Tamseel's Weblog

Latest

  • On Arbitrariness of Labels

    In

     There’s something fundamentally wrong with classifications/categories/genres or as you can call them generally, labels. What’s wrong about them is two-fold. People’s inability to understand the arbitrariness of the process of putting things into different buckets to the extent that they don’t like things which aren’t too easy to put into a bucket, and also people who create things that are too easy to put in a bucket. Labels (classification / genres / categorization) rarely do justice with a good work.

    I wanted to specify it down, but I am unable too. Every good creative work (book/movie/game/whatever) has such of its identity derived from its unique idiosyncracy that it feels wrong to use the regular labels with them.

    But the strange thing about this is that labels are useful. They serve a useful function to make a lot of information manageable, which makes me realize that the problem arises when people start giving labels more importance than what they hold, and that probably happens because most of mass content has too little of an identity of its own that defining them with labels become the norm.

  • RIP SBP Data Authenticity ft. Aaj ke hisaab se

    In

    Edit: I later found out the mistake was my own. Details in end.

    ————-

     Today was the first time, the intertemporal currency value converter I had made a while ago came to my use, when during a phone call with my father, I instantaneously converted a past amount he mentioned to its current value, and I told him about this, and he liked the idea. When I told him, that for present value it’s using 2024’s last quarter gold prices to compare against historic gold prices, he asked me to get present gold price updated and share it with him. This time, I figured out extracting data from Business Recorder is pretty straightforward using an AJAX call that the webpage was making under the hood.

    Last time, I wasn’t able to find out a good source for recent data. Since, BR had this data since 2011, I thought about comparing it with the data from SBP that I had used in conversion in my converter, and voila, the two data series are not validating each other.

    The price growth rates for the two sources are not even remotely close.

    And then they say, I am too paranoid about data sources.
    Now, what am I supposed to do plug in my converter? Does authentic data even exist?
    ——-
    Edit 14/07/25: So, I found out the data from both sources is actually consistent. The issue was how I was grouping it. The SBP Handbook explicitly mentioned the years were fiscal years. But when grouping BR data to get annual average, I had grouped by annual year instead of fiscal year. When I grouped BR data by fiscal year, it was pretty much the same.
    PS. None of this matters because I think my decision to use gold prices as deflator was pretty stupid one, because gold prices have increased much more sharply then overall prices, probably because demand for gold was not much high pre-2000 but it increased sharply in the next 2 decades; that’s my hypothesis. But anyways, I have found old CPI data for Pakistan, and will plug it when I find time.

  • Miyazaki’s movies & kids not being taken seriously

    In

    An interesting pattern Nabeel Qureshi mentions about what makes Miyazaki’s movies so special:

    [00:10:22] Dan: Another director you cited that you love before, Miyazaki. What do you think that he understands that maybe Disney, other animation studios are overlooking, and they don’t quite get?

    [00:10:31] Nabeel: Oh, yes. This is one of my favorite topics. I think Miyazaki just makes movies for adults that are also for children. He really takes children seriously as full beings, if you will. That’s very important. If you watch interviews with him, he’s always saying, I think kids have a very good sense of the issues that we think of as adult issues. Life and death is a simple example. Even a movie that’s relatively on the child-like side of his canon, like My Neighbor Totoro, it’s actually a pretty serious plot because the mother is on the verge of death, and she’s sick the whole time. It’s showing how these two children cope with that.

    Another example is Kiki’s Delivery Service. It’s charming, right? It’s this teenage girl, she’s going to become a witch, and she’s going to learn to fly. I feel like Disney would take this in a very whimsical childlike direction. Actually, it’s a drag, she moves to this Stockholm-like city. She has to get a job and work. It’s a grind. She gets sick. Nobody cares about her. There’s all these things that happen that you wouldn’t really expect to happen in a kid’s movie. Yes, I think his secret is he takes children very, very seriously, which I think most adults do not by default. He makes movies for children as though they were fully conscious beings.  [Emphasis Mine]

    Source: https://www.danschulz.co/p/nabeel-qureshi

  • Writing raw vs processed thoughts

    In

     It was interesting that I just simply had to nudge FA that he should write down his thoughts somewhere on the internet, and he would start doing that. It was easier in his case, because there was no need to convince or explain him, he already knew why he should be doing that. So, this was already something he thought, he should be doing, but as is the reality of this world, just desiring things or thinking that things should be a certain way, does not automatically translate into actually doing that thing. Other than reason or desire, one also needs to be agentic about actually doing that thing for that thing to be done, and sometimes, even the generally high-agency people need some sort of nudge to start doing the thing that they themselves think they should be doing. For this reason, I think, nudging people into doing something that they themselves think they should be doing is a good thing, and more people should be doing this.

    Anyways, that was the preface. What I actually was thinking about, was the difference between writing raw/unprocessed thoughts and structured/processed thoughts. If you have read about Visakan‘s ideas about it, notably do 100 things, he leans more on the idea of writing unprocessed thoughts and just doing things which is basically a better version of quantity leads to quality thing. On the other hand, if you have read Henrik Karlsson‘s work, notably the essay where he explains that the conventional internet wisdom of quantity leads to quality is flawed, and Anu Atluru’s work, notably Make Something Heavy, you might have impression that one should spend more time into processing his writings before publishing them.

    Now that I am thinking about how to phrase what’s in my mind, different frameworks are coming, so let’s try them one by one.

    If you actually read all these essays, you will realize there are differences in the context in which all of them are talking about. Visa is talking about doing something in a sense of enjoying the thing itself:

    “Do 100 Crappy Things For No Reason, With No Agenda To Live Up To, At Whatever Pace Feels Comfortable, However You Like.”

     While, if you read Henrik, he is framing it completely in his own personal context of doing writing that leads him to better thinking:

    “I also don’t think that optimizing for growth is a healthy way to write; a better metric for me is how much my thinking improves.”

    Meanwhile, if you read Anu Atluru, you realize, she is talking about creating something heavy — something meaningful and durable, in a broader sense but around context of current internet culture running on light things, resulting in this strange feeling of unfulfillment caused by not creating something heavy. So she has framed it as sort of a counter-effect of that internet culture, however, she also hints in between on the cycles of creation process.

    “At any given time, you’re either pre–heavy thing or post–heavy thing… Your gut knows what state you’re in. And the cycle repeats… No one wants to stay in light mode forever. Sooner or later, everyone gravitates toward heavy mode—toward making something with weight. Your life’s work will be heavy. Finding the balance of light and heavy is the game.”

    “Weight is not restricted to “work” in a traditional sense but to every arena of meaning.”

    Actually that incorporates both the frameworks I had in my mind which were:

    1. Understand the context or framework of thinking of the author and see how they apply to your own specific context, if they do.
    2. When an author makes a point, they are specifically framing it in context of people who are on a certain side of a specific point of reference. If you are standing on the other side of that point of reference, that specific framing might not apply to you even though the underlying concept of the author is totally valid.

    If I think about my own personal context specifically regarding writing, the closest it is, is to Henrik’s context. But I had been having this problem that when I intended to write processed thoughts, my mind would auto start processing those thoughts before I had the time to even write them out, and what resulted was that I would not be able to think about them in a concrete and structured way because the thoughts remained as vague pre-concepts in mind without being materialized into type. So for me, the first step was to first build the habit of being able to write down the thoughts first, because when I did that, no matter if they were much processed or not, I was able to restructure or reframe that idea in my mind later on could see the flaws where they were.

    This was a reasoning behind my idea of starting to write in weblog instead of my earlier idea of writing essays. However, even for me being able to write a weblog (where thoughts being properly structured is not a strict requirement), there were some hidden intermediaries, which were my physical diary and a secret blog I had shared with no one else. Interestingly, Henrik despite being inclined to writing processed thoughts, realized that it too had its benefit:

    What has delighted me about the shit blog is how abundant it has made me feel. I sit down and type as fast as I can, and the results—well, they suck, but they don’t suck that much. They have a certain breeziness and some insights, too—insights of a different kind than I have in the serious essays. Which means I have underestimated my capacity! I can actually just sit down, without energy, without ideas, and if I frame the task in the right way, I can extract something of value from myself. The sense of scarcity I felt previously—feeling that to write the actual essays, I needed hours of high energy, which is scarce since we homeschool our kids, and I work, and the 2-year-old wakes up at night screaming, and feeling, because of this, that I needed to use my limited energy on good ideas—this feeling of scarcity has, I realize, kept me from doing more and better work.

  • Certain about means, uncertain about end, and vice versa

    In

     Since I’m about to graduate in a while, and since these are the times when a lot of people start asking you about your future plans, I thought to write about this. Though most of the people are such that you just have to give them an artificial answer to satisfy them because they are not in the mood of trying to understand the idea I want to convey, there are also some people who are in this mood, and so I might be sharing this with them.

    Most of the things we do are a means to a certain end. What happens many a time is that people get so engrossed in trying to figure out the means that they do not give much thought about exactly what kind of end they would want to achieve, and thus it’s no wonder that many a times, when people end up achieving the exact end that they themselves wanted years ago, they regret it, or at-least do not like it as much as they thought they would.

    So, I believe it’s a better thing to start from the end, and then retrace back the means. But there are issues with this approach as well. Because when most people think about an end, they are not thinking about the true end itself but also another means a level high up. For instance, if you ask college students about what kind of work do they ideally want, they would list out different sorts of careers, but what they are actually targeting via any specific career path is a vague concept of the work that would allow them live their ideal kind of life. For instance, two persons might want to adopt two different careers, but both are actually targeting careers that allow them to make the most amount of money, but even making most of money is a means for another end, one of which could be to lead a comfortable life, or another could be to have a certain status in society.

    Living a comfortable life or having status in society etc. are actually very simplistic notions, but I’m mentioning them because these are the closest thing to a true end that most people would have thought about their work. In reality, targeting for one or two concrete variables to optimize their life for, is often not a good idea, and subconsciously, people try to optimize against a wide range of preferences with different priority and weightage.

    Now, the apparent problem with these preferences is that they change over time. But first, it needs to be realized that a lot of times, what changes over time is not your underlying preferences about the kind of life you want to live, but rather your reasoning about adopting what means allow you to optimize for those underlying preferences. Now, when I think about the instances when your underlying preferences actually change, it’s only because of you having imperfect information, sometimes about the world and sometimes about yourself. In some of the cases, this information gap can be somewhat improved by thinking out about these things in a structured manner, but the true bridging of this gap happens in the real world, when you actually dip your toes in the water. So, the solution, to the changing preferences problem is to increase the surface area of interaction of your inner self with the external world and then being actually thoughtful and perceptive about those interactions so that you gather maximum information about your inner reality and the external reality and their interaction. In other words, it can be said, you have to speed-run experimentation around your life, but not in a blind manner, but rather a perceptive and thoughtful manner, so that your learnings from the previous experiments guide you about what next to experiment.

    This was a very long tangent, but I think it can be used as a fair enough idea for explaining where I currently stand.

    Over the last three years, I have been doing experimentations, but initially I was not very thoughtful about them, and I was definitely not speed-running them, however, over the years, it has improved. Owing to these, I have got some idea of my underlying preferences or the true end I would want to achieve, however, I have yet to figure out what exact form of work would allow me to optimize for them.

    The words means and the end give the impression that one follows the other in time. But if we consider, a given point in time, or a certain kind of life one is living, then we can distinguish that certain form into two things, let’s call them essence and the body. The body represents a specific concrete form of exterior reality, while essence represents the core idea which is abstract and sort of undefinable in a sense — something difficult to enclose in a jar of words.

    So, I have somewhat learned about the essence of the kind of things I would want to work on in life, but I am trying to figure out what exactly the specific body or structure of the work would look like.

    Now, that I am thinking, maybe, that essence is not actually that undefinable. There are a certain elements of it which are difficult to structure out in words, but still a lot of it can still be structured out in words. It seems I wrote that this was undefinable because I did not want to define it, either here in this blogpost, or maybe I was just playing around with myself not wanting to think structurally about it, or maybe, that phrasing “difficult to enclose in a jar of words” was interesting enough that I did not want to delete it, but anyway, I think this pretty much sums up about my current situation.

    I have got an intuitive sense of the kind of work I would want to do in life, but I have no idea of specifics of that yet. Maybe, I will try to wordify this intuitive sense some other time.

  • Playing the Devil’s Advocate

    In

     I noticed on Saturday while talking with a good teacher, that I play something like Devil’s Advocate pretty often. This happens specifically when discussing about things I have given a good enough thought but I have still been unable to form confident conclusion about any particular side.

    I think this practice is helpful in the sense that it allows one to explore various points in the explanation-space of a specific problem or phenomenon in order to ultimately find out which point in that complex space is closer to the truth.

  • First Filter

    In

    Few days back I told GZ who had recommended me Steins;Gate anime based on my interest in science that I had watched it and I liked it, and asked him if there are other anime like this, and he recommended me to watch Dark which is a TV series not an anime (because it also involves time travel). I told him I had watched the first season, but I did not like it. He was really astounded, and so I explained. Dark is a really good TV series actually from all aspects but one. It’s cinematography, acting, music, direction, all of the things are top-class. When they do depiction of old timeline from 60s or 80s, that too has been done excellently. Most people do not consciously think about these things, but it takes consideration of multitudes of factors to make a movie or TV series give that certain effect that it has, and Dark had that effect. But the single thing that didn’t click to me was the core story. SPOILER AHEAD. The core story, at-least of the first season, was revolving around a town of people constantly cheating over one another. I’m sure they would have developed the story further in the next seasons, but the one for first season wasn’t interesting enough for me to continue on. Now, the core story makes one element of the multiple of elements, and if let’s say I intended to learn cinematography or film-direction, I maybe wouldn’t have bothered about it. But in general, it seems I have something that I would henceforth call first filter which outright eliminates some options out of the table regarding things I might be spending time on. I will explain more about what it means later. So, the core story of a piece of creative work sort of acts as a first filter for content I would want to consume. I probably wouldn’t have coined a new term for this if it was just about this, but as I was having my dinner few minutes ago, a pattern popped up in my mind, because I had applied a similar rule on a different thing as well. A little while ago, maybe one or two months, someone had asked me about potentially working with a specific FinTech company and I just started analyzing that company to understand what it actually does. So, their core product was basically a mobile app for saving committee but when I looked into it, it was just a credit app disguised as a savings committee app. It had much better UX and credit terms than most of the other credit apps, but this idea just couldn’t settle with me, that they are a credit app but they disguise themselves as a savings committee app. It might not seem that much of a problematic thing to most, but to me, it certainly is. Their targeted userbase belongs to a culture where borrowing on interest is disallowed by their religion. What the company is essentially doing by this branding, is extending credit to those people who do not wish to borrow money because of their religious ideology or whatever, by tricking them into thinking that it’s not credit. I probably wouldn’t have a problem if it was just a credit app as my problem here is not any specific religious ideology, but the fact that the company does not respect its customers’ religious ideology — it’s like a restaurant that serves alcohol in name of non-alcoholic beverage to a Muslim or serves beef in name of chicken to a Hindu. Anyways, I was just casually asked about this company and it was not like a job offer or something, so it was just something that popped up while I was trying to figure out the business model of that company. But after this thinking, I was certain, that this was not a company where I would like to work, essentially because it failed the first filter. Probably, more examples might up come up in my mind, but based on this, this is the basic idea of first filter. There are a lot of criterion and variables based on which one decides and prioritizes between several options, but there are some things, the presence of which just outright eliminates an option out of consideration; that specific thing is the first filter. Now, since first filter is what eliminates an option regardless of how much it desirability it has in other aspects, it means that first filter is and should be only applied on a problem regarding some core idea. For instance, with the first example, the core plot is the core point of the whole film or series (unless you are watching for specific reasons like learning cinematography) and if that feels sickening to me, I just don’t bother with it, and with the second example, the core idea of a product company is what their product is, and if that product is sold around a deception, it doesn’t make sense for me to work at such company. As I mentioned one could be watching a series to learn cinematography, one might also work at a company for some very specific reasons, and in that case he might not bother about the core business of the company. For things like work, these specific reasons do not exist for long-term, so I think it is a good idea to decide not to work at a place where you feel problem with the core idea of their whole business.

  • Ideal Weblog Platform

    In

     Blogger is not the right platform for a tech person to use for long-term. It is only good for non-tech people. If someone has the technical expertise the best way is to compose the blog posts as .md files on Obsidian or VS code and have a static site generator convert them into html files to host somewhere.

    But there is one major flaw I have seen in all SSGs is their closed look. By default all of these have an index of posts (maybe along with some excerpts) but not an open feed view of the weblog. This exactly is why Blogger felt compelling to me. This old-school bloggish feeling is not just about nostalgia, there’s an open feel to it, because you don’t get a list of posts that you can open if you wish to. Rather, you automatically get a feed of recent posts by which you get to know what the person has been recently upto. Interestingly, Substack also has this feature but this is specifically an issue with the idea of weblog and not essays, it isn’t much useful there.

    Consider the experience of reading a book. You randomly swift through pages, and if something catches your attention, you stop and read around. You do not first have to select a topic from the TOC to decide what to read. Weblog should be like this, where entries are open by default, and the homepage should be a feed of atleast around 20 recent entries from which the person can swift through by scrolling. And if someone wants a list of entries, there should be a table of content for that too (which is something blogger is missing).

    Also, I still need to think whether I would want /essays or /weblog in subdomain of the site or in the slug. This is difficult to decide.

    I am leaning towards latter because the written content is supposed to be the substantial thing on weblog.tamseel.pk/. What else am I thinking of having on my site? I can’t think of anything. So, why not have it under the main site. Because different subdomain is kind of a different site.

    Apparently, it seems no pre-cooked solution would satisfy my needs. I will have to tinker with these SSGs to specifically achieve what I want. I can’t tolerate Blogger’s stupidly long permalinks for long.

  • Vibe Researching

    In

     Even though I don’t like it for myself, I thought about trying to do a full-fledged thesis work for a friend with the help of LLM. But unfortunately, I failed. Either it was a prompting skill-issue or that LLMs haven’t reached that state yet.

    My thoughts lean towards the latter. Although Claude is fairly good at writing functional code to achieve the desired result, it like its counterparts is not even nearly close at understanding statistical nuances. I think, it’s about datasets and training. Training dataset even for complex programming problems is, I suppose, available in plenty amount, and all the documentations for programming languages do not have any conflicts. But firstly, the training amount on real world dirty statistics is available in less amount, and that different textbooks are written for different levels of understanding. And many a times, different authors use different terminologies for same things, and…

    My brain’s auto-forming connections with former thoughts, but yeah, LLMs basically extract the concepts from labels, and wherever the labels are not universally consensually defined for concepts, it might give LLMs a hard time extracting the concepts.

    I don’t think there’s something fundamentally different about statistics. If it’s tacit so is programming. The only thing I can think of is the log, or in other words, data for training. Getting LLMs to do non-nonsensical statistical work by giving them published research papers, will only be as good as showing LLMs only the GUIs or command outputs of programs instead of actually including the code for the program in the training dataset.

    The concept of reproducibility in research does have been something people have started to talk about, but still, it’s scarce. What we need is a log of statistician — something like a field diary, which contains all assumptions, hypothesis, decisions, the rationale behind decisions, the bad results, the deciding of an alternate route, or in short, the whole truthful process of the work. And I believe it’s not limited to statisticians only. More fields should be giving a thought about getting the tacit ideas documented down somewhere, so that we are able to build better LLMs.

  • Lingap – Extending the definition

    In

     Earlier in a thread, I had coined lingap as a situation when there’s a linguistic conflict, not a conceptual one.

    Here I redefine — or maybe, just rephrase the concept.

    Picking from my earlier entry, there are concepts and then there are words. Concepts are the abstract entities that reside in our minds within the thoughts. Words are the labels we give to those abstract entities in order to communicate those thoughts. Concepts can also be related to physical objects, like water — the molecules of which are made of hydrogen and oxygen, but still when I talk about water, I’m talking about that because there’s an abstract concept of water in my mind too. I have already gone too far across this tangent in my head, and I don’t need to put that down.

    Anyways, so I was saying there are four types of situations. When two persons are talking about:

    1. Same concept and the same label (normal communication)
    2. Different concept and different label (normal communication)
    3. Same concept but different label (like that folkstory about grapes where three persons were fighting about what to eat, and they were all saying grapes but in different languages)
    4. Same label but different concept (this is what I call lingap)

    I recently noticed this phenomenon when I shared the Intelligence essay by Talha with two individuals and I had to explain to both of them, how intelligence is a label for an abstract concept, and they should focus on what abstract concept the writer is trying to convey, and should not confuse it if against that label, we index a different concept in our mind. If so, we should forget what label/word/code the writer is using for the concept and focus entirely on the concept, and then decide whether we disagree or agree with the concept. After that, we have all the time to argue what would be a good label for it.

    So, lingap is a situation where two persons are using a same word/label but both mean different concepts by the same word/label. But the problem, is not just that. The problem is that people by default don’t distinguish between label and concepts. They think of them as intertwined. So the more people try to communicate/argue/discuss about the concepts, the more the discussion gets entangled, and the only way to resolve such a communication is to understand this phenomenon (which I have only labelled as lingap) and realize that labels and concepts are not intertwined and are completely separable.