Inbox

Warning

This is where I brain dump any fleeting thought to make sure that I don’t forget it. They’re a glimpse of a seeds until I can process them into something coherent. I used to keep this list private, but decided that it’s useful to be fully transparent.

---

• John Ousterhout’s mention of “coherent systems are inherently unstable”, but I’m thinking more about security and stability in computer systems and not just biological systems. Homogeneous issues like CrowdStrike’s impact. Think about the relationships here and rephrase.

• Reputation systems, ranking algorithms, social media, gamification, eternal September

• Interconnectedness of attribution, identity, reputation, trademarks, importance of a name.

• The burdens of open source, entitlement, expectations, pressure, https://xkcd.com/2347/

• Photographic memory considered a curse, but you don’t want to not remember things either. Related to this site and thinking outside of my brain, it’s okay to forget things as it’s your brain’s way of compressing information, but also the halting problem…when does learning things at a deeper level become a bad idea for forward progress. Also related to “collector’s fallacy”?

• Mathematical perfection, attractiveness of symmetry, golden ratio, but not always a universal appeal, for instance, with piano tuning (see jacob’s piano is out of tune)

• Piano tuning, asmr, and synergy of the strings

• If it takes more time to sort a collection of items compared to the amount of time it takes to search through the whole thing, it’s not worth doing, because that’s not an optimization. My sock drawer made me think of the concept from Algorithms for Life book, but I need to look up exactly what it said again about the tradeoff. There was practical advice surrounding closet organization and has changed my habits to prefer LRU.

• iconology, radiation symbol for warning, universal understanding, simplicity of language, space probes

• anthropomorphism of ChatGPT, gender, politeness, actual effect on results

• My learning style, breadth vs depth, hobbies, just enough to be dangerous

• Inbox Zero (clearly I need work for this inbox)

• Allure of Tensegrity Structures. I’m curious if there are applications in disciplines outside of a physical field (like in graph theory or network optimization). Interestingly the term was coined by Buckminster Fuller, who also pioneered the field of Synergetics. Also tied to cellular automata and how simple elements can lead to complex outcomes.

• Jobs I’ve done…vanity, doc, smiths, customer service, telephone.

• Wolfram Research, AnyBot, remote work

• Why I’m an engineer?

• Serendipity and luck.

• Story of Dad’s half-brother, genealogy, Evergreen Cemetery in Troy Center

• Distilling of information into something intuitive is important for teaching (and learning?). Makes me think of Steve Mould’s videos like “You can’t bounce a ball under a table” where he uses a simplified discrete demonstration to explain a continuous more complex interaction (like Newton used for calculating orbits), and relates to first principle’s thinking. Jacky also wrote something about this.

• Story of quality and the goal of making 100 of something in a month versus making 1 of something in a month that is a representation of your “best work”.

• xkcd Ten Thousand https://xkcd.com/1053/ and taking things for granted as “common knowledge”. Too many things to learn and always new info. How else can we quantify added knowledge and how much we don’t actual know (or ever will)? Also related to importance of lifelong learning.

• Three interrelated topics on spelling alphabets, mnemonic encoding (pgp, signal fingerprints), and word lists. Clarity of communication (password character ambiguity), related to encoding schemes (Code book), cultures, internationalization. Licensing as well (Lexibomb word list).

• Preservation of knowledge or work (word lists), how do you make an archive discoverable? Related to lost knowledge, bit rot, and what is important to retain from history?

• Important design goals with CI surrounding spellout. Related to journey of using CUE, then reusable workflows, not needing local dependencies for a release process, etc.

• Fascination with precision and measurement being about relation to other things. Re-watch Adam Savage’s gauge block video, he touched on a lot of interesting points. Part of my appreciation of watchmaking and complex mechanical systems.

• Modern day life vs speed of evolution. Instincts regarding food, relation to The Comfort Crisis.

• The sunblock article and conspiracy theories.

• Learning at the right level of difficulty to push yourself; connection to progressive overloading for progress in exercise. Related to deliberate practice?

• Visualizing muscle growth while working out is actually effective for hypertrophy. Linked to making time estimates and visualizing the process required (versus focusing on the end result) leads to more realistic outcomes.

• The frequency of checking in on investments and the correlation with happiness. Conversation with Scott after a book, but I forget which one. Figure this out.

• Why am I drawn to words like “kaizen”, “ikigai”, and “wabi-sabi” to summarize concepts that may or may not directly translate into English? Is it purely length? Can I truly understand the concept as a non-native speaker, or am I misrepresenting something by not knowing the cultural context? Does it matter?

• ChatGPT as the ultimate rubber duck. Any time with no delay or ego, infinite patience, with the ability to ask for clarification and dive deeper into questions, be a sanity check.

• in code, every dependency can be viewed as a liability…this ties to trust, packaging ecosystems, reliability of access, open source, security, supply chain attacks, not invented here syndrome, but also to sharing reliable implementations.

• the problem of reinventing the wheel and it’s relation to lost knowledge. also tied to the idea of standing on the shoulders of giants, but how do we know those giants are there? we don’t know what we don’t know, so if some knowledge is gained, how do we distribute it most broadly to prevent others from making the same mistakes?

• the benefits of progress through disruptive technology are not always evenly distributed

• all knowledge in $\pi$ (https://github.com/ajeetdsouza/pifs)

• apex domains, www vs bare

• longevity of thought: how long does knowledge last or hold up? any such thing as universal truth? relativity vs quantum mechanics

• aside from history and philosophy, is there any reason to study earlier advancements

• bootstrapping technology or civilization, it’s relation to understanding stuff from first principles, charles petzold book

• link to ted chiang’s short story about ai versus humans and progress

• perfectionism being a problem, self-pressure, the idea of good enough, 80/20 rule, tied to personal finance journey as well, productivity, 4000 weeks book, the simple thing unintitively be more optimal (referenced in Algorithms for Life book), perfectionism = procrastination (related to wabi-sabi), find recent YouTube video on doing the most things.

• why static asset allocation for life. find recent article about research suggesting you should’t change asset allocation in retirement to be more conservative.

• learn in public, even though it’s for me, not commercializing, chronological organization killed curated creativity and exploration (Amy Hoy’s How the Blog Broke the Web; Stop Giving af and Start Writing More). Does the consistent structure imposed by a new tool halt innovation?

• summarize references of garden vs stream

• An “Inbox” is a mixed metaphor, but commonly understood. I think this can just be a note where I brain dump ideas (like this list) without having to think of note titles or “de-streaming” concepts. Or did I see a chatgpt plugin that might help create titles, so a dedicated note might not be necessary? I just want to reduce friction of capturing thoughts.

• ribozome structure of interconnected network, not a hierarchy (from author of Quartz)

• How do people logistically research history and collate facts? Recent example: Amy How’s How the Blog Broke the Web, but there are lots. Deep Work’s references, etc.

• timeless way of building, a place must be lived in and adapted over time to reflect it’s actual usage before it can become “alive”. similar to a desire path. you can’t plan everything out perfectly ahead of time, you have to make adjustments as you go. related to perfectionism/analysis paralysis/best practices.

• even following pattern languages, you can’t do it blindly without understanding the intent behind the rules.

• I don’t understand politics, lack of transparency, perverse motivations, common sense issues like inability to read laws before passing

• takeaways from the dictator’s handbook

• privacy is important, nobody knows all the rules, rules change, metadata analysis is powerful and could be used against you.

• what about times I’m not a stickler for privacy, like owning a tesla. Brad posted about court case https://therecord.media/class-action-lawsuit-cars-text-messages-privacy

• learning in public at odds with privacy? cancel culture, acknowledgment that I’m probably wrong, but open to hearing about it. scientific process is adapting understanding based on new information.

• appeal of minimalism, tiny house movement too extreme but lots to learn from.

• passive homes

• a pattern language; has made me find modern aesthetic less appealing, too clinical

• thoughts on stoicism

• thoughts on atheism

• can businesses refuse to accept cash? cash has a link to privacy

• moving from customized to tools with better defaults (defaults matter). vim → kakoune → helix, bash → zsh → fish

• related to the above, I love auto-formatting tools gofmt, rustfmt and prefer them to be opinionated with less configuration. eliminate bikeshed arguments, focus on doing the work/solving problems and not minor details. consistency is more important than 100% agreement on aesthetics. They ease my own cognitive load.

• optimize for doing/producing

• mostly flat organization, minimal structure, but maybe references could have folders (books, games, movies, videos, papers, etc.) Maybe just tags for that? What is the dataview plugin for querying, would that help to maintain things automatically?

• Something may be thought provoking, not necessarily something I agree with.

• what’s lacking is the federated aspect of easily sharing these ideas?

• self-directed learning for statistics

• Why Bayesian Statistics (vs Frequentist)

• self-directed learning for piano

• chatgpt and education ramifications; the cat’s out of the bag, must adapt to new tools. progression of spell check, grammar check, grammarly/hemmingway, chatgpt. how best to learn?

• should one try to minimize usage of idioms?

• chatgpt as a force multiplier (human plus ai…yes, I know it’s not ai). tapping into a high base level knowledge across literally all fields. can help make connections, be a good teacher, probe for deeper understanding, proof reader, helpful rubber duck debugger.

• long term ramifications of ai, artistic expression/creativity, lack of new training data, echo chamber, reversion to the mean.

• resonsibility to reduce bias vs accurate reflection of cultural biases

• ai replacement of jobs. universal income, collaborate to increase productivity, or adapt to better uses of human’s natural abilities? Link to “Wirth’s law” to explain the steady state of increasing our ambitions toward hard tasks by becoming more efficient?

• HOWTO Publish an Obsidian Vault for Free; this would be a typical old-me blog post, but I do need to write up what I’ve done for my own reference. Maybe it would become afruit, or maybe it doesn’t deserve as much polish?

• Do I not appreciate quotes as much as I used to? They’re by definition taken out of context, but can still be great. “Sucking at something is the first step towards being sorta good at something.”

• statistics are counter-intuitive; tied to interesting paradoxes, inability to assess risk, errors in conventional wisdom

• algorithms for life; practical applications for some mathematical rules of thumb.

• optimizing the wrong thing. personal finance journey to a one-fund portfolio, a single human error could wipe out any minor advantage, sort of like automating a process, beauty in the simplicity which ends up being more optimal. kind of like index funds versus actively managed funds.

• Where to capture pure HOWTO references or facts, like my copyright notices gist that I look up regularly. Things that I’d find useful to store as a personal knowledge base, but not necessarily original thoughts to expand upon. Is that even important to document anymore, considering chatgpt might be faster? Is it faster to search the broader internet than my own notes? Maybe if it’s a resource that I wish existed and couldn’t easily find?

• should index show list of recently touched pages? what about the pages that haven’t been updated in the longest? is there an ability to show a random page to help myself remember older ideas?

• comfort crisis ties into a lot of ideas; rucking

• a second brain; I’ve Googled for answers before, found them, realized that I wrote the answer and just don’t remember.

• How my memory works (figuring things out from my own perspective).

• following news is harmful; sensationalist, clickbait, made to stir up controversy and/or keep you in an echo chamber, overwhelming. people can’t estimate risk and think outliers are more likely than they are (crime).

• axiom of trusting the expertise in what you read while seeing obvious errors in topics you know deeply.

• pull ideas and raw notes out of old nvalt archive?

• how can we know anything…recall the book read after conversation with Scott.

• replication crisis in academic research

• any concepts tied to hobbies, photography, watchmaking, drawing, exercise

• value of personal metrics; are they actionable, a curiosity, or a waste of time

• the case for optimism…another book and Scott discussion.

• classic tuxedo

• short feedback loops

• graceful degradation…what doesn’t work when javascript is off?

• explore if writing is not shared, does it still have value? remember reading someone making that point that it should always be shared, but writing is how we can think through problems ourselves. that said, there are many examples of great discoveries from people’s notes only found after their death.

• lost knowledge, can old technology be adapted to new technology by understanding how it works. Thinking of Corridor Crew https://youtu.be/8q_HpSq95bw or https://www.bricetebbs.com/2023/11/09/varifocal-mirror/. Constraint breeds creativity. Think video game consoles.

• digital representation of analog mechanisms. great learning resource is the mechanical watch breakdown, allows you to visualize inside during simulation like cut-aways of physical items, but there’s still something visceral about handling things directly. Slide rule.

• what is it that makes something objectively terrible? person/place/thing might be all different. for people, is it external impact, intent, lack of remorse (sociopath next door), attitude, something else? what is the tipping point of redemption?

• cancel culture link; can we separate work from the person? statistics flowers data set. does appreciation of a work mean support of the person? what if they’re dead, does it still support them through notoriety, etc.?

• chatgpt used to infill connections between ideas (just like images).

• Tapping on Glass…universal way to seek attention. Is the woodpecker seeking my attention and I’m misunderstanding. Is there anything considered universal language with animals, maybe kindness? The Universal Greeting. Or just an idiot. What about other universal language (nuclear waste, space probes).

• Is We Learn Through Mistakes related to work organization and autonomy? We need the freedom to be creative, but the structure in how we work and the process can be more organized to make it efficient.

• Why I prefer a Digital Garden over a Zettelkasten…too much polish, the term “permanent notes” implies static and unchanging where I think having seed/sapling/fruit together is all fine.
• Keep a flat structure without hierarchy…promotes serendipity, less cognitive overhead in trying to apply and maintain a taxonomy. Tags can be useful, in moderation.
• Decision making…even smart people make terrible mistakes: https://fs.blog/smart-decisions/#smart_people_make_terrible_decisions
• Importance of boredom
• I don’t know where this link belongs…it’s similar to the point of using the seed/sapling/fruit tagging system, but more explicit. Epistemic statuses are lazy, and that is a good thing by Devon Zuegel