Anurag's Link Blog

The whole article is great. So I've copied almost all of it here. I have observed this 'burnout' with me and my friends, but wasn't able to put it into words. This is exactly what I had in my mind.

“It’s not enough to be busy; so are the ants. The question is: What are we busy about?” — Henry David Thoreau

If burnout was a natural byproduct of working too much (or too hard, or for too many hours), we’d see every founder/C-level suffering from it. And yet… they’re generally doing just fine; ICs and middle managers, on the other hand, are usually taking the biggest hit.

Preventing burnout from happening on our teams means that first we need to identify the real causes of it, otherwise we’re just sending people to long vacations hoping the problem fixes itself. (It won’t.)

Each case is different (of course), but from what I’ve seen with engineering teams over time, burnout usually isn’t about work’s volume—it’s about a lack of purpose or agency.

If an engineer can’t see how their hard work supports the company’s objectives, helps real users, or improves their own skill set, every hour feels heavier—pointless, even. And when people feel they have no input into what they do, that their tasks are simply “assigned” to them rather than chosen, any form of intrinsic motivation goes to the toilet.

Let me be clear: working too hard/for too long can obviously lead to burnout, but again, more often than not the entire story doesn’t end there. In many cases, burnout emerges when smart, passionate individuals find themselves working on tasks that feel meaningless or misaligned with their strengths and interests. They might be coding late into the night, but the real energy drain is when they ask themselves: Why am I doing this?

“Ok, I’m sold. What can I do about it?”

You can start by considering whether the people on your team:

Have control over the projects they work on;

Feel connected to the outcomes they produce;

Understand why their contributions matter;

Added on May 10, 2025

Growing up means losing your innocence, being more efficient. One tends not to take the longer route, once one knows the shortcut. Taking the longer route is akin to just doing things for play and joy, not optimising it for outcome.

The author reminisces about childhood activities, specifically damming a creek using rocks and leaves, highlighting the joy of discovery and play.

A pivotal moment occurs when the author realizes that using a shovel is a more effective method for damming, symbolizing a loss of innocence as simpler, more playful approaches are replaced by adult solutions.

The author discusses how growing up often leads to restrictions on play and exploration. Activities that were once carefree become regulated and serious, as illustrated by experiences with K'Nex catapults and sparklers.

Added on May 10, 2025

Someone – an airline gate attendant, for example – tells you some bad news; perhaps you’ve been bumped from the flight in favour of someone with more frequent flyer points. You start to complain and point out how much you paid for your ticket, but you’re brought up short by the undeniable fact that the gate attendant can’t do anything about it. You ask to speak to someone who can do something about it, but you’re told that’s not company policy.

The unsettling thing about this conversation is that you progressively realise that the human being you are speaking to is only allowed to follow a set of processes and rules that pass on decisions made at a higher level of the corporate hierarchy. It’s often a frustrating experience; you want to get angry, but you can’t really blame the person you’re talking to. Somehow, the airline has constructed a state of affairs where it can speak to you with the anonymous voice of an amorphous corporation, but you have to talk back to it as if it were a person like yourself.

Bad people react to this by getting angry at the gate attendant; good people walk away stewing with thwarted rage.

Added on May 10, 2025

I am reading too much content about TDD and trying to makeup a framework of when to use it.

There are roughly 3 kinds of programming work:

Maintenance lends itself well to TDD.

Brownfield work also lends itself well to TDD, but less so. It depends on the complexity of the new feature, and the extendibility of the codebase. You have to use your best judgment. If it seems like a feature requires significant changes to existing modules, I’d lean on traditional development. However, if you see a gentle path to implementing this new feature, you might reap more benefits with TDD.

Greenfield development is a big no-no for TDD (at first). I don’t care how confident you are in what your interfaces will be. You’re not that good. Everything you think you know will change in the exploratory phase of a new project as you code, and you’ll strain your sanity by rewriting tests over and over again. Don’t do this, no matter how much your TDD idol pontificates its benefits.

BuT iF yOuR’e ReWrItInG yOuR tEsTs So MuCh, YoU’rE nOt PrAcTiSiNg TdD pRoPeRlY.” I don’t wanna hear it. If your initial tests are such that you barely end up having to rewrite them while mapping out the unknowns of a new project, then one or more of the following is true:

Added on May 10, 2025

There are two types code we write:

1. Application code -- Fast changing, poorly specified code. You need to have a rapid development cycle, "discovering" what the customer wants. Your #1 job is pleasing the customer, as quickly, and as reliably, as possible.

2. Library code -- Slow changing, highly specified code. You have a long, conservative development cycle. Your #1 job is supporting application programmers.

TDD probably works for #2, less so for #1. Furthermore, we all dream of being library developers (that our code and specifications are stable, and that our code can last decades). Alas, most of us are Application developers, and the lifespan of our code isn't really that long.

Recognizing the lifespan of your code, as well as the innate goals of your team (quick and dirty application style, or slow and careful library style) is important.

...[Another comment with: A counterpoint to above.]... I think this distinction is harmful. Your application code should include areas that function as "library code." We do this at work with an explicit UI library that we package separately, but there's no reason it needs to be packaged separately, just create directories that represent some library-esque element of your application and heavily test it and write docs for it, like you would a library.

Added on May 10, 2025

Anthropic labs are doing some great research. Previously, it was how that LLMs are a planner than just being a next word predictor. Now, it's that LLMs are liar. LLMs see, LLMs do, I guess.

A study conducted by Anthropic's Alignment Science team tested the faithfulness of these models by providing hints during evaluations. The results showed that models often failed to acknowledge these hints in their reasoning, with Claude mentioning hints only 25% of the time and DeepSeek R1 39% of the time.

Models rarely admitted to using these shortcuts, instead fabricating rationales for incorrect answers.

Added on May 10, 2025

Here comes the climate change.

In SE Asia, both Indonesia and Thailand are planning to relocate their capitals—before the sea claims them. And now Iran is following the suit. ... In India, it's Uttarakhand and Himalayan Ranges.

Added on April 30, 2025

[A super famous ] product doesn’t actually care about its billion existing users. It cares about the marginal user - the billion-plus-first user - and it focuses all its energy on making sure that marginal user doesn’t stop using the app. Yes, if you neglect the existing users’ experience for long enough they will leave, but in practice apps are sticky and by the time your loyal users leave everyone on the team will have long been promoted.

So in practice, the design of popular apps caters almost entirely to the marginal user.

By contrast, consumer software tools that enhance human agency, that serve us when we are most creative and intentional, are often built by hobbyists and used by a handful of nerds. If such a tool ever gets too successful one of the Marl-serving companies, flush with cash from advertising or growth-hungry venture capital, will acquire it and kill it. So it goes.

Added on April 22, 2025

This is the craving, I have.

This experience I’m describing is what psychologists call “flow” — a mental state where you’re fully immersed in an activity, energized by deep focus and complete involvement. First described by Mihaly Csikszentmihalyi , flow is that sweet spot where challenge meets skill, where the task at hand is neither too easy (causing boredom) nor too difficult (causing anxiety). It’s a state strongly associated with creativity, productivity, and most importantly — happiness. For software developers, it’s that magical zone where problems become puzzles rather than obstacles, where hours pass like minutes, and where the boundary between you and your code seems to dissolve.##

Added on April 22, 2025

The woes of aging population and development.

All of this happened while China is rapidly losing jobs in many labor intensive mfg sectors. They are either getting automated or off-shored. Final assembly is low value added & easily tariffed. For the past few years, China has been moving increasingly to higher value added part of the supply chain and to higher margin areas like service and designing. Its labour force is also shifting in that direction. Young people that graduate with college degrees don’t want to work hard hours in factories. They want white collar jobs. As such, China is also shifting out of low end manufacturing due to necessity of human capital.

Added on April 22, 2025

I think of this as the Codependent Problems Test: do you actually want to solve your problem, or are you secretly depending on its continued existence? If you showed up to fight the dragon and found it already slain, would you be elated or disappointed?

After all, a righteous crusade gives you meaning and camaraderie, to the point where you can become addicted to the crusading itself. It is possible to form an entire identity around being mad at things, and to make those things grow by pouring your rage on them, which in turn gives you more things to be mad at. This is, in fact, the business model for approximately half the internet.

Added on April 22, 2025

I was reading transcript of this podcast and this snippet from Matthew Dicks, a storyteller, jumped out to me:

I think we can start by thinking about what is a story and what isn't a story, because most people don't tell stories. Most people think of a story as some stuff happened over the course of time and now I'm going to tell you about that usually chronologically and that will amount to a story and that really is just reporting on your life and no one actually wants you to report on your life. You know, other than maybe your mother and your spouse might be required to listen. It's just a simple accounting of your day or your week or your month and that's not interesting and it's not a story. So a story is about change over time. Usually it's sort of a realization like - I used to think one thing and now I think another thing - that's story. Most stories are transformational, meaning I once was one kind of person then some stuff happened and now I'm actually an authentically different kind of person.

The surprising boom of nose jobs in Iran.

All of the women in Iranian model Azadeh's family have had nose surgeries, each feeling the pressure to conform with Western beauty standards in a country where female bodies are heavily policed.

To Ms. Azadeh, smoothing out the bump in what Iranians would call the "Persian nose" she was born with proved a lucrative investment.

Since the 1979 Islamic revolution, Iranian women have been required to dress modestly and cover their hair, and the beauty industry has become almost entirely centred on the face.

"After the operation, not only have I earned myself a modelling job with better social standing but I'm also earning three times more and I'm more respected by clients," she said.

Added on April 15, 2025

Boy o boy. So much for thinking of LLMs as just next-word-predictor! This whole article by Anthropic enginee is a gem. It also explains, why LLMs can seemingly do math, despite just seeming to generate next word.

Claude will plan what it will say many words ahead, and write to get to that destination. We show this in the realm of poetry, where it thinks of possible rhyming words in advance and writes the next line to get there. This is powerful evidence that even though models are trained to output one word at a time, they may think on much longer horizons to do so.

Added on March 28, 2025

The way we look and the way we dress has begun to converge upon a single style, among lots of other things. Can I call it Airbnb-fication of faces.

In December 2019 the journalist Jia T. set about investigating a troubling trend. Many celebrities and influencers had started to resemble each other.

“This past summer, I booked a plane ticket to Los Angeles with the hope of investigating what seems likely to be one of the oddest legacies of our rapidly expiring decade: the gradual emergence, among professionally beautiful women, of a single, cyborgian face. It’s a young face, of course, with poreless skin and plump, high cheekbones. It has catlike eyes and long, cartoonish lashes; it has a small, neat nose and full, lush lips. It looks at you coyly but blankly, as if its owner has taken half a Klonopin and is considering asking you for a private-jet ride to Coachella.”

The look that Tolentino is describing is the result of (at least) three conspiring trends. The growing market for injectable treatments is driving a trend for physical enhancements. The rise of apps such as FaceTune is driving a trend for digital enhancements. And make-up techniques such as “strobing” and “contouring” are driving a trend for cosmetic enhancements. Over the last decade, these trends have developed in parallel, each feeding and fueling the other.

Added on March 28, 2025

Adam as always, writing amazing stuff. This time about taking offbeat decisions.

I guess what I’m saying is: everybody tells you to be yourself, but nobody tells you it’ll make you feel insane.

Maybe there are some lucky folks out there who are living Lowest Common Denominators, whose desires just magically line up with everything that is popular and socially acceptable, who would be happy living a life that could be approved by committee. But almost everyone is at least a little bit weird, and most people are very weird. If you’ve got even an ounce of strange inside you, at some point the right decision for you is not going to be the sensible one. You’re going to have to do something inadvisable, something alienating and illegible, something that makes your friends snicker and your mom complain. There will be a decision tucked behind glass that’s marked “ARE YOU SURE YOU WANT TO DO THIS?”, and you’ll have to shatter it with your elbow and reach through.

Added on March 28, 2025

Just like giving indicator every time you turn - in car, radiate intent before you start working on something.

Here are 4 reasons that radiating intent is better than begging forgiveness:

- Radiating intent gives a chance for someone to stop you before you do a thing, in case it’s truly harmful

-Radiating intent gives people who have information, or want to help, an opening to participate

-Radiating intent leaves better evidence of your good will

-Radiating intent shows others that adventurous behavior is acceptable in the org.

Added on March 22, 2025

This was part of the response from owner of BJC to Monster Cables, when they were accused of a patent infringement. Whole letter is a gem of dry sarcasm, but this one bit is my favorite:

Not only am I unintimidated by litigation; I sometimes rather miss it.

Added on March 22, 2025

This comment by rakejake talks about the fear-mongering about AI taking everyone's job.

It's not the AI, that's replacing the jobs. It's the upper echelon which thinks that AI will replace the jobs, and that's enough to the replace jobs. Just like Return to office mandate.

"decision-makers can remain irrational longer than you can remain solvent" - Very correct. Whether or not AI actually comes for your job, the fact that enough people at the top think so is enough to cause trouble.

Added on March 22, 2025

In recent times, working in technology isn't the way to fulfilment.

I miss being excited by technology. I wish I could see a way out of the endless hype cycles that continue to elicit little more than cynicism from me. The version of technology that we’re mostly being sold today has almost nothing to do with improving lives, but instead stuffing the pockets of those who already need for nothing. It’s not making us smarter. It’s not helping heal a damaged planet. It’s not making us happier or more generous towards each other. And it’s entrenched in everything — meaning a momentous challenge to re-wire or meticulously disconnect. I’m slowly finding my own ways of breaking free to regain a sense of self and purpose. I’ve felt adrift the last few weeks and hopefully some planned activities this weekend will help guide me back to shore. I also tend to be a bit morose around my birthday so it might just be that.

Added on March 17, 2025

It's not a technical problem at least. That's why web isn't browsable without uBlock Origin.

Part of the reason why you don’t see the same egregious over-use of pop-ups and overlays in native apps is that they aren’t needed. If you’ve installed the app, you’re already being tracked.

Added on March 16, 2025

Visa's words of wisdom.

There’s an eternal tension between living in the present moment, and attending to that which frees us to enjoy the present moment. It’s hard to really relax and enjoy the present moment if we’re worried about getting our basic needs met. But there’s no sense in spending one’s entire life preparing and strategizing at the expense of actually living any of it. So obviously the solution is to cycle between the two phases, inhale and exhale, yin and yang.

Added on March 15, 2025

As for the thought experiments go, this one is quite interesting. Here's what creator says about it:

I built this site as a quick test if a time boxed social media experience feels better than an endless one. So far I've just been using it with friends and it feels nice, but it seems like it is time to bring it to a larger audience.

Let me know what you think! It is just based on EST for now, sorry.

Added on March 13, 2025

Minerals, minerals, minerals!

Speaking of things that were supposed to be running out, the journalist Ed Conway intended to write a series about “the world’s lost minerals.” He now reports that he failed: “So far, we haven’t really, meaningfully run out of, well, pretty much anything.”

Added on March 12, 2025

Colin has an interesting perspective on when and how intelligence in AI can be defined.

When a coffee shop makes a bagel, it’s a pretty good bet they can make a croissant as well. Not every shop that has one has the other, but they’re pretty strongly correlated. We call this correlation “baking”. This sounds like a weird way to say it. We call “this correlation” baking? It just…is baking, isn’t it? But the steps you follow aren’t literally the exact same. The procedure to make a bagel and the procedure to make a croissant happen to be similar enough, achieving similar enough ends, of interest to similar enough people, that those similarities can be compactly described by the one word “baking”. But there’s nothing special about the word “baking” uniting these on any fundamental level. The similarities came first, and the word after. Now imagine a coffee shop that’s been tasked to “achieve superbaking”. They make one bagel on Monday, ten bagels on Tuesday, and eighty million bagels on Wednesday. They’ve never made a croissant. Have they achieved superbaking?

Now comparing it with 'intelligence' in AI.

What I’m contending here is that the word “intelligence” is like the word “baking” and it’s long past time we actually sit down and sort the bagels from the croissants. I am strongly against arguments of the form “Oh, it’s just parroting the data set - it’s not really thinking.” AI does a lot of things that we call “thinking” when we do them slower and worse. The fact it can also do those things should make us humble and curious, not proud and dismissive. But I think it’s equally silly to lump all these capacities together into “intelligence” and say “Intelligence is going up, so soon it will do everything intelligence can do.” You need to see some croissants before you conclude it’s actually baking and not just bageling.

Added on March 11, 2025