more to come. By preparing users for more identity verification and tying it up with a “this way we know who’s real” bow, they’re able to get ahead of these rules. And being the primary identity broker in the social space during this new era will provide some business security.

I don’t think a move to verified identities is good. Real names don’t improve online community health, and attaching identities to the things people post will have a chilling effect on activists and people from vulnerable communities in particular. It’s also not true that social media is bad for younger teenagers, contrary to popular perception of the research. It’s a trend that leads to more surveillance and less open speech. But if it’s going to happen, Meta, I’m sure, will be delighted to turn it into a moat and a profit center if it can.

Watching Bari Weiss Murder Investigative Journalism at CBS

Spencer Ackerman has had a 60 Minutes piece spiked and has withstood White House demands to stop a story. And he has some thoughts about Bari Weiss spiking the 60 Minutes story about CECOT.

“That's where my mind goes when I read about Weiss "delaying" the 60 Minutes CECOT piece. To what it would have meant for The Guardian leadership to cave on the Snowden stories. What it meant for the Times leadership to have self-censored its reporting on NSA bulk surveillance and CIA torture. The lesson to any journalist attempting to reveal the highest-stakes stories is that you must fight your outlet if you're going to do the work your audience needs to understand the situation it's in.”

Ackerman takes great pains to indicate that CBS was never the bastion of independent journalism that commentators with rose-tinted glasses might have suggested: it had a cozy relationship with the CIA and was indisputably an instrument of power. Still, what Bari Weiss is doing with it rises to some other level: a mouthpiece for right-wing talking points.

It’s also worth saying that this matters less and less. Broadcast news has been slowly diminishing for a generation. CBS, in other words, is not where most people are getting their news. That’s not to say that it couldn’t be again, if it transformed itself into an authentic, trustworthy voice that meets audiences where they’re at today. Which is the opposite of that’s happening; and regardless, it doesn’t make what’s happening any less egregious.

EchoFeed Profiles - Robb Knight • Posts • Atom Feed

2026-01-02T13:48:18.000Z

I'll say this up front: I don't want EchoFeed to be a social network. There are many platforms better suited to that and EchoFeed isn't it. Having said that, I do think there's something nice about being able to link to a profile that says "Yeah I use this thing, I like it, look at a photo of my face and stuff" so I made profile pages that can have a photo of your face and stuff.

There's not much here, by design. Go to the profile editor in the dashboard from the dropdown in the top right and you can set your username, bio, website, and choose a colour for your profile (and your open graph image). Your profile will show if you're an EchoFeed Pro subscriber, how many echoes you have, and how many posts you've made. It will never expose what feeds you have enabled or posts you've made — I may add something related to this in the future but it will be off by default^[1]. You can turn off your profile at any time by unticking "Profile Enabled".

Speaking of open graph images, to get this out quickly I used simonhamp/the-og which is a "An OpenGraph image generator written purely in PHP". It pretty simple to use and had a built-in layout that was mostly what I wanted already. I will update these in the future to have more information but for now, they do the job nicely.

I'm seeing a couple of instances where opengraph images won't load I think because of caching (which I need for the stampede problem) so if you see something off, let me know. For now, you can check out my profile at echofeed.app/@robb.

Tangentially related to this but I'm moving all EchoFeed announcements to here instead of maintaining a separate blog which is overkill. I'll setup redirects and import the existing posts to here in the next few days.

hey tech bros, this is called "consent". ⤾

The Case for Blogging in the Ruins - Westenberg

2026-01-01T22:56:52.000Z

In 1751, Denis Diderot began publishing his Encyclopédie, a project that would eventually span 28 volumes and take more than two decades to complete. The French government banned it twice. The Catholic Church condemned it, Diderot's collaborators abandoned him, his publisher secretly censored entries behind his back, and he worked himself into periodic breakdowns trying to finish the damn thing.

When people talk about the Enlightenment as if it were an intellectual garden party where everyone sipped wine and agreed about reason, they're missing the part where producing and distributing ideas was (in fact) dangerous and thankless work.

Diderot has been on my mind lately, spending the Xmas period scrolling through the dwindling numbers of social media platforms that haven't yet been purchased by an "eccentric" (read: race-science obsessed) billionaire or banned by a foreign government.

Diderot's project was fundamentally about building infrastructure for thinking. He wanted to create a shared repository of human knowledge that anyone could access, organized in a way that invited exploration and cross-referencing. He believed that structuring information properly could change how people thought.

He was right.

270 years later, we have more information than any civilization in history. But aside from Wikipedia, we've organized the sum total of our collective knowledge into formats optimized for making people angry at strangers in pursuit of private profitability.

Something has gone terribly wrong.

And I think the fix, or at least part of it = going backwards to a technology we've largely abandoned: the blog, humble // archaic as it may seem.

The Pamphlet Problem

Before social media ate the internet, and before the internet ate everything else, and before everything else ate itself, blogs occupied a wonderful and formative niche in the information ecosystem. They were personal but public, permanent but updateable, long-form but informal. A blog post could be three paragraphs or thirty pages. It could be rigorously researched or entirely speculative. It could build an argument over weeks or months, with each post serving as a chapter in an ongoing intellectual project that readers could follow, critique, and respond to.

The blogosphere of the mid-2000s had its problems: It was insular and often smug, prone to flame wars between people who agreed on 95% of everything but found the remaining 5% absolutely unforgivable. But it also produced actual intellectual communities.

Remember those?

People wrote long responses to each other's posts, those responses generated further responses, and you could follow the thread of an argument across multiple sites and weeks of discussion. The format rewarded careful thinking because careful thinking was legible in a way that it simply isn't on platforms designed for rapid-fire engagement.

Social media removed the friction of publishing, and in doing so removed the selection pressure that separated signal from noise. We "democratized" the ability to publish (good?) while simultaneously destroying the conditions that made publishing meaningful (bad!).

Diderot spent twenty years on his infrastructure.

We handed ours to advertising companies and - like Pilate - washed our hands of it.

The Architecture of Attention

When you write a blog post, you're creating a standalone document with a permanent URL. It exists at a specific address on the web, and that address doesn't change based on who's looking at it, when they're looking at it, or what algorithm has decided they should see next. The post is there, stable, waiting for whoever wants to find it.

Compare this to a tweet (by God I'll not call them "X's") or a Facebook post, which exists primarily as an item in a feed, algorithmically sorted, personalized to each viewer. Your post might appear at the top of someone's feed for an hour and then disappear into an infinite scroll of other content, never to be seen again. The platform has no interest in whether your post is found next week or next year; it has a vested interest in keeping users scrolling through new content right now.

When I write a blog post, I'm writing for an imagined reader who has arrived at this specific URL because they're interested in this specific topic; I can assume a baseline of engagement; I can make my case over several thousand words, trusting that anyone who's made it to paragraph twelve probably intends to make it to paragraph twenty.

When I write for social media, I'm writing for someone who is one thumb-flick away from a video of either a hate crime or a dog riding a skateboard. Everything I produce has to compete, in real-time, with everything else that could possibly occupy that user's attention. The incentives push toward provocation and emotional activation. The format actively punishes nuance, which means that a thoughtful caveat reads as weakness and any acknowledgment of uncertainty looks like waffling.

Diderot understood that the container shapes the contents. The Encyclopédie was a collection of facts, yes, but more fundamentally it was an argument about how knowledge should be organized. Cross-references between entries were themselves a form of commentary, connecting ideas that authorities wanted kept separate. We've restructured the presentation of ideas around the needs of advertising platforms, and not by accident, and we're living with the consequences.

Montaigne's Heirs

Michel de Montaigne arguably invented the essay in the 1570s, sitting in a tower in his French château, writing about whatever interested him: cannibals, thumbs, the education of children, how to talk to people who are dying. He called these writings essais, meaning "attempts" or "tries." The form was explicitly provisional. Montaigne was trying out ideas, seeing where they led, acknowledging uncertainty as a fundamental feature rather than a bug to be eliminated.

The blog, at its best (a best I aspire one day to reach) is Montaigne's direct descendant. It's a form that allows for intellectual exploration without demanding premature certainty. You can write a post working through an idea, acknowledge in the post itself that you're not sure where you'll end up, and invite readers to think alongside you. You can return to the topic weeks later with updated thoughts. The format accommodates the actual texture of thinking, which is messy and recursive and full of wrong turns.

Social media flattens all of this into statements: Everything you post is implicitly a declaration. Even if you add caveats, the format strips them away. What travels is the hot take, the dunked-on screenshot, the increasingly-shitty meme, the version of your argument that fits in a shareable image with the source cropped out.

I keep thinking about how many interesting folks have essentially stopped writing anything substantial because they've moved their entire intellectual presence to Twitter or Substack Notes. These are people who used to produce ten-thousand-word explorations of complex topics, and now they produce dozens of disconnected fragments per day, each one optimized for immediate engagement and none of them building toward anything coherent.

It's like watching someone who used to compose symphonies decide to only produce ringtones.

What Makes a Blog Actually Work

Most blogs are abandoned after three posts. The ones that persist and accumulate value share certain features worth naming:

They have a perspective, not just a topic. A blog about "marketing" is forgettable. A blog about "why most marketing advice is wrong and what actually works" has a point of view that gives every post something to push against.
They build. The best blogs create posts that reference and extend earlier posts, developing ideas over time rather than starting from scratch each week. Gwern's site is an extreme example, with entries that get updated for years, accumulating evidence and refinement. But even a modest version of this works: a body of work that compounds.
They're written for someone specific. Not "everyone interested in X" but "the person who already knows Y and is trying to figure out Z." Specificity creates resonance.
They exist at a consistent address. A blog at your own domain, with permanent URLs, can be found and referenced for years. Content locked inside platforms disappears when the platform changes or dies.
Their authors accept that most posts won't go viral, and that's fine. The value is cumulative. A blog with fifty solid posts is an asset even if no single post ever breaks through. Most social media content has a half-life of hours; a good blog post can draw readers for a decade.

The Discovery Problem (And Why It's Overstated)

There are rote objections: nobody reads blogs anymore. The discovery mechanisms are broken. How is anyone supposed to find a new blog when they're competing against algorithmic feeds specifically designed to capture and hold attention?

Etc, etc, etc.

But there are a few things worth noting:

Search engines still index blogs far better than social media posts. A well-written blog post on a specific topic can draw readers for years through Google (or Kagi // DuckDuckGo if you're nasty, and by nasty I mean excellent); a tweet is lucky to get attention for twelve hours. Hell, call it six. Hell, call it three and call me an optimist at that. If you're trying to build a body of work, or to create something that will outlast the platform of the moment, a blog is simply a better tool.

What else?

RSS never actually died. It went underground. Feedly, Unread, NetNewsWire, and other readers still have millions of users. The people who read blogs tend to be the people worth reaching: curious, patient, willing to engage with longer arguments.

Newsletters are still a discovery layer, no matter how many people pronounce their untimely death. You can write on your own site and distribute via email, getting the permanence of a blog with the push distribution of a newsletter. The writing lives at your domain; the email is notification infrastructure.

And the fragmentation of social media is actually creating demand for alternatives. Every time a platform implodes (Twitter's ongoing collapse, Instagram's slow retirement // decay into a metaphorical Floridian condo, TikTok's uncertain status, Facebook's demographic hollowing) people start looking for more stable ground. The infrastructure exists. It's waiting.

A Room of Your Own

Virginia Woolf wrote about the importance of having a room of one's own: physical space for creative work, free from interruption and control. A blog is a room of your own on the internet. It's a place where you decide what to write about and how to write about it, where you're not subject to the algorithmic whims of platforms that profit from your engagement regardless of whether that engagement makes you or anyone else nebulously smarter.

Diderot built the Encyclopédie because he believed that organizing knowledge properly could change how people thought. He spent two decades on it. He went broke. He watched collaborators quit and authorities try to destroy his work. He kept going because the infrastructure mattered, because how we structure the presentation of ideas affects the ideas themselves.

We're not going to get a better internet by waiting for platforms to become less extractive. We build it by building it. By maintaining our own spaces, linking to each other, creating the interconnected web of independent sites that the blogosphere once was and could be again.

So:

Start a blog. Start one because the practice of writing at length, for an audience you respect, about things that matter to you, is itself valuable. Start one because owning your own platform is a form of independence that becomes more important as centralized platforms become less trustworthy. Start one because the format shapes the thought, and this format is good for thinking.

The blog won't save us. But it's one of the tools we'll need if we're going to save ourselves.

Start Today

If you're convinced, the practical options:

Write.as is minimalist, privacy-focused, no-frills. You can start anonymously and upgrade to a custom domain later. The editor gets out of your way. Good for people who want to write without fiddling with settings.

Bear Blog is extremely lightweight, fast-loading, no tracking. Free tier is generous. The aesthetic is deliberately simple, which enforces a focus on writing over design tinkering. Privacy-first.

Ghost (powering this blog) is more fully-featured, with built-in membership and newsletter tools. You can self-host (free) or use their managed hosting (paid). Good if you want to eventually build a paid subscriber base but want to own your infrastructure. Open source.

Micro.blog was built explicitly as an alternative to social media, with cross-posting, a community timeline, and support for short and long posts. Has an indie web ethos baked in. Good if you want the social layer without the algorithmic manipulation. Manton (founder) is one of my favourite folks to follow // read lately. Good insights.

All of these let you use a custom domain, which you should do. Buy yourname.com. It costs ten dollars a year and your writing will live at an address you control, regardless of what happens to any particular platform.

Pick one. Set it up this week. Write something. Send me a link (email joan@joanwestenberg.com) and I'll read it. The first post doesn't have to be good; it has to exist. The second one can be better. That's how this works. That's how it's always worked.

What is Instagram’s Adam Mosseri really saying in his year-end memo? - Werd I/O

2026-01-01T21:44:32.000Z

[Om Malik]

Om has some smart observations about Instagram’s messaging at the end of a year that was dominated by AI-generated content:

“AI is flooding the system, and feeds are filling with fakes. Visual cues are no longer reliable. Platforms will verify identities, trace media provenance, and rank by credibility and originality, not just engagement.

It starts by verifying who is behind an account, embedding provenance in media, and rewarding trust signals. Over time, Meta may tighten control and aim to be an identity broker for everyone. Instagrams want you to be prepared for this new era of tighter control over identity, authenticity, and content provenance.“

[Link]

Note published on January 1, 2026 at 5:13 PM UTC - Molly White's activity feed

2026-01-01T17:13:05.000Z

grateful for the books that made 2025 a little less hard

Last book: James S.A. Corey’s Abaddon’s Gate
Longest book: Fyodor Dostoevsky’s Crime and Punishment
Shortest book: Martha Wells’ Artificial Condition
Books read: 119
Pages: 50,481
Top genres: Fantasy (41), science fiction (38), mystery (37), thriller (32), crime (29)
Top authors: John Sandford, Martha Wells, Matt Dinniman
5-star reads: Becky Chambers’ The Long Way to a Small, Angry Planet; Matt Dinniman’s Dungeon Crawler Carl; John Steinbeck’s East of Eden
Average rating: 3.56 stars" />

The new year - James' Coffee Blog

2026-01-01T16:31:31.000Z

Watching as condensation drips from the window panes – the inside being so warm, and the outdoors being almost freezing – I notice the pink sky which signals that, soon, the sun will set. My first instinct when seeing the pink sky – the colour of the sky only intensifying as the minutes pass – was how much the colours of the sky means to me, not just this year but in general.

At sun rise and sunset, the boundaries are gradients: the sky changes between many colours. There are magical moments in between.

Last year, I wrote down a few “goals” I had for the year. I didn’t think of them as “goals,” naming them so only because I couldn’t think of a better word. Perhaps the word for which I was looking was “directions.” I wanted to write down what I felt at the start of the year to serve as a direction. I referred to the list a few times especially as 2025 came to a close, not as a “to do list” but more so as a guide.

I have written a similar list for this year, too, writing down a few directions for myself for the new year. I don’t think of these as resolutions, but a little something that I can look at to remember this moment – this time of new year. At the top of the list is to be as kind to myself as I can – oh anxiety! – followed then by completing my first year in university. I am so excited to start university.

Like the sunrise and sunset, I have changed and almost always am changing, even if in subtle ways. Although unlike the sunrise and sunset, change for me begins with noticing, both the small and the big things. Oh, my coffee is a little bit bitter today. I wonder what I can do to make it better tomorrow // I wonder if I could play a chord on the guitar. // I think I need to challenge myself more.

Having a few directions for last year served as a guide, but left room for all the moments in the middle – those times when I think oh! and have an inclination to try something new, or to explore new paths and build conviction in what I want to do next. Indeed, many of the things I want to do have actually been in my mind for a while – to learn German, for example – and that started last year. Change can happen at any time.

To keep going is another matter. And so, in addition to starting on a big new adventure – university – many of my directions this year are to keep going with those things I enjoy; to learn more.

Often I don't see the start of the sunrise and the sunset: I notice the change after it has already begun. Change takes time. As do many of the best things.

The sky outside is now a beautiful blue, a colour that both evokes memories of the last year and further back, and calls to mind all of the colours to be seen this year.

Wherever you are in the world, I wish you the happiest new year!

2025: The year in LLMs - Werd I/O

2026-01-01T15:38:15.000Z

[Simon Willison]

For what it’s worth, I agree with him here:

“The most impactful event of 2025 happened in February, with the quiet release of Claude Code.”

It really is a different way of writing software:

“Claude Code for web is what I call an asynchronous coding agent—a system you can prompt and forget, and it will work away on the problem and file a Pull Request once it’s done.

[…] I love the asynchronous coding agent category. They’re a great answer to the security challenges of running arbitrary code execution on a personal laptop and it’s really fun being able to fire off multiple tasks at once—often from my phone—and get decent results a few minutes later.”

There’s a lot more here besides. If you want to get up to speed, or see what you missed, it’s a great place to start.

[Link]

The next big thing in 2026 will be... - Werd I/O

2026-01-01T15:28:54.000Z

[Nikhil Basu Trivedi at Next Big Thing]

ClassDojo CEO Sam Chaudhary suggests:

“The atomic unit of execution inside companies collapses to a one-person squad; a ‘full-stack operator.’ As AI reduces the cost of execution to near zero, the right person can out-build a 50-person org. The bottleneck becomes finding full-stack operators with curiosity, taste, and judgment.”

To emphasize: the important thing in this prediction is finding people who can operate across the stack; their curiosity, taste, and judgment become pivotal.

Rebecca Kaden, General Partner at USV, suggests:

“The next big thing in 2026 will be a fresh boom of consumer platforms. Boredom with older platforms + curiosity at what new tools can create + continued desire for connection and the new will usher this in. After a long stretch of it being too hard to interrupt incumbents, the cracks are showing and the newer, the weirder the better as we crave new consumer digital experiences and connection.”

And I think most importantly, Homebrew Partner Hunter Walk adds:

“The next big thing in 2026 will be Kindness. Apolitical, nondenominational, online and offline kindness. Mainstream tech population is exhausted by the escalations of the last decade. Begins investing time and energy in more community, civics, and charity. Leaves the most toxic social spaces to trolls and bots.”

But I’m pro-human. To the extent that these predictions put human experience, communities, and values first and foremost, I’m excited for the future they describe.

[Link]

200 Years Ago - A Prediction of a Deadly Comet Impact - Terence Eden’s Blog

2026-01-01T12:34:22.000Z

While digging though some old journals in a fruitless side-quest, I came across this delightful description of what I think is the Comet Encke.

It is quite an astonishing prediction, and the last line is perfection.

In 1926, several journals and almanacs syndicated a column discussing this comet. The above is from The New Jerusalem magazine and theological inspector which has added "This is very alarming!".

Most of the other reprints leave off this last line - but The Christian Advocate" adds "But who expects that the earth will endure four millions of years!"

Only a few thousand more years until we will know if the prediction was accurate!

Weeknote #1981 - Robb Knight • Posts • Atom Feed

2025-12-31T22:01:35.000Z

Final day of the year, final weeknote of 2025.

The sticker shop is live as of a couple of weeks ago with a handful of stickers I think you'll like.

I'm into a new journal for tomorrow and this time I'm using a Leuchtturm1917 Bullet Journal I got for a steep discount. The paper isn't as thick as the STM paper I used last year but it's certainly nice to write on. I definitely have some more thoughts about the Bullet Journal brand but that's for another time.

I am getting so many ads on YouTube for fucking AI generation, AI business leader training, and it's all shit. Stop it. Just stop.

KFC have a merch shop and I want this hat. I don't need it but also, KFC hat.

The song "Louie Louie" has been covered upwards of 1600 times. "Three chords and the most mundane beat possible. Any idiot could learn it, and they all did." is a very funny description of the song. I just went to get a link for the song, having read the wikipedia page not actually knowing what the was, only to realise I do know it. I'm assuming everyone does.

Malcolm in the Middle is back and I am here for it.

Superman by Goldfinger went gold earlier this year and they performed it with Tony Hawk on vocals at the Warped tour. As I always do when I mention this song, check out Mike Herrera's acoustic version and his version of No Cigar while we're at it.

Spigen made an entirely unnecessary Classic Mac case for the Mac Mini and I want it.

Mel put together now-updater which automatically updates your now page. Imagine this Rube Goldberg machine I built but it's easy to use and better in every way. It's that.

Game Font Library is a new site listing fonts from games, 259 of them as of this writing.

It's not for me right now, but Year Compass is an interesting idea. (via Juhis)

I linked to this National Park-style font pack in weeknote 1973. Turns out there's another similar one available with seven weights.

Adam made this printable one-page calendar a while back and I printed one to stick in my journal. If you want to see some other versions or find out why Adam is the worst, it was on Hacker News this week. Ellane also has some markdown calendars if that's your thing.

I love this outtakes post from anh.

Finally these Pigeon Post combination letter/envelopes are really cool and I'm buying some for sure. Maybe I'll send some of my stickers to random buyers.

Good Riddance, 2025 - Kev Quirk

2025-12-31T18:48:00.000Z

I'm glad to see the back of 2025, so let's focus on 2026.

As I write this, it’s 18:49 on New Year’s Eve. I’m sat in the lounge, with the fire going, annoyed that our NYE plans have been ruined by me having the flu.

I feels like an appropriately shitty end to a very shitty year.

In 2025, I:

Worked way too much
Spent too much time away from home travelling for work
Neglected those I love
Helped my Mum, where possible, through recovering from a heart attack
Lost my sister

Yeah, it’s been rough. But toward the end of the year I made some decisions to make my life more sustainable. So I’m going to try and not dwell on the past, but instead look forward to the future.

As I look to 2026, I want to:

Eat healthier and lose weight
Exercise more
Spend more time with loved ones
Stop being such a shitty friend and make more of an effort

There’s more I want to do, but if I can get these done, I’ll consider 2026 a win.

We’ll see…

Thanks for reading this post via RSS. RSS is great, and you're great for using it. ❤️

You can reply to this post by email, or leave a comment.

Watching Bari Weiss Murder Investigative Journalism at CBS - Werd I/O

2025-12-31T16:51:59.000Z

[Spencer Ackerman at Forever Wars]

Spencer Ackerman has had a 60 Minutes piece spiked and has withstood White House demands to stop a story. And he has some thoughts about Bari Weiss spiking the 60 Minutes story about CECOT.

“That's where my mind goes when I read about Weiss "delaying" the 60 Minutes CECOT piece. To what it would have meant for The Guardian leadership to cave on the Snowden stories. What it meant for the Times leadership to have self-censored its reporting on NSA bulk surveillance and CIA torture. The lesson to any journalist attempting to reveal the highest-stakes stories is that you must fight your outlet if you're going to do the work your audience needs to understand the situation it's in.”

[Link]

A 2026 checklist - Werd I/O

2025-12-31T14:46:33.000Z

I'm not sure I believe in New Year's resolutions as such: too easy to break. But there's something powerful about a checklist, so I've been thinking about what it might look like to build one for every single day next year.

Here's what I have.

Every day

Close my Apple Fitness rings.

Go for a walk, even if it's just a mile.

Eat nutritious, plant-heavy food, if possible.

Read a book, even if it's just one page.

Write creatively, even if it's just one sentence.

Communicate with a friend, even if it's just a one-line text.

Tell the people I love that I love them.

Write down at least one thing I'm grateful for.

Write down at least one self-affirmation.

If I'm at home

Use the rowing machine, even if just for 10 mins.

This does not need a checkbox, but give my buddy the biggest hug.

Get to bed before 11pm.

Nothing bad will happen if I don't do these things. But they're a north star: a way to remind myself of what is important. I'd like them to stick.

If you have something similar, I'd like to read yours.

A new RSS feed! - Hey, it's Jason!

2025-12-31T14:44:27.000Z

The Obsidian Publish RSS feed is not great. So I am making my own! This should be a test entry. Real feed to follow!