Daily Log Digest – Week 27, 2025
2025-07-06
AI, ML and Deep Learning
PyTorch in One Hour: From Tensors to Training Neural Networks on Multiple GPUs #ai #ml #deeplearning #tensors
I was working my way through this article when I came across a nice, simple definition of the different categories - AI, ML and Deep Learning
AI is fundamentally about creating computer systems capable of performing tasks that usually require human intelligence. These tasks include understanding natural language, recognizing patterns, and making decisions. (Despite significant progress, AI is still far from achieving this level of general intelligence.)
Machine learning represents a subfield of AI (as illustrated in Figure 2) that focuses on developing and improving learning algorithms. The key idea behind machine learning is to enable computers to learn from data and make predictions or decisions without being explicitly programmed to perform the task. This involves developing algorithms that can identify patterns and learn from historical data and improve their performance over time with more data and feedback.
Machine learning has been integral in the evolution of AI, powering many of the advancements we see today, including LLMs. Machine learning is also behind technologies like recommendation systems used by online retailers and streaming services, email spam filtering, voice recognition in virtual assistants, and even self-driving cars. The introduction and advancement of machine learning have significantly enhanced AI’s capabilities, enabling it to move beyond strict rule-based systems and adapt to new inputs or changing environments.
Deep learning is a subcategory of machine learning that focuses on the training and application of deep neural networks. These deep neural networks were originally inspired by how the human brain works, particularly the interconnection between many neurons. The “deep” in deep learning refers to the multiple hidden layers of artificial neurons or nodes that allow them to model complex, nonlinear relationships in the data.
Unlike traditional machine learning techniques that excel at simple pattern recognition, deep learning is particularly good at handling unstructured data like images, audio, or text, so deep learning is particularly well suited for LLMs.
Oliver Burkeman on Insecure Overachievers
The Imperfectionist: Acting because you don't have to
The spiritual teacher Michael Singer says somewhere that the basic stance most of us take toward the world is that we try to use life to make ourselves feel OK. And this is certainly true of the type psychologists label ‘insecure overachievers’, who often accomplish plenty of impressive things, but who do so, deep down, because we don’t believe we’d have earned the right to feel good about ourselves, or to relax into life, if we didn’t.
It’s a soul-crushing way to live, not least because it turns each success into a new source of oppression, since now that’s the minimum standard you feel obliged to meet next time…
Most productivity advice, I think, caters to people mired in this mindset. It promises ways to help you take so much action, so efficiently, that you might one day get to feel good about yourself at last. Which isn’t going to work – because the real problem isn’t that you haven’t yet done enough things, or got good enough at doing them. The real problem is the fact that for whatever combination of reasons in your childhood, culture or genes, your sense of self-worth and psychological safety got tethered to your productivity or accomplishments in the first place.
This is a really good take on ambition, and continuing to be ambitious without being constantly in insecure-overachiever mode
One of the most important consequences of all this, for me, has been the realisation that when you begin to outgrow action-from-insecurity, you don’t have to give up on being ambitious. On the contrary: you get to be much more effectively and enjoyably ambitious, if that’s the way you’re inclined.
I’ve long been allergic to the notion, prevalent in self-help circles, that if you truly managed to liberate yourself from your issues, you’d ideally spend your days just sort of passively floating around, smiling at everyone, maybe attending the occasional yoga retreat, but not much more. “The more I heal, the less ambitious I become” is a phrase I’ve encountered multiple times online in recent months. And yes, sure, if your ambition was only ever a function of anxiety, becoming less ambitious would be an excellent development. Then again, the desire to create remarkable outcomes in your creative work, relationships or community – or even just in your bank balance – might just be an authentic part of who you are, once the clouds of insecurity begin to clear.
So you don’t need to choose between peace of mind and the thrill of pursuing ambitious goals. You just need to understand those goals less as vehicles to get you to a future place of sanity and good feeling, and more as things that unfold from an existing place of sanity and good feeling. (Besides, I’ve got to believe that ambition pursued in this spirit is far likelier to make a positive difference in the world.)
How do we boost birthrates
How do we Boost Birth Rates? - by Alice Evans #fertility #demography #children
Six factors need to be addressed
Women procreate if they expect rewards such as personal fulfillment or social approval.
People are more likely to have kids if it’s fun relative to other alternatives.
Economic and housing incentives from governments must outcompete other alluring alternatives.
Community plays a role by creating social expectations and shared activities among families.
Films and cultural portrayals could make parenting seem more desirable.
The rise of singles and solitude makes it harder to raise children alone and affects dating prospects.
2025-07-08
The Perils of ‘Design Thinking’
The Perils of ‘Design Thinking’ - The Atlantic #design #politics #culture
The concept of design, as the French philosopher Bruno Latour observed in a 2008 lecture, has had an “extraordinary career.” No longer is design about making objects more beautiful and useful; instead, he suggested, “design is one of the terms that has replaced the word ‘revolution’!” That might be the problem. “Our contemporary idea of design,” Gram writes, is often used to convince ourselves “that positive social change could be achieved without politics and government action; that problem solving could be both generative and profitable.” But most ambitious changes on the societal level require political consensus, and what’s profitable for some may not be beneficial for all. Design may be a distraction from the real work.
In Praise of “Normal” Engineers
In Praise of “Normal” Engineers – charity.wtf #engineering #normal #productivity
I really liked the part How do you turn normal engineers into 10x engineering teams?
Shrink the interval between when you write the code and when the code goes live.
Make it easy and fast to roll back or recover from mistakes.
Make it easy to do the right thing and hard to do the wrong thing.
Invest in instrumentation and observability.
Devote engineering cycles to internal tooling and enablement.
Build an inclusive culture.
Diverse teams are resilient teams.
Assemble engineering teams from a range of levels.
Decentering work
Is it finally time to decentre work? | Dazed #work #hustle #capitalism
“The journey of emerging into adult responsible contexts [the working world] involves at some point the need to do some sort of deprogramming and unlearning as the fantasy of your working life becomes deconstructed by reality,” 25-year-old Olivia tells Dazed. Olivia graduated with her MA in 2023 and started working at her university shortly after, but she recently quit her job. Her last job made her acutely conscious of the structural violence of late-stage capitalism, and shaped her plans for the type of job she’d like to do next. “[It is important to look at] how you are being valued at a place where you swap your time, skills, mind and body for money,” she says. “And I use those words with a lot of intention.”
Framing work in the way that Olivia has is beneficial to understanding the oppressive nature of work under late-stage capitalism because you are exchanging your limited time on earth and your health, which can be seriously jeopardised by doing a desk job, for money. And this is, of course, not by choice. Helen Hester and Nick Srnicek argue in their book After Work: A History of the Home and the Fight for Free Time, that “we are coerced into work on pain of homelessness, starvation and destitution.” In other words, we work because we have no other choice. This isn’t to say that work isn’t fulfilling and deeply enjoyable for some. It can provide a sense of purpose and optimism, especially when you’re doing something you love and feel passionate about. But the amount of time we’re expected to expend on our jobs (and lack of choice) can significantly sour that devotion.
I would argue that young graduates’ dissatisfaction with work is potentially connected to the fact that their priorities differ from those of past generations. As older people were marrying young and able to buy their homes and have children on their salaries, they were able to acquire symbols of the “good life”, as theorist Sara Ahmed describes it, even if it didn’t really make them happy. Gen Z (and millennials) struggle to receive the economic benefits that previous generations achieved through work, which potentially makes them less resilient to its brutality. It’s also important to note that the markers of the “good life” have changed in the social media age, where we are confronted 24/7 by influencers who are always on holiday, live in big homes and wear enviable clothing. They could not afford their lifestyle through a traditional nine-to-five job, and it makes one question the point of having one when you could just become a content creator.
As we’ve already established, under capitalism, everyone needs to work; and detailing the ways work under this system is coercive and detrimental to one’s health doesn’t change that fact, if anything, it’s just depressing. The intention of this article is not to depress anyone, but rather to prompt an examination of our feelings about work, as it often defines so much of our self-worth, creating fears about how others see and value us. You are not a failure if you can not find work in our incredibly fucked-up job market. You are more than the work you do and what you produce, and the same energy that goes into your working life should be expended on your personal life. Of course, our personal lives do not provide us with income, but they do equally (if not more so) provide us with the tools we need to stay alive: our friends, family and our communities (if we invest in them).
2025-07-10S
Rigidity in Islamic Societies
State Power & Punishment - by Alice Evans #islam #religion #power
Nobel laureates Daron Acemoglu and James Robinson have a brilliant new paper, published in the Journal of Economic Literature, proposing a dynamic interaction between culture and institutions. They suggest that a society’s values legitimise particular economic and political institutions, which then shape emergent configurations of culture. In their theory, each cultural set has a set of ‘attributes’ (which can be abstract or specific). These sets can be ‘free-standing’ or ‘entangled’
Contrasting cultural evolution across world regions, they suggest that Islam has some ‘highly specified’ attributes, which are also ‘entangled’. Sharia reigns supreme, since it is the word of God, revealed by Mohammad. Given this fundamental set of attributes, reinterpretation or critique becomes illegitimate. Instead, Muslims usually gravitate to scriptural literalism, and the religion thus becomes relatively ‘hard-wired’, less open to contestation or institutional reform. Rulers then cement their authority by invoking and entrenching Islamic values.
Alice Evans critically engages with Acemoglu and Robinson’s theory on the dynamic interaction between culture and institutions, focusing on Islamic societies. She argues that state power, prestige, and punishment are key engines sustaining cultural norms, especially through religiously sanctioned enforcement mechanisms like the office of hisba.
Key Takeaways
• Islamic cultural rigidity is not "hard-wired" but enforced through state power and punishment.
• The office of hisba historically policed morality, legitimizing rulers and controlling social behavior.
• Power, prestige, and punishment together shape cultural evolution and institutional legitimacy.
“Positive” Masculinity
The problem with ‘positive masculinity’ | Dazed #masculinity #gender #patriarchy
Masculinity is always in crisis, and the crisis is always new. In 1100, chronicler Oderic Vitalis railed against pointy shoes, for initiating young men into lives of effeminacy and sexual deviance. In the 1930s, George Orwell blamed the suburbs and middle-class morality. Today, it’s violent misogynist Andrew Tate and his legion of off-brand masculinity influencers.
TIL about the term "toxic masculinity"
‘Toxic masculinity’ was not a concept born of feminism. The phrase was coined, at least in print, by Shepherd Bliss, a member of the mythopoetic men’s movement – a mostly 1980s and 90s phenomenon that tried to explain why men felt like shit.
Can we really solve patriarchal violence with good patriarchs, who deserve their authority and always get things right? Envisioning a way out of oppressive masculinity is obviously better than defending it, but the positive masculinity movement hasn’t done much envisioning. ‘Healthy masculinity’ means both everything and nothing at the same time. Plus, some of its advocates are startlingly conservative: a recent BetterHelp ad features a man with a soothing voice being asked what ‘healthy masculinity looks like in 2025’, and telling the listener “maybe the tropes of provide and protect still hold,” clarifying that “maybe it’s less about muscles right now and maybe it’s about protecting your partner’s spirit or protecting their emotional safety.” Some are pushing back against traditional gender roles in more meaningful ways, but the language of ‘healthy masculinity’ leaves their endgame a great big mystery box. Do they want the borders of masculinity to be less policed, or do they just want to slightly expand the territory of what is acceptable for men? And where, if at all, do women fit into this?
Notes from an experienced software dev
All high value work is deep work, and all motivation is based on belief. #software #programming
Advice about software engineering that is worth repeating in full
Senior SWE, 12 YoE. The discourse around software development is incredibly chaotic and anxiety-inducing. I deal with the same emotions as everyone, but I manage to keep going despite having worked in a very poorly run company for a long time on a severely neglected product amidst product cancellation, brand cancellation, mass layoffs (one of which affected me), mismanagement, offshoring, you name it. I have managed to stay actively learning new tech, engaged on challenging problems, and having positive interactions with my coworkers consistently, even when one or more parties are being difficult to work with (which we all can be guilty of, myself included).
Here, I am to about share what keeps me grounded within all the noise.
This post itself is not a statement of fact, but a belief. But it keeps me going through all the noise and bullshit.
Also, a caveat: The claims I am making aren't the only claims to be made, and there are other important things to know. For example: It is true that all high value work is deep work, but it's not true that all deep work is high value work. A rectangle isn't necessarily a square.
All high value work is deep work, and all motivation is based on belief.
High value work is differentiated work. It's your moat. Not everyone has the grit, the attitude, the determination, and the ability to focus on challenging problems involving abstract concepts, especially when there is no immediate gratification, and when there is significant adversity in the environment. This is true of the population at large. But even within engineering/development, there are levels to this. Most people refuse to read. Most people refuse to do research. Most people panic when they see big log messages or stack traces. Most people give up when their code won't compile after googling for 20 minutes, if they even try googling at all. If you're the opposite of that kind of person, you will always be valuable in development.
All motivation is based on belief. Use this fact to be a leader, and use this fact to motivate yourself. All hard workers work hard because they believe they will benefit from it.
For some people, it is enough benefit to simply get in a flow state and enjoy solving a problem. But there is something deeper. Ask yourself what it is for you. Some examples:
ego boost (I am so smart wow)
prestige/praise (he/she is so smart wow)
distraction/addictive pattern (my marriage/family/health/social life sucks so bad, I need to forget for a while)
raw gratitude (or is it cope energy?) (I am grateful I get this fat paycheck to sit inside in comfortable temperatures and ergonomics, safely on a computer with no risk of injury or death, no one berating me constantly, no dealing with unreasonable patrons/patients/customers/schoolkids etc, just to solve challenging problems and be in a flow state, and if I could earn this money in a band or as a gamer I would but I can't so I'm just grateful for this opportunity so I can focus on myself and my family and my hobbies outside of work and build a nest egg for my family)
social (I love the people I work with, I genuinely have fun at the office with these cool people and I would still hang out with these people even if I weren't being paid)
Find out what motivates you, understand it, contextualize it, and ACCEPT it. Once you do that, you can have the space to figure out the same for others and help them along. I recommend taking the gratitude route. Gratitude can apply pretty broadly. It is actually a major life lesson in happiness.
Also, yes, corporate America is toxic. But you choose to work there. Every day you choose to work there, you should 100% double down on acceptance, or 100% double down on trying to find another job. Anything in between is total misery. Don't live life in resistance to what is. Accept what you can't control and work hard on what you can control. Either go to a startup and accept the risks, become politically active and solve the problem that way, or accept that you want the money badly enough and that the greedy, lying toxic charlatans running corporate America are the ones most able to give you the fat paycheck you signed up for.
Find what it is that motivates you in this field, and use that motivation to power some deep work so that you have some staying power in this field. It all starts in your own mind.
I know this devolved into a ramble. Just my two cents, hope it helps.
2025-07-11
Why I don't want a boyfriend
Why I don't want a boyfriend - by Sky Fusco - Unsupervised #boyfriend #relationships
I’ve been dating men for twenty years. I’ve merged lives with brooding musicians, flamboyant jocks, hard-working farmers, single dads, tortured professors, treehouse builders, and award-winning chefs. I tried to love these men for who they were, and not for what they did, but it was impossible. Their actions towards love were clouded by ego, coercion, control, dominance, manipulation, self-loathing, reactivity, weaponized incompetence, and cowardice. With one exception, I don’t remember any of them taking the time to sit still, self-reflect, or even jot some thoughts down in a journal.
Currently, I see multiple generations of men who haven’t done any inner work, and generations of women who have done all of it. Men can barely look at themselves, and women are taught to look too much. This disparity is sad for everyone involved, but because of it, I rarely feel the benefits of the “love” men try to offer me, especially when compared to the nourishment and deep intimacy provided by my friendships with everyone else.
It turns out that love starts with the self, and it requires courage, attention, and devotion. Deeply loving someone, and being loved by them in return, requires radically loving ourselves. In my experience, most men don’t even like themselves. It’s no wonder I don’t really feel their love.
Ok this made me chuckle
An exception to my boredom is when I’m romancing academics—men who read profusely and have knowledge to share—or men who are phenomenally skilled and obsessed with their work. I learn so much from them, but both of these archetypes ultimately make difficult partners, namely because they’re workaholics. They are, however, good temporary lovers, and can turn into the best peers.
These days, I cherish my friendships with men who aren’t trying to court or possess me—mentors, dads, friendly neighbors, helpers, or the partners of my friends. They almost see me; there’s no fantasy for them to project onto me, blocking their view. And if there is, I don’t have to know about it.