2025-04-23

Banking and Crypto

Crypto Might Get Some Banks - Bloomberg #crypto #stablecoin

But narrow banking pops up elsewhere. One important modern form is stablecoins. A stablecoin is a crypto form of banking: You deposit dollars with a stablecoin issuer, it gives you back tokens entitling you to get your dollars back, and meanwhile it does whatever it wants with the dollars. In the unregulated early days of crypto, “whatever it wants” could be quite spicy indeed, but these days stablecoins are a big business and there is something of a norm of parking the deposits in very safe short-term dollar-denominated assets, ideally Treasury bills or reverse repos or a BlackRock money market fund. If you launched a new stablecoin today and said “we will take your dollars and use them to make loans to emerging crypto entrepreneurs,” you’d have a hard time competing with the big incumbent stablecoins that say “we will take your dollars and use them to buy Treasury bills.” (Especially if, like most stablecoins, you didn’t pay interest.)

The Future is Augmentation

Why LLM-Powered Programming is More Mech Suit Than Artificial Human #llm #coding #software #programming

There is a view in many circles that LLMs will replace programmers. I am hesitant to say that this will never happen, becuase a lot of things with LLMs have surprised me recently, and I expect more surprises to come. For now, however, I don’t see LLMs effectively replacing programmers; but they are transforming how we work. Like Ripley in her Power Loader, we’re learning to operate powerful new tools that extend our capabilities far beyond what we could achieve alone.

This transformation will change what we value in developers. Raw coding ability becomes less important; architectural thinking, pattern recognition, and technical judgment become more crucial. The ability to effectively direct and collaborate with AI tools emerges as a vital skill in itself.

The developers who thrive in this new environment won’t be those who fear or resist AI tools, but those who master them—who understand both their extraordinary potential and their very real limitations. They’ll recognise that the goal isn’t to remove humans from the equation but to enhance what humans can accomplish.

In my view, that’s something to embrace, not fear. The mech suit awaits, and with it comes the potential to build software at scales and speeds previously unimaginable—but only for those skilled enough to operate the machines in ways that don’t harm themselves or those around them.

---

2025-04-21

AI Phobia

AI Phobia Is Just Fear That ‘Easier’ Equals ‘Cheating’ #ai #phobia #skepticism

Why do these businesses care so much? I suspect they aren’t really worried about AI—they're clinging to an old belief that if work isn't visibly difficult to produce, it must be less valuable. When we dig beneath the surface of "no-AI" policies and detection tools, we find an age-old assumption that worth must be measured in struggle. This mindset shows up again and again, from “hustle culture” and the “rise and grind’ mindset that defined the 2010s to recent return-to-office mandates that prioritize presence over performance. In a culture that values butts in seats and availability on Slack, it becomes easy to mistake friction for effort and effort for worth.

Oddly enough, the very thing we’re resisting—the ease of AI—might be what sets us free. AI isn’t the first tool to challenge how we think about work, but it may be the most direct. By shifting the locus of effort, AI forces us to confront our dysfunctional relationship with work. It holds up a mirror to our culture’s deeply rooted belief that struggle equals value—and in that reflection lies a rare opportunity: to reimagine work in terms of outcomes, not optics; human flourishing, not performance theater.

Anne Helen Petersen once memorably described this as "LARPing your job"—performing a theatrical version of productivity. Workers engage in elaborate displays of "being at work": staying visible on Slack, responding to emails at all hours, and maintaining a digital presence that signals industriousness. The tools have evolved—from software that monitors keyboard activity to AI that analyzes facial expressions in video calls—but the underlying philosophy remains pure Taylorism.

The irony is that these measurements often have little correlation with value creation. Knowledge work rarely follows linear patterns. Our most valuable contributions often come from reflection, seemingly "unproductive" conversations, exploration of dead ends, and invisible mental processing.

A Survey of Reinforcement Learning

The State of Reinforcement Learning for LLM Reasoning #llm #rlhf #rl #reinforcement

This piece is by Sebastian Raschka of Build a Large Language Model fame.

Slowly making my way through this. Found it via a tweet and this is a lot of signal for the relatively short length of the article. Really enjoying it!

---

2025-04-20

Review Code and LLMs

Alex Bird: "I have found that reviewing code is harder than w…" - Mastodon Canada #llm #code #reviews

I have found that reviewing code is harder than writing code. Increasingly, I am trying to write code that is easier to review -- that when someone looks at it, it is easier for them to tell if it does what it ought.

One of my objections to widespread LLM use for code generation is that we replace the easier task of code writing with the harder task of code reviewing.

(I include ensemble/mob programming here as a type of code review -- continuous code review)

---

2025-04-18

Plastics

Plastics are greener than they seem

Although the drawbacks of the world’s reliance on plastics are all too apparent, the benefits they provide, in the form of reducing waste and cost, are all too easily overlooked. Plastics have made possible a bewildering range of new materials that can replicate the properties of existing ones, and can do things they cannot, while being lighter, more durable, and cheaper and easier to manufacture. These materials have become vital in everything from building to carmaking to consumer electronics.

Take food as an example. Plastic packaging prevents perishable foodstuffs from spoiling, making possible global trade in meat, fish, fruit and vegetables. It enables essentials like rice, cooking oil and powdered milk to be stored and distributed safely and cheaply. A one-litre plastic bottle weighs 5% as much as a glass one; plastic packaging thus reduces shipping costs and emissions.

Plastics have also eased the world’s reliance on older materials, and on the living beings from which many of them came. There are perhaps 10m pianos in the world. If all their white keys were made of ivory, how many elephants would remain?

---

2025-04-17

Intelligence in Birds and Mammals

Intelligence Evolved at Least Twice in Vertebrate Animals | Quanta Magazine

Intentionally Make Close Friends

Post 43: Intentionally Making Close Friends — Neel Nanda #friendships

Good experiment with some interesting ideas.

---

2025-04-16

Situationships

Bad Romance #situationships #love #dating

A friend of mine is in a situationship where neither of them talks about their dating lives, so she doesn’t know if he’s seeing other people, and he doesn’t know if she is. Instead, she’s piecing together context clues to figure out his status. This is an untenable situation, and I was surprised she was okay with it. But I think the arrangement satisfies some of her needs while keeping alive the hope of real commitment. Sometimes, uncertainty is preferable when seeking clarity could mean learning that what you want is impossible. It’s like the Prisoner’s Dilemma: staying in a mutually tolerable situation feels safer than risking “losing it all” by asking for the truth.

AI as a normal technology

AI as Normal Technology #ai

We articulate a vision of artificial intelligence (AI) as normal technology. To view AI as normal is not to understate its impact—even transformative, general-purpose technologies such as electricity and the internet are “normal” in our conception. But it is in contrast to both utopian and dystopian visions of the future of AI which have a common tendency to treat it akin to a separate species, a highly autonomous, potentially superintelligent entity.

Autonomy and Connection

The Age of The Social Paradox - Rob Henderson's Newsletter #books #review

A fascinating new book, “The Social Paradox: Autonomy, Connection, and Why We Need Both to Find Happiness” by William von Hippel, a social psychologist, offers some insight. Drawing from decades of research in social and evolutionary psychology, von Hippel argues that human beings are pulled by two needs: autonomy (our desire to control our own life) and connection (our need to belong). For our hunter-gatherer ancestors, connection took precedence because small, close-knit communities depended on collaboration and mutual trust. Personal freedom in such societies was constrained by interdependent reliance on kin and friends.

Over time, however, modern lifestyles have elevated autonomy at the expense of connection, leaving many people struggling to balance these twin drives.

Von Hippel argues that the pursuit of autonomy often comes with a trade-off: If we want both independence and companionship, we may feel the need to influence or control others to align with our preferences. We want to live on our own terms, but we also don’t want to be alone — so we might try to guide others toward our way of thinking or doing things. As he puts it, “Only when others conform to your preferences can you meet all your autonomy needs while maintaining your connections.” But this approach is rarely sustainable. Relationships built on control rather than mutual understanding create tension and, ultimately, a hollow sense of connection.

VERT - online file converter

VERT.sh

File converters have always disappointed us. They're ugly, riddled with ads, and most importantly; slow. We decided to solve this problem once and for all by making an alternative that solves all those problems, and more.

All non-video files are converted completely on-device; this means that there's no delay between sending and receiving the files from a server, and we never get to snoop on the files you convert.

Video files get uploaded to our lightning-fast RTX 4000 Ada server. Your videos stay on there for an hour, or after they're converted (in the case of the input which you upload) or downloaded (in the case of the output which is to be downloaded), whichever comes first.

Their code is open source as well: GitHub - VERT-sh/VERT: The next-generation file converter. Open source, fully local* and free forever.

The rise of end times fascism

The rise of end times fascism | Far right (US) | The Guardian

Inspired by a warped reading of the political philosopher Albert Hirschman, figures including Goff, Thiel and the investor and writer Balaji Srinivasan have been championing what they call “exit” – the principle that those with means have the right to walk away from the obligations of citizenship, especially taxes and burdensome regulation. Retooling and rebranding the old ambitions and privileges of empires, they dream of splintering governments and carving up the world into hyper-capitalist, democracy-free havens under the sole control of the supremely wealthy, protected by private mercenaries, serviced by AI robots and financed by cryptocurrencies.

The Map is Not The Territory - Social Media Edition

cutting through the image - by Adam Aleksic #map #territory #social-media

However, as I’ve previously written, social media platforms want us to think that their map is the territory, because that helps their business model. The more we confuse their “content” with reality, the more we identify with it. Over time, you might really find yourself pursuing regular basket-weaving instead of underwater basket-weaving, simply because that’s what’s available to connect with—but that makes you easier to target as a consumer, since you’re now aligning your identity with the kind of broad metadata the algorithm is able to work with.

Once you identify with the algorithmic version of reality, the manufactured values of the platform become synonymous with your actual values. Content is presented as if it’s “good”—after all, it’s targeted for you, and has lots of “likes” from other people—but these metrics are made up. They reflect the platform priority of engagement optimization, rather than actually being intrinsically “good” or targeted to you.

We might prefer to compartmentalize our “algorithmic selves” from our “real selves” like we’re characters on Severance, but the truth is that they’re both constantly influencing each other. You get basket-weaving videos because you have a latent urge to pursue underwater basket-weaving, and then you ultimately take regular basket-weaving classes because you identified with the spectacular presentation of reality.

Our memes and language are similarly always evolving online and offline, with both mediums constantly influencing each other. The territory affects the map we draw, and then that map affects how we interact with the territory. This is inevitable—it’s just useful to remember which is which.

---

2025-04-15

Missed a bunch of daily logs because I was locked in. Hopefully I can resume today onwards.

Oliver's Burkeman's Latest Newsletter - No Escape

The Imperfectionist: No escape (from 2022)

I think the general point here, beyond the specific question of how to get writing done, is that we desperately want to be saved. We want to find some person, or some philosophy of life, that will spare us the fear or discomfort or self-doubt or tedium that so often seems to come along for the ride, whenever we try to make progress on things we care about. We hate feeling yoked to reality in such an unpleasant way; we long instead to soar above it, in a realm free from problems. And it’s the mark of a bad self-help book, a dodgy spiritual guru or an incompetent therapist that they’ll be only too happy to encourage the illusion that this might one day be possible.

I don’t think people generally seek out such quick fixes or back-door solutions out of laziness or entitlement. It’s not because they think they shouldn’t have to put in the same effort as everyone else. Rather, most of the time, it’s the opposite: it’s that they feel so inadequate and unqualified for the task ahead of them (of writing, of marriage, parenting – whatever) that they believe they absolutely need a miracle technique, some sort of edge over other people, some secret from a book, if they’re to have half a chance of not screwing everything up. They don’t realise that everyone else is just winging it, too – and that all they need to do is the straightforward thing that’s been staring them in the face all along: to just write for a few hours a week; to sit down for a few minutes and meditate; to be the most loving spouse or parent they’re capable of being on this particular day, and so on.

AI's impact on the job market

The Post-Developer Era • Josh W. Comeau #ai #jobs #software #programming

Josh Comeau with a solid take.

If you’re a job-seeker, you know that there aren’t as many high-quality job listings as there used to be, and the good ones get swamped with applications. It’s very hard to get an interview, let alone an offer.

But I don’t think this is because companies are actually replacing their developers with autonomous AI agents. As I’ve shared, the real-world experiences I’ve read just don’t support that hypothesis. So what gives? Why is it still so brutal out there?

I think there are a few factors:

1. Macro-economic stuff. Interest rates are still relatively high, making it harder for startups to attract the funding they need to grow and hire developers. For several years now, the general economic sentiment has been that we’re on the cusp of a recession.

2. Layoffs. Big tech companies laid off hundreds of thousands of workers over the past couple of years, for a variety of reasons. This means that there are tons of highly-qualified devs out there, looking for work.

3. AI myths. Some companies are still operating under the belief that AI really will make developers obsolete soon, and so they’re not hiring as aggressively as they otherwise would.

That last point is particularly frustrating. Companies are not hiring the developers they need because they’re convinced that AGI? is right around the corner, and when that egg hatches, we won’t need human developers at all anymore. “It’ll be any week now”, they’ve been saying for years. 😅

---

2025-04-09

Inquiring Together

Why not inquire together more? - Marginal REVOLUTION #inquiry #questions #discussion #ideas

I find that “inquiring together” works best when you are traveling together, and confronted with new questions.  They can be as mundane as “do you think the two people at that restaurant table are on a first date or not?”  From the point of view of the observers, the inquiry is de novo.  And the joint inquiry will be fun, and may make some progress.  You both have more or less the same starting point.  There isn’t really a better way to proceed, short of asking them.

For most established social science and philosophy questions, however, there is so much preexisting analysis and literature that the “chains of thought” are very long.  The frontier point is not well maintained by a dyadic conversation, because doing so is computationally complex and further the two individuals likely have at least marginally separate agendas.  So the pair end up talking around in circles, rather than progressively.  It would be better if one person wrote a short memo or brief and the other offered comments.  In fact we usethat method frequently, and fairly often it succeeds in keeping the dialogue at the epistemic frontier.

Shopify CEO Memo

This memo of the Shopify CEO mandating AI usage is doing the rounds.

The skeptic in me wonders if something that is obviously useful really needs a top-down mandate. Wouldn't developers automatically adopt it if AI was that transformational? It sounds to me like a bit of a bait and switch where the true agenda is something else - like justifying hiring less and dumping more work on the currently working engineers at Shopify, and using AI as an excuse for keeping up the same level of productivity and output.

On the other hand, this could one of those memos which goes down in technology lore as something that changed computing forever, like the famous Bezos memo: API Mandate: How Jeff Bezos' memo changed software forever | Kong Inc.

AI coding mandates are driving developers to the brink

AI coding mandates are driving developers to the brink - LeadDev

This dropped a day or so after the AI memo above, and is very skeptical about AI Adoption.

For software developers specifically, there are concerns that AI coding tools are introducing errors into their code, failing at many tasks, and compounding technical debt. But they also feel that misguided mandates are inhibiting the successful adoption of AI tools. While AI coding assistants can be helpful, it’s clear that how leaders approach and support engineering teams makes all the difference.

The HN Discussion around it is pretty insightful as well. From the comments:

I think a lot of confusion and frustration about this is the assumption that all programming is the same thing.

I have seen areas with just tons of boilerplate, straight forward UI stuff, basic test skeletons, etc that I guess could work with AI.

But personally, in 30 years I’ve just never done much of that kind of work. Probably because I find it boring. I go look for hard or novel problems, or data structures and APIs that are causing problems (overgrown and hard to use, buggy, etc). A lot of the time is just figuring out what the authors expected the code to do, and anticipating what we will need for the next few years. I don’t see AI helping much with that.

This pattern is maybe 20% about AI specifically and 80% about low-trust leadership.

LLMs require comparatively little training to use, and in fact training (like how to optimize prompts etc.) is probably a waste of time because the model behavior and interfaces change so frequently.

This puts them in the class of tools where the benefit is obvious once there is actually a real benefit (and not, say, the foreshadowing of some future benefit). This sort of tool doesn't require a mandate to gain adoption.

So, in the subdomains of software development where the LLMs are already useful, developers will naturally pick them up, assuming the business has secured the policy support and funding. In the areas where they aren't useful, businesses should trust developers and not waste their time with mandates.

And I say this as someone who uses LLMs all day every day for Python, Go, Bash, C/C++, pretty much everything. But as an active user I see the limitations constantly and wouldn't mandate their use upon anyone.

---

2025-04-08

Dealing with an internet outage where I am at.

---

2025-04-06

How Gatsby foretold Trump’s America

How Gatsby foretold Trump’s America #trump #oligarchy

Celebrating its centenary on April 10 2025, The Great Gatsby endures as the defining portrait of Jazz Age extravagance and glamour — a shimmering vision of energy, affluence and possibility. Away from the brightness, however, lies a world where dreams wither beneath corruption and malice.

Observing the deepening faultlines in American society in the early 1920s, F Scott Fitzgerald guessed right: he foresaw tragedy in the country’s impulse towards grandiosity and self-destruction in its reckless dishonesty. While Gatsby doesn’t predict the Trumpian politics of 2025 in any literal sense, it perfectly captures the society that would embrace such politics a century later. The novel’s prescience lies not in foretelling specific events but in diagnosing a culture where power enjoys impunity and cruelty rubs out its traces — a society run by careless people.

Fitzgerald’s verdict on them at the end of Gatsby has become for many a definitive statement on unaccountable elites today: “They were careless people . . . they smashed up things and creatures and then retreated back into their money or their vast carelessness, or whatever it was that kept them together, and let other people clean up the mess they had made.”

Gatsby reaches beyond the moral failures of its characters to expose carelessness as a political force. This includes not only the oligarchy’s immunity from consequence, but also the way extraction was equated with success. The unheeding brutality of so-called world-builders has returned most recently in the dark fantasies of Trumpism, and in Silicon Valley’s fatuous motto, “move fast and break things”.

Women and Work

Women, work and why we’re getting it all wrong #books #workplace #women #feminism

I have tried to articulate my own response to these questions after reading three recent books that explore work in contrasting ways. Patriarchy Inc by Cordelia Fine goes deep into biology and culture to debunk the common (and, in the second Trump era, ascendant) idea that men and women are genetically predisposed to desire different jobs. Charlie Colenutt has produced an oral history of UK workers in their own words, 50 years after Studs Terkel’s US version, Working. And Emily Callaci reanimates the 1970s Wages for Housework movement, a challenge to capitalist assumptions about which labour carries economic worth.

Fine’s book follows her bestselling Testosterone Rex (2017), in which she examines theories that this hormone is what creates inequality between the sexes: in Fine’s analysis, it’s culture, not nature. Patriarchy Inc takes these arguments into the workplace, to demolish the contention that our genetic and hormonal differences mean it’s natural that women and men do different jobs (this would neatly explain why there are so few women in corporate leadership, for example). Fine pithily calls this the “Different But Equal” argument. In her analysis — and it’s one I agree with — we have, rather, been acculturated from birth to expect one type of work or another. Fine calls this “mindshaping”. And it can be reversed.

Once men decide they want to do a job in large numbers, she argues, it becomes high-status. Fine outlines the early history of software programming, when women were considered suited to its demands as “a kind of puzzle-solving — like crosswords”, and held senior roles. In 1957, programmer Elsie Shutt became pregnant and was forced to leave her job at Raytheon Computing in Massachusetts. She carried on freelancing, recruiting other mothers as her workload grew. Shutt’s resulting business, Computations Inc, was rooted in collaboration between “physically dispersed part-time workers, in the days before email and Zoom”. It was very successful.

There’s no definitive answer, Fine says, to explain why women fell away from programming, after peaking in the mid 1980s — 37 per cent of computer science graduates in the US in 1984 were women. Software programming was rebranded as “software engineering” and, as more men rushed in, it became a high-status job. Screening tests became “dubious and gender-based aptitude tests and personality profiles”. Fine’s wider point is that when a significant number of a dominant group take desirable jobs, “social closure” occurs. Men hoard the roles for other men; the same would happen whenever there is a dominant group, because of our inbuilt affinity for others like us: “homophily”. It’s often not conscious, it’s how humans work, left unchecked.

Fine deals in the structures that underpin corporate life. For what happens “on the ground”, I turned to Charlie Colenutt’s Is This Working?. He interviewed 100 people across the UK, and 68 made it into the book, identified only by age and job description. Throughout this long (probably too long) book, the common thread is that it is fate or chance meetings that determine many working lives, rather than planned careers.

Colenutt groups the interviews into sections: “Sales work”, “Bosswork”, “City work” and so on. An online sex worker is with a panel beater, a security guard and a warehouse worker, in “Bodywork”. She’s a student, supplementing an inadequate loan, and finds the work “freeing” — but her account is a glimpse into the invisible, underpaid sectors that service those in well-paid corporate jobs.

The biggest question about work, which Colenutt and Fine approach, focuses on what and who we value. Emily Callaci’s Wages for Housework goes further: it explores a feminist campaign to pay housewives. Why do we tolerate the lack of economic worth assigned to caring? And is a different view possible?

The book is, in part, a chronological account of Wages for Housework in the wider landscape of women’s liberation in the early 1970s, and also contains chapter-long biographies of five key figures in the global movement. I previously knew nothing about this important strand of second-wave feminism and the women behind it, including the American activist and anti-racist organiser Selma James, and Mariarosa Dalla Costa, an Italian academic and political radical.

“Wages for Housework,” she writes, “is perhaps best understood as a political perspective, which starts from the premise that capitalism extracts wealth not only from workers, but also from the unpaid work of creating and sustaining workers.” As well as their employee, an employer gets “the labor of the second person who is at home sustaining him”. When the movement emerged in the 1970s, that worker was presumed to be a him.

LLMs from an operator's perspective

LLMs: an operator's view - by James Stanier #llm #operator #manager

In the midst of the current LLM explosion, we as operators find ourselves amongst:

• A blistering pace of improvement in the capabilities of LLMs. New models and products are being released at a rate that is hard to keep up with.
• Immense noise and hype online making all sorts of claims, good and bad, about what the future holds.
• An expectation from our companies to go full-on with "AI", which typically means LLMs, both in developer tooling and in customer-facing products. AI is the new data is the new cloud.
• Echoes in the industry that we are all now overstaffed as a result of productivity gains: that everyone should do more with less, and that AI is the answer to that.

What we'll cover related to LLMs is:

• The (real) rising floor of developer productivity.
• The changing size of organizations.
• The increasing importance of code reviews.
• The changing nature of interviews and identifying talent in short spaces of time.

As an operator, up-skilling your team to use these tools is now essential. Securing the necessary budget to give everyone access to the Pro tiers of ChatGPT, Cursor, or whatever tools represent the best fit for your team is a table stakes activity. And yes, this does mean that your budget will increase, but the productivity gains from an existing team will more than make up for it. Trade the cost of hiring new people for the cost of acquiring tooling.

You should also take the adoption of this tooling seriously. It is not just a case of giving everyone subscriptions and hoping for the best. You need to invest time and effort into training your team on how to use these tools effectively.

ML Inference Optimization Learning Path

One way to find a career in core ML/AI stuff #ai #llm #ml #career #learning

be me june 2024 learn ML model optimizations inference mostly, training a bit optimize SDXL inference post about it find cuda, learn it blogs on kernels for sgemv, softmax, etc. get in-depth knowledge on GPUs fal reaches out offering a role optimize inference of 4 models back to back feel grateful

How to Do Real Work With LLMs

How to Do Real Work With LLMs, Part 1 - by Jon Stokes

LLMs are the next phase of this exact same phenomenon. LLMs enable human language to eat the software that is still eating the world.

---

2025-04-05

Four Thousand Weeks by Oliver Burkeman - Cosmic Insignificance Therapy

This is my second time reading the book. Actually, it is the first because the first time I listened to it as an audio book. I am also reading this book right now as I slowly embark on a reset of long held belief and patterns of behavior around productivity and achievement.

These excerpts from the chapter titled Cosmic Insignifance Therapy really hit home.

The hazard in any such discussion of “what matters most” in life, though, is that it tends to give rise to a kind of paralyzing grandiosity. It starts to feel as though it’s your duty to find something truly consequential to do with your time—to quit your office job to become an aid worker or start a space flight company—or else, if you’re in no position to make such a grand gesture, to conclude that a deeply meaningful life isn’t an option for you. On the level of politics and social change, it becomes tempting to conclude that only the most revolutionary, world-transforming causes are worth fighting for—that it would be meaningless to spend your time, say, caring for an elderly relative with dementia or volunteering at the local community garden while the problems of global warming and income inequality remain unsolved. Among New Age types, this same grandiosity takes the form of the belief that each of us has some cosmically significant Life Purpose, which the universe is longing for us to uncover and then to fulfill.

Which is why it’s useful to begin this last stage of our journey with a blunt but unexpectedly liberating truth: that what you do with your life doesn’t matter all that much—and when it comes to how you’re using your finite time, the universe absolutely could not care less.

It’s natural to find such thoughts terrifying. To contemplate “the massive indifference of the universe,” writes Richard Holloway, the former bishop of Edinburgh, can feel “as disorienting as being lost in a dense wood, or as frightening as falling overboard into the sea with no-one to know we have gone.” But there’s another angle from which it’s oddly consoling. You might think of it as “cosmic insignificance therapy”: When things all seem too much, what better solace than a reminder that they are, provided you’re willing to zoom out a bit, indistinguishable from nothing at all? The anxieties that clutter the average life—relationship troubles, status rivalries, money worries—shrink instantly down to irrelevance. So do pandemics and presidencies, for that matter: the cosmos carries on regardless, calm and imperturbable. Or to quote the title of a book I once reviewed: The Universe Doesn’t Give a Flying Fuck About You. To remember how little you matter, on a cosmic timescale, can feel like putting down a heavy burden that most of us didn’t realize we were carrying in the first place.

The evolutionary angle here is interesting

These self-centered judgments are part of what psychologists call the “egocentricity bias,” and they make good sense from an evolutionary standpoint. If you had a more realistic sense of your own sheer irrelevance, considered on the timescale of the universe, you’d probably be less motivated to struggle to survive, and thereby to propagate your genes.

Truly doing justice to the astonishing gift of a few thousand weeks isn’t a matter of resolving to “do something remarkable” with them. In fact, it entails precisely the opposite: refusing to hold them to an abstract and overdemanding standard of remarkableness, against which they can only ever be found wanting, and taking them instead on their own terms, dropping back down from godlike fantasies of cosmic significance into the experience of life as it concretely, finitely—and often enough, marvelously—really is.

Goya Journal - MTR in Bengaluru and Iftar in Lucknow

The latest edition of the Goya Journal newsletter landed in my inbox and a couple of article caught my eye - one about Lucknow, a city I grew up in and about MTR in Bengaluru.

• From Sehri to Iftaar in the Streets of Lucknow — GOYA
• The Story of Bengaluru's Iconic MTR, the Original Udupi Hotel — GOYA

Tiny Experiments by Anne-Laure Le Cunff

Started reading this book on a whim because I saw Oliver Burkeman recommended it. I didn't expect much, and it turned out to be not that different from a typical book in the self-help genre. But skimming it does yield some nice gems, like this quote from Ron Finley

When people ask me what I do for a living, I tell them I breathe - I'm a professional breather. We create those false transitions, we make it all look like it's separate, just like we separate a garden, from a state, from a country, and ourselves from nature.

---

2025-04-04

Stop expecting your partner to fulfil all your needs | Dazed #dating #relationships #expectations

But what if we’ve now swung to the other extreme? Many of us now expect our partners to be everything: our best friends, sexual playmates, therapists, running buddies, cheerleaders, and more. We don’t accept any traits that jar with our own or a single conflicting value – and it’s likely this is making dating harder for us all.

But by approaching dating as nothing more than a tick-box exercise, we drive away possible moments of intimacy and connection. Contrasts in personalities, temperaments, and interests provide the push-and-pull needed to allow relationships to thrive. This doesn’t mean settling for someone who treats you poorly or forcing yourself to persevere with dating someone who doesn’t excite you. But if you like wild Friday nights and your partner prefers Netflix marathons, go party with your friends. If you love Sunday morning hikes and they don’t, go solo or find a buddy. A partner doesn’t have to be your everything.

I’ve been single for a while (aside from a minor situationship detour), and when people ask if I miss dating, I can honestly say I don’t. I get everything I need from my friends, sisters, mum, dogs, and – ahem – vibrational support. But most importantly, I’ve learned to rely on myself. If I want to see a play or catch a comedy show, I go solo. I can still grab lunch at that cute café I’ve been eyeing or curl up somewhere with a coffee and a good book. I stopped expecting one person to be my whole world, romantic or otherwise.

---

2025-04-03

Matt Levine on Stablecoins

Stablecoins Are Growing Up - Bloomberg #stablecoin #money #finance #crypto

I have always thought of stablecoins as basically unregulated non-interest-paying crypto banks. The idea of a stablecoin is that there is a company, the stablecoin issuer, and you send the company dollars, and it gives you back tokens (“stablecoins”). Each token is supposed to be worth a dollar, and if you hand the issuer back a token it will hand you back a dollar.1 But mostly you hold onto the stablecoins and the issuer holds onto your dollars. You both arguably benefit from this:

• You can use the tokens like dollars on in crypto applications, where actual dollars are hard to use: The stablecoins live natively on various crypto blockchains, can be sent over the blockchain, can be used in smart contracts, etc., in ways that don’t really work with dollars held in bank accounts.
• The issuer has the dollars, which it can invest in super-safe short-dated US Treasury bills, or anything else, to make money. It doesn’t pay you any interest; any profit it makes from its investing is the issuer’s to keep.

But yesterday Circle Internet Group Inc., the big US-based issuer of the USDC stablecoin, filed publicly for an initial public offering. It has previously filed confidentially, but its prospectus is available now. It is an interesting read, and it has caused me to update my model of stablecoins in two respects:

1. Circle is not an unregulated bank; it is a tech front-end for US banks and asset managers.2
2. I am not as sure as I once was about “non-interest-paying.”

AI Ambivalence

Great articulation of how it is like to use AI for software development.

Imagine you’re a Studio Ghibli artist. You’ve spent years perfecting your craft, you love the feeling of the brush/pencil in your hand, and your life’s joy is to make beautiful artwork to share with the world. And then someone tells you gen-AI can just spit out My Neighbor Totoro for you. Would you feel grateful? Would you rush to drop your art supplies and jump head-first into the role of AI babysitter?

This is how I feel using gen-AI: like a babysitter. It spits out reams of code, I read through it and try to spot the bugs, and then we repeat. Although of course, as Cory Doctorow points out, the temptation is to not even try to spot the bugs, and instead just let your eyes glaze over and let the machine do the thinking for you – the full dream of vibe coding.

I do believe that this is the end state of this kind of development: “giving into the vibes,” not even trying to use your feeble primate brain to understand the code that the AI is barfing out, and instead to let other barf-generating “agents” evaluate its output for you. I’ll accept that maybe, maybe, if you have the right orchestra of agents that you’re conducting, then maybe you can cut down on the bugs, hallucinations, and repetitive boilerplate that gen-AI seems prone to. But whatever you’re doing at that point, it’s not software development, at least not the kind that I’ve known for the past ~20 years.

The entire conclusion section is worth reading as well. Here is the last couple of paragraphs

So there’s my overwhelming feeling at the end of this post: ambivalence. I feel besieged and horrified by what gen-AI has wrought on my industry, but I can no longer keep my ears plugged while the tsunami roars outside. Maybe, like a lot of other middle-aged professionals suddenly finding their careers upended at the peak of their creative power, I will have to adapt or face replacement. Or maybe my best bet is to continue to zig while others are zagging, and to try to keep my coding skills sharp while everyone else is “vibe coding” a monstrosity that I will have to debug when it crashes in production someday.

I honestly don’t know, and I find that terrifying. But there is some comfort in the fact that I don’t think anyone else knows what’s going to happen either.

Hamish Mckenzie on Chaos Media

From the temple to the garden - by Hamish McKenzie #media #chaos

Today, we live in the age of chaos media.

Traditional media’s rigid order has been replaced by mayhem. Conflict supersedes consideration. Speed overwhelms verification. This system is, in many senses, a marvel, with massive democratization potential. Anyone can have a voice, and your idea, if the winds blow just right, can reach billions of people in an instant. One of its great virtues is that it demands that everyone—even the powers that be—speak to each other directly, and others can talk back. Now the savviest politicians sit down for long interviews in nontraditional formats and show their thinking.

Hamish (as the founder of Substack) has a vested interest in the shift away from traditional media. So understandably, this piece is a bit tepid when it comes to outlining a clear alternative. He just seems to just gloss over the corrosive effects of chaos media.

The chaos of our current media moment cannot last, but no one knows exactly what the new landscape will look like when it stabilizes. That’s precisely why your choices today matter so much. Every subscription, every share, every minute of your attention is a vote for the culture you want to see flourish. You can choose to invest in a system that values deep relationships over the flimsy validation that chaos media offers. You can reclaim your attention from the doomscroll feeds and pour it like water onto the seedlings of a better future. These actions aren’t just about getting better content or contributing to a healthier media economy—they cultivate a richer, more thoughtful culture capable of addressing the complex challenges of our time. It’s a culture worth subscribing to.

---

2025-04-02

Means of Production vs Means of Connecting

The Algorithmic Holiday - by Brett Scott #capitalism #gig #economy

The Gig Economy platforms, by contrast, run a lean version of this. A company like Uber knows that it doesn’t actually need to own tens of thousands of cars - the Means of Production - provided that it owns an information and management architecture that stands between those productive assets and society: Uber owns the Means of Connecting, rather than the Means of Production, which means they can still be a gatekeeper between workers and consumers, while outsourcing finance costs to the individual workers.

Choosing Next.js

Build Times - You should know this before choosing Next.js #nextjs #next #javascript

A breakdown of how Next.js seems to be tied to Vercel in many ways, who don't seem to care about being good open source stewards.

• Lack of adapters
• No official serverless support
• Vercel-Specific, Undocumented Code Paths
• Poor Security Incident Handling
• Lack of transparency in governance

Beauty and Relationships

Does beauty make people good partners? | Dazed #beauty #relationship #love #feminism

So much of romantic relationships is about pressing yourself into a shape you do not recognise, as Faye so accurately puts it. We are told that women are meant to be hyper-feminine and that the right man will put you in your “feminine energy”, as divine feminine con-influencers incessantly express online. But this isn’t a natural state of being as we are so often told, but a naturalised one. We act this way because we are told to and because it gets the best reception from both our desired partners and from the public.

Nobody knows what Leclerc and Mleux’s relationship is like, and yet they assume it’s good because she is beautiful and feminine and adheres to her gender correctly. Conversely, so does he – Leclerc is seen as being good-looking and is immersed in the hyper-masculine and perilous world of motor racing. We praise certain relationships and those within them for their ability to keep up with appearances.

This is not surprising. As beauty critic, Jessica DeFino wrote in her newsletter last year, as a society, we tend to have the moral code of Disney cartoons. We superficially believe that beauty is good; thus, a beautiful couple must have a healthy relationship (look at the contrast with how people talk about and aspire to Dua Lipa and Callum Turner’s relationship). When we see people we perceive as “ugly” (ie who sit outside narrow European beauty standards), we are quick to assume the worst. That their relationship must be bad, that something must be wrong, that they are not really in love. This is exactly why people feel the need to change and transform themselves beyond recognition. It seems contrary to what we ask of people when we are looking for love in the first place. We ask that people genuinely love us for who we are, but the world repeatedly proves that we will only be rewarded for looking a certain way, for not being or looking like ourselves.

In her book Right Wing Women, radical feminist Andrea Dworkin writes that “the tragedy is that women so committed to survival cannot recognise that they are committing suicide.” Dworkin was making a specific reference here to right-wing women who make accommodations to male domination for their own survival. But when I read that quote, I immediately thought about beauty and relationships and the ways we transform ourselves entirely for love and survival. But we deserve to do way more in this life than just survive.

We “survive” by adhering to our gender roles, with their own built-in violence that often result in tragedy. We “survive” by adhering to beauty standards, where people go under the knife and sometimes do not wake up or have lifelong complications due to surgeries. We “survive” by changing ourselves, making ourselves smaller and more easily digestible, to the point where we look in the mirror and can no longer recognise who is staring back at us. It may seem like survival, but as Dworkin would put it, it is a type of death.

Not all people feel forced to look a certain type of way in their relationships; but many of us can relate to feeling pressured within these dynamics to keep up with appearances. Love and the pursuit of it (albeit challenging) are supposed to be fun, exhilarating, and life-affirming, but rarely ever feel that way. So, what really makes a good partner? Maybe once we deprioritise the importance of physical appearance from our lives (and gender), we will genuinely know the answer to that question.

---

2025-04-01

Travel day. Digital nomadding for at least a month.

Mumbai Recs

A lovely friend sent along some very hyperlocal food recs for Mumbai #food

Chembur - my most fav vadapav in front of post office opens 4 or 5pm

Matunga - Classic pav bhaji or DPs Pav bhaji if you’re interested to try South Indian: Ramashray or madras cafe, there’s a soya chaap wala near madras cafe that’s also v good. Gupta pani puri near the matunga station. Too many Guptas check for the right one haha

Sion : gurukripa samosa. Very famous supplies to most theatres and schools in Mumbai

Dadar : Ashok Vada pav, sandwich also might be good in most places Bombay veg toast here.

Near marine drive : panini at Raju sandwich in front of KC college and K rustoms icecream is famous for the nostalgia value. I like the walnut crunch flavour there tho hahah

---