Bits and Paradoxes - 08

Beliefs, Thoughts and Writing

Jun 28, 2020

Hi Folks,

How are you doing, really?

The lock-down has become less-restrictive (way less restrictive). So, I finally went out of the house after almost 2 months to get the guitar strings (after having broken string for 2 weeks). I still had the stamina for walking fast. So, yeah! That's that.

This issue 08 is super long. Be prepared.

Many things happened this week in my mind-cave that I can't refrain from writing in depth to most of the sections here, especially about beliefs, Machine Learning landscape, and a concept of "living documents".

I spent the majority of the week working long hours, and even during non-office hours, I kept on looping back frequently to the problems I was solving (one of my not-so-good habits :/ ). In the long run, I procrastinated on my bi-weekly personal goals I had set up (hint: I had been trying to study basic economics).

A strange thing did happen in our locality. I found that few "leaders" from the locality reform committee (my mom included) started distributing this so-called "immunity enhancers" medicine. I tried really hard to convince mom to not fall for that who-knows-what-medicine. But failed. Eventually, I found out that the same thing had been done throughout other regions too. This made me believe how much we all seem desperate to get out of this "COVID-19" rut.

Another event was me trying to have a conversation with an elderly neighbor about the impact of the lockdown on a daily-wage worker. I was again dismissed of the position to conversate because of the age gap, which shouldn't at all be relevant.

These two events particularly forced me to contemplate about "beliefs" in general - from an individual to a group and society, and how we all choose to believe in our beliefs and that almost all of us follow some type of Belief Template.*

I realized that I have been having the same belief templates for years now, be it the coding paradigms or my thinking patterns. My slow thinking and the inability to do fast context-switching is also having certain "unhealthy" impact on my mental health. When someone says about something (both non-technical and technical), I often stay there nodding, as Bijay mentioned in his Things of note 05, accepting things at "face value". And contemplate the matter later.

Throughout the years, I have coped with this, particularly, going inside my mind-cave, writing things down, analyzing various things, and trying to recall and simulate events in my head. Most importantly, reading, listening, and simply observing. There were times back then when I too got offended from certain things (but not in recent years). Eventually, I came to an "inconclusive conclusion" (an oxymoron I presume) that Life is all about constantly updating our prior beliefs and having infinitely minuscule transformations. (broken thoughts I had written in Evernote around October of 2019)

Nevertheless, writing has profoundly impacted the way I think, and it’s more than a tool now. A habit to think critically.

BTW, I have restarted my habit of maintaining a daily journal and using Listen Notes fully now.

*Belief Template
I wonder if I can do footnotes in the newsletter. It has turned into more like long-forms. Also, the reason might be there are lesser clicks, opens only. So, this format is still an experiment in my writing style. Anyway, what I am trying to convey is, belief templates come from how we are structured in our lives, our society. Again, this is me being highly influenced by David Foster Wallace. This also made me recall when someone said they “hated” something. But that wasn’t even close to hatred. “hate” is a pretty strong word that should be used carefully. That’s why I don’t “hate” anything. There are only different vresions of “dislikes”, the preference of which might change later.

Image Source: moretothat

~Reading~

Do You Really Believe What You Believe?

Lawrence Yeo | moretothat | 13 min

The question “Do You Really Believe What You Believe?” seems simple, but the answer isn't. It's a bit hazy to what extend anyone believes their beliefs. Often times, we believe in our beliefs because of concerns and reputation.

What is so concerning about belief systems is our inability to judge our own thought processes, but doubting others. When it comes to others, we tend to simply go to certain depths to test their beliefs -- through various tools of questioning -- and that we seem to become a bit "less" critical when it comes to our own. This process of introspection, however, is not linear. At least, we can try to reinforce, time and again, our actions, words, and thoughts -- a concept of [[Belief Triangle]] -- to really have a wider perspective.

The Dragon In My Garage

Carl Sagan | 4 min

This is a thought-provoking experiment on belief.

Suppose I say to you that there is a fire-breathing dragon living in my garage. What would your thought processes be? You attempt to test the hypothesis with reasoning. But I reject every explanation (test) you propose logically with explanations like the dragon being invisible, cannot be detected by other sensors and such.

This thought experiment reminds me of the conspiracy theories out there, scattered within an individual's psyche or that of some group (cult). Most of them seem to not have foreseeable evidence and that it's hard to even disprove a few of them. But that doesn't even remotely prove those conspiracies to be true. Maybe it's fine to reject those circumstances and put that belief on hold for the future when we get more data.

If there's no conceivable experiment to disprove the experiment, what does it mean to say that the dragon exists?

Your inability to invalidate a hypothesis is not at all the same thing as proving it true.

Another aspect of this thought experiment is to think critically. Maybe the belief you have, which seems a bit off and cannot be proved (or disproved), can also be reasoned with a rational process. Maybe your belief in the belief can be justified by looking at your mental situation. What has made you believe in it?

Living Documents

Sam Squire | 10 min

Note: This is a mix of the author’s original idea and my interpretations.

Like any tools that are shaping our lives, we are probably doing the same with our thinking processes, especially the interactions we have with our thoughts through writing. From jotting down our "broken thoughts" to taking notes to writing full-fledge long-forms, it's certain that we are overloading a ton of information in our mind, and offloading them if not the least.

The platforms where we write tend to affect how we “think critically”. The fact that our ideas mutate and evolve through time is directly correlated to how we "write". And it's this spot that kinda makes a way for a new emergent paradigm - Living Documents.

The traditional writing paradigms seem to restrict a user to a single dimension (if you work on X, keep on working on X and when you have another idea Y just move to another document and simply archive X). However, that's not even close to how people think.

When people write, they do not write or think linearly.

And that's where one of the properties of living documents come in - [[Divergent Writing]]. This paradigm enables people to disperse their thinking to multiple dimensions and give a way for “non-linear” writing processes which catalyzes the chaos of ideas in one's mind (in a good way). This, while seems obvious, opens up a new graph-like thinking mindset where each node represents splits in connected-ideas.

For me, this concept seems to be a turning point. Having recently discovered tools like Roam and Obsidian, it feels natural to write ideas and think upon them critically. Of course, it doesn't mean traditional software systems aren't good. They have their own sets of "niche". And this brings us to the other two properties of living documents:

Embed Anything
Natural Habitat

Living documents make it seamless to embed anything, even dissonant interfaces. It's like a loose coupling of various echo-systems, each residing in their own natural habitat, and interacting through some non-restrictive mechanism. One example of this is to see how web pages can embed source codes that are executable inline. Of course, traditionally you would copy the source to some other environment, execute and see the output. But for the aforementioned properties to hold true, you should be able to simply embed the source code in the webpage and the execution happens using some type connection with virtual machines right there in the web page.

This brings us to the last properties of living documents: seamless data collection and relationships.

For the documents to be "living", it should allow users to add data without any restrictions from where relationships could arise. Relationships help to understand (i) existing ideas clearly (ii) new ways to think.

Examples of living documents

Wikis are pretty much an effective implementation of Living Documents.
A movie script can also be a form of living document where the interpretation is open to anyone.

Example of documents that aren't living

Dictionary

Encyclopedia

Another, food for thought can be to ask if any constitution is a living document or not. The US constitution says it’s a living document.

In all, everything seems to stem from the concept of Divergent thinking and writing, where contents aren't static but keep on evolving.

What I learned from looking at 200 machine learning tools

Chip Huyen | 10 min

This is an interesting read I'd say from a technological standpoint. The stats do make sense given the nature of AI Hype train at present moments.

Tentatively, each machine learning project (from a business standpoint) has 4 parts:

Project Setup
Data Pipeline
Modeling and Training
Serving (into production)

So, the author has analyzed the status of ML projects and applications from these 4 grounds. I am going to talk about some major takeaways I found interesting (which might not be new for most of the people in this domain).

The landscape over time

Before 2012 (Pre-AlexNet era), it was more like data analytics where the space was dominated by tools like scikit-learn, Apache Spark, and such.
Development Phase (2012-2015) revolved around data pipeline tools.
From 2016 to present, it’s mostly Deep Learning with the eco-system being hyped up extremely.

Still, this landscape seems to be under-developed. Most of the companies focus on [[Vertical AI], solving a very narrow problem. The tools developed within this space are mostly niche and business-oriented.

However, there is an increasing flow of tools around Machine Learning Operations.

ML Operations

There is a large space on ML Operations which is still in development. Only a handful of tools are available that provide generic ways to build machine learning models. Other than that most of the tools are simply oriented towards data pipeline. To have a perspective on this, there are only two major frameworks that allow training of deep neural networks - PyTorch and TensorFlow.
Other than that, there is still a large gap between doing research on ML models and putting them into production. Most probably, due to separate teams handling both of them. There is a huge opportunity to bridge this gap.
Nevertheless, a lot of tasks seem to have their own sets of research:
- Model Compression
- Hyper-parameter tuning and tracking
- Monitoring
- Data Labelling
- Deployment and scaling

This brings us to the last part of the AI Landscape - Open Source tools. The majority of the "bigger" companies are moving on to open sourcing some parts of their pipeline to pick up the hype train. For example Netflix's Metaflow, spaCy from Explosion AI, Google's colab notebooks.

However, open-sourcing these tools/frameworks are for two-way benefits:

First, it helps to expand the AI ecosystem and pushes the research more openly.
Secondly, latent intention to (strategically) shade lights on non-open-source applications. Say, Explosion AI has prodigy for enterprises, Google's colab notebooks boosting TPUs and such (you know how it is, right? :D)

So, what does it mean for an individual?

Probably, if you are in this space, you are either an ML Researcher or an engineer. But that's an overstatement. For small startups (like I am working in/for/with), this boundary is a bit intertwined. We seem to go in-depths during our experiments (but not that deep...more reason to feel dreaded about not knowing anything :/), and when the time calls we are simply putting things together, hacking, stitching the wings of a plane hoping that it flies. So, that's the path I guess.

Inside 2 Stoopid Triangles– 17 Years of ‘Chup Laga Timi

HallaNepal | 3 min

In the era of Nepal’s Civil War, this band did a ton of shit with Punk Rock. Quite fascinating.

Vishal: This album, to me, is a time capsule. We wrote about what was happening to us and around us. There are lyrics about the civil war, girlfriends, gangs, our hatred for local metalheads, our beef with other bands, everything—so much stupid stuff. A lot of it makes me cringe now.

Loved this album. It took me to the teenage period when I used to listen to Blink 182, Sum 41, Green Day…

~Listening~

Lex Fridman: Michio Kaku - Future of Humans, Aliens, Space Travel & Physics

Lex Fridman | Michio Kaku | 1 hour

I absolutely enjoyed this conversation that circled around civilizations (especially about [[Type IV Civilization]] leveraging dark energy), space travel, aliens, mind, and above all science.

I grew up watching Michio Kaku and various other people, speculating about the universe and existence. But then somewhere along the line, I kind of lost those things; I stopped watching/listening to the same old repetitive speculations. So, listening to this made me kind of nostalgic.

Nevertheless, I thoroughly enjoyed Michio Kaku's explanations and insights on many ideas, especially getting an insight into the scale of the speed of light; You could zoom around the solar system for a day and get back to your room in the evening.

However, the conversation felt one-sided and everything Michio was talking about oriented towards same old "TV Shows"-like dialogues. It was unnatural to hear him throughout the conversation, and Lex could barely conversate. Perhaps, Michio is asked the same questions repeatedly throughout the years? That might be the reason for the monotonicity of the conversation.

John O'Donohue - The Inner Landscape of Beauty

John O'Donohue | The On Being Project

I absolutely loved this conversation. I almost felt into a "meditative" state while listening to this. Discovering this very podcast was like witnessing a wonderful phenomenon.

John O'Donohue was an Irish poet, author, priest, and Hegelian philosopher.

John O’Donohue seemed so much articulated in his thoughts; the sensation is as if he is in front of you and could touch your soul through his beautiful view of the world and the inner beauty. It could perhaps be that he was highly influenced by the way he had lived as a priest for years, ultimately going into a more philosophical realm.

Beauty is all about more rounded, substantial "becoming", emerging fullness, a great sense of grace, a deeper sense of depth, homecoming for the enriched memories of your unfolding life.

Also, his calm voice reminds me of Alain de Botton from The School Of Life.

~Watching~

The Neighbors' Window - Oscar Winning Short Film

21 min

Beautiful.

For me, the window is a metaphor for our emotions of envy, jealousy, and wants. Our perceptions seem to be very malleable. Nevertheless, each of us has some parts in us that want to have an alternate life. This might be seen in how we try to compare our lives with others and feel more dreaded.

I guess our needs define our beliefs; how we want to see the world around us. Maybe it's better to not interpret renditions (and arts) like these. Truly marvelous.

I quit my PhD

Tibees | 19 min

I found this very inspiring. I can't comprehend how hard it might have been for Tibees to get out of that dilemma to pursue her dream for science education. The fact that life, career, and goals are not linear in their entirety and that you are fascinated by almost everything, makes it even more difficult to resolve the dilemma. Maybe it is about one's own Comfort Zone that breeds the passivity; the sort of mundane life you feel you have to get out of from.

Additionally, Tibees says that despite having not to think about Ph.D., she was still procrastinating the whole time even if the focus was to create her YouTube videos. Maybe, that's what everyone can relate to; procrastination kinda makes it so comforting for all of us I guess.

Nevertheless, I feel if you are definite about what you are going to do with your life and that your educational path like Ph.D. isn't so satisfying (or maybe work is kinda dull...or whatever it is), the thought of quitting should be assessed seriously. One should seriously think of other ways of generating income and being engaged.

For me, I am not sure. The feeling of this indefiniteness of life makes me more anxious. So, I am just going with the "flow" for now.

Animation: Astartes

Syama Pederson | 15 min

This is mind-blowing because the whole project is done by a single person. Damn.

Plus, the war tactics are so articulated.

The amount of detail put into each and every scene, with that level of fighting precision is bizarre.

Someone actually called him Shakespeare of the 21st century.

Loved this scene.

~Fascinating Things~

Website: Tools for better thinking

Untools is a collection of thinking tools and frameworks to help you solve problems, make decisions, and understand systems.

This HN Thread is also insightful.

Generative Artistry

Exploration of generative art.

~Poetry~

Everything Is Waiting for You

David Whyte

Loved it, especially the second verse.

Put down the weight of your aloneness and ease into the
conversation. The kettle is singing
even as it pours you a drink, the cooking pots
have left their arrogant aloofness and
seen the good in you at last. All the birds
and creatures of the world are unutterably
themselves. Everything is waiting for you.

~Music~

What Colour Is Your Raindrop - Tajdar Junaid

Beautiful. Happy-Sad

Also do listen to this Sofar Sounds performance. Enchanting.

Hungry Ghosts - Alone, Alone

Post-rock. Calm.

Darren Hanlon - I Wish That I Was Beautiful For You

5 min

Beautiful.

But a wish is just a pin
In a swimming pool of needles
If for one night you need the room
I wish to be your tomb
You can lie in me

Ending Thoughts

I want to end this “belief” of my newsletter with something Charlie Munger said:

"The ability to destroy your ideas rapidly instead of slowly when the occasion is right is one of the most valuable things. You have to work hard on it. Ask yourself what are the arguments on the other side. It’s bad to have an opinion you’re proud of if you can’t state the arguments for the other side better than your opponents. This is a great mental discipline."

Love,

Nish

PS: I’d really appreciate it if you shared your thoughts or better shared this newsletter with anyone who finds these things “interesting” enough. This has become a way for me to think aloud. Also, since I don’t have any social media, I think of you as a “meaningful connection” that can impact the direction of my writing too. But then, it might not matter. Everything is absurd if you go down the rabbit hole. Thank you for sticking to the end.