Skip to content

Instantly share code, notes, and snippets.

@ivan

ivan/2023_reading.md

Last active July 7, 2024 18:18
Show Gist options
  • Save ivan/5095670735ba941a6090a69fce4183df to your computer and use it in GitHub Desktop.
Save ivan/5095670735ba941a6090a69fce4183df to your computer and use it in GitHub Desktop.
Download ZIP
2023 reading list
Raw
2023_reading.md

[This page is best viewed with https://github.com/ludios/expand-everything, which will load all the comnents below.]

Wherein I try to prioritize reading for the limited amount of time I have this year, and to remind myself to read more than just comments on the Internet. Because of problems of time and shifting interests, I will consider this a success if I read a third of the list. I'll reflect on the reading and deviations from the plan in Jan 2024.

{+} = added after initial planning

  • Albert Camus - The Fall/ audio
  • {+} John Kennedy Toole - A Confederacy of Dunces/ audio, go to 6m44s to skip past the introduction spoilers
  • {+} pirate aba - The Wandering Inn/ audio
  • William Olaf Stapledon - Star Maker/ audio, go to 12m35s to skip past the introduction spoilers
  • Tae Kim - A Guide to Japanese Grammar
  • Noboru Akuzawa - Japanese Sentence Patterns Training Book for JLPT N5
  • Noboru Akuzawa - Japanese Sentence Patterns Training Book for JLPT N4
  • Jay Rubin - Making Sense of Japanese: What the Textbooks Don't Tell You/ the romaji is miserable; may have useful grammar insights
  • struggle through Japanese Wikipedia for some topics I know about
  • Daniele Minnone - A learning handbook for Joyo Kanji/ the first third, pg. 1 - 98

(my initial source for learning Japanese is https://cijapanese.com/ and not any of the reading.)

Lectures

maybe in 2024? not sure

  • {+} Paul Bourke - Fractals, Chaos, Self-Similarity
  • {+} Alex Komoroske - The Compendium / after I convert the Firebase export in code/websites/compendium-cards-data/db.json to a single HTML page
  • {+} James Betker - Non_Interactive
  • {+} Denny Britz’s Blog
  • {+} Robert Root-Bernstein - Discovering: Inventing and Solving Problems at the Frontiers of Scientific Knowledge
  • {+} Steven H. Strogatz - Infinite Powers: How Calculus Reveals the Secrets of the Universe
  • {+} Lexi Mattick & Hack Club - Putting the “You” in CPU
  • Lou Keep - The Uruk Series
  • Knut Schmidt-Nielsen - How Animals Work (via)
  • Edward O. Wilson - The Diversity of Life
  • James L. Gould, Carol Grant Gould - The Animal Mind (via)
  • Symbols and mental programs: a hypothesis about human singularity/ printed
  • Robert Yarham - How to Read the Landscape
  • Richard Powers - The Overstory/ audio
  • Rigdzin Shikpo - Openness Clarity Sensitivity/ printed
  • Michael R. Canfield (editor) - Field Notes on Science & Nature (via)
  • Sabine Hossenfelder - Existential Physics
  • George Soros - The Alchemy of Finance/ printed
  • Eric Gill - An Essay on Typography/ printed; I know he's bad
  • {+} Richard Hamming - The Art of Doing Science and Engineering

unplanned cool things read

unplanned and abandoned

  • Chuck Klosterman - The Nineties/ audio
  • Rick Rubin - The Creative Act/ audio
  • Mike Rinder - A Billion Years: My Escape From a Life in the Highest Ranks of Scientology/ audio
  • Sarah Steel - Do As I Say: How Cults Control, Why We Join Them, and What They Teach Us About Bullying, Abuse and Coercion/ audio
  • Benjamín Labatut - When We Cease to Understand the World/ audio
  • Kathryn Petras, Ross Petras - Awkword Moments: A Lively Guide to the 100 Terms Smart People Should Know/ audio
  • Adam Galinsky, Maurice Schweitzer - Friend & Foe: When to Cooperate, When to Compete, and How to Succeed at Both/ audio
  • Han Kang - The White Book/ audio
  • Niccolò Machiavelli - The Prince/ audio
  • Anthony Bourdain - Kitchen Confidential: Adventures in the Culinary Underbelly/ audio
  • Kristie Macrakis - Espionage/ audio
  • Christopher Winn - Legal Daisy Spacing (via)
  • Justin E. H. Smith - The Internet Is Not What You Think It Is/ audio
  • Alice Schroeder - The Snowball: Warren Buffett and the Business of Life/ audio (~77% in)
  • Morgan Housel - Same as Ever/ audio
  • Amanda Montell - Cultish: The Language of Fanaticism/ audio
@ivan
Copy link
Author

ivan commented Dec 17, 2023

leverage Screenshot 2023-12-17 013539

Kevin Mullet, Darrell Sano - Designing Visual Interfaces: Communication Oriented Techniques, p. 47

@ivan
Copy link
Author

ivan commented Dec 17, 2023

pros and cons of doing things

pros: things
cons: doing

https://twitter.com/keta_mean_/status/1733209721471983735

@ivan
Copy link
Author

ivan commented Dec 17, 2023

Programmers who can add functionality, without making the codebase unnecessarily complicated, are *incredibly* valuable. If someone has properly absorbed this aesthetic and knows how to do it, I would hire that person on the spot regardless of perceived lack of other skillsets.

I am on this team, but there is one drawback: this type of engineer/programmer tend to have acute complexity/bullshit intolerance so they avoid it for their own sanity. The drawback is that they rarely like working with a foreign codebase.

https://twitter.com/stephc_int13/status/1731722904419684839

@ivan
Copy link
Author

ivan commented Dec 17, 2023

traders Screenshot 2023-12-17 024428

Larry Harris - Trading and Exchanges: Market Microstructure for Practitioners

@ivan
Copy link
Author

ivan commented Dec 17, 2023

Hunter Walk recommends that folks avoid “snacking” when they prioritize work. If you’re in a well-run organization, at some point you’re going to run out of things that are both high-impact and easy. This leaves you with a choice between shifting right to hard and high-impact or shifting down to easy and low-impact. The later choice–easy and low-impact–is what Walk refers to as snacking.

When you’re busy, these snacks give a sense of accomplishment that makes them psychologically rewarding but you’re unlikely to learn much from doing them, others are likely equally capable of completing them (and for some of them it might be a good development opportunity), and there’s a tremendous opportunity cost versus doing something higher impact.

It’s ok to spend some of your time on snacks to keep yourself motivated between bigger accomplishments, but you have to keep yourself honest about how much time you’re spending on high-impact work versus low-impact work. In senior roles, you’re more likely to self-determine your work and if you’re not deliberately tracking your work, it’s easy to catch yourself doing little to no high-impact work.

https://lethain.com/work-on-what-matters/

@ivan
Copy link
Author

ivan commented Dec 17, 2023 •
edited
Loading

My most productive and probably my most valuable days are me just sitting in the backyard thinking about stuff. Sounds like something you would do when you're high, but I'm not high, I'm literally purposely and intentionally thinking about some topic. I might walk around the pool 500 times. Whatever I'm doing, I'm thinking about this one problem and that will lead me inevitably to other problems or other things I want to think about, and I've got a phone, and I'm just dictating or writing notes the entire time. And that is the best time that I spend. But a client's not going to pay you to think. They only pay you to do stuff. Especially for lawyers who bill time, it's "what am I going to do next? what am I going to do next? what am I going to do next?" and they miss the critical first step: "I'm going to think about what I should do, what's happening, how I can fix it." Those are the days where I really think I do the best job.

https://www.youtube.com/watch?v=Otq9wvcigtI

@ivan
Copy link
Author

ivan commented Dec 17, 2023 •
edited
Loading

Are there any studies done on the psychological aspect of how smart people run into common failure modes? Ones that spell out the common ways in which they dig themselves into ruts (preferably psychologically / financially) would be very interesting reads.

I think the big root for many of these has to do with a tendency to not rely on others when it comes to decision-making etc, overestimating how much can be solved by direct conscious solutions or control

https://twitter.com/Westoncb/status/1732527670402121996

Preferential use of abstract analysis is often useful when dealing with the many evolutionary novelties to be found in modernizing societies; but is not usually useful for dealing with social and psychological problems for which humans have evolved ‘domain-specific’ adaptive behaviours.

https://medicalhypotheses.blogspot.com/2009/11/clever-sillies-why-high-iq-lack-common.html via https://twitter.com/prathyvsh/status/1732554262851600574

@ivan
Copy link
Author

ivan commented Dec 17, 2023

Amazing how often agency is just a problem of knowledge. There was a time when I’d just never asked: “You mean I can just [create a festival / start my own support group / learn fundraising / learn how to socialize / found a magazine / run a class / go to a new place w no plan]?”

A lot of people are discouraged from taking initiative when young, and it sticks. Often, people who try to lead get envious rebuke just for trying, and people who get good at something are cut down with criticism. That kind of thing can stick for a long time, well into adulthood.

https://twitter.com/StupendousGrav1/status/1732411314759422410

@ivan
Copy link
Author

ivan commented Dec 17, 2023

I asked a student who always stresses me out to take the tension out of their throat before asking me a question and it worked :-)

https://twitter.com/_brigid_f/status/1732226528656638122

@ivan
Copy link
Author

ivan commented Dec 17, 2023 •
edited
Loading

Deprogram yourself from the idea that the more intense an experience is, the more healing it must be

once upon a time @bjtoomey said something along the lines of "a diffuse sense of confusion is the real indicator of therapeutic progress, not catharsis" and i've been chewing on it since

i interpret this to mean that confusion is an indicator that a part of you is at its edge, open to newness, ready to actually learn and change. the healing thing about catharsis in my experience is actually what happens afterwards, what you do with all the spare capacity that’s been freed

https://twitter.com/QiaochuYuan/status/1732147007966589201

@ivan
Copy link
Author

ivan commented Dec 17, 2023

are you in the right headspace to receive information that could possibly cause you to adopt opinions that are less popular than your current opinions

https://twitter.com/VesselOfSpirit/status/1731963644165161364

@ivan
Copy link
Author

ivan commented Dec 18, 2023

it is short, easy to read, and introduces crayon eaters to a slightly higher level of thinking

https://twitter.com/Idsbraam/status/1730989100788822225

@ivan
Copy link
Author

ivan commented Dec 18, 2023

Why does this meme work so well?

Part of the reason is the symmetry. There's something both topwit and bottomwit share that midwit doesn’t. But what is it? It’s not knowledge. He has that.

What they share is self-knowledge. Both the bottomwit and the topwit know what they do and don't know. The midwit knows what he knows but doesn’t know what he doesn’t know.

The midwit is foolish. That's the deep memetic payload of this meme: it carves out and illustrates the difference between intelligence and wisdom.

Being that the arrogance of the intellect is precisely how Satan preys on you there's layers and layers of depth here.

midwit GAQ5KjAXcAAHQ9H

https://twitter.com/nosilverv/status/1730571639954313315

@ivan
Copy link
Author

ivan commented Dec 18, 2023 •
edited
Loading

The FBI isn't supposed to use its most controversial spy tool to snoop on emails, texts, and other private communications of Americans or anyone located in the United States. However, that didn't stop the FBI from sometimes knowingly using its Foreign Intelligence Surveillance Act (FISA) Section 702 powers to conduct warrantless searches on US persons more than 280,000 times in 2020 and 2021, according to new disclosures.

[...]

Olsen described reforms that have already been implemented to prevent improper queries on Americans, which he described as "mistakes." For example, the FBI changed the default settings in its systems to force employees to "opt in" to querying Section 702 information, which helps prevent "inadvertent" searches. The FBI also now requires "specific, written justifications before accessing 702 information from a US person query," Olsen said. Previously, personnel chose general justifications from a drop-down menu, and now justifications must be case-specific.

https://arstechnica.com/tech-policy/2023/05/fbi-misused-foreign-surveillance-law-280k-times-to-snoop-on-people-in-the-us/

@ivan
Copy link
Author

ivan commented Dec 18, 2023 •
edited
Loading

Question: What did you say when Dexter Shoes came up? Were you for it or against it at the time?

Charlie: Well, I didn’t look at it very hard. The company, it was loved by all the retailers. It was the number one supplier to JCPenney. It had surpassed everything. It was a solid earner. It dominated Maine. They were nice people.

And of course, the Chinese hadn’t come up by that time and they just came up so fast. And they just took no prisoners in the shoe business. And they weren’t just cheaper by a little. They were half priced. And of course, the shoe business is not that easy a business. Of course, people bought the half-priced shoes. And the business just went to hell very fast.

[...]

I just think, if you just keep going, you’ll make some mistakes and, of course, you’ll learn from it. How could you not learn from that one? We learned how awful it is to have somebody who’s really way lower priced come in hard and how no amount of managerial skills could protect us.

https://novelinvestor.com/charlie-munger-dexter-shoes-handling-mistakes/

@ivan
Copy link
Author

ivan commented Dec 18, 2023

Qualcomm is a very active firm with over 24,000 active patent families and a patent portfolio of 140,000 global patents.

https://blog.withedge.com/p/the-secret-behind-qualcomms-margins

@ivan
Copy link
Author

ivan commented Dec 18, 2023

I feel like society genuinely just never talks about the problem of "What if there aren't enough of those?"

Most people want more great therapists: compassionate, well-researched, open-minded, unafraid of liability, willing to be paid relatively little, etc. Assume the optimal ratio of therapists in society is something unrealistic like 1 for every thousand non-therapists. Do we have good evidence there are that many people born with the inherent skill and attributes to be great therapists per capita? If not, then even offering free therapy college programs wouldn't meet the demand.

Most people want more great cops. Non-corrupt, free of bigotry, physically fit, willing to risk serious danger, able to make hard decisions in high stress situations, etc. Assume the same ratio, 1:1,000 cops to civilians. What evidence is there that enough people who fit that profile even exist per capita?

Ditto teachers, doctors, scientists, politicians, soldiers, everything we consider an "important" job. For so many of these, the solutions people turn to are spending more money, either for better salaries or better training, or less arbitrary red tape. And I think all of those could definitely be tried and lead to positive results.

But I always find myself wondering... what's the actual threshold we "should" be willing to accept? When does our expectation just become utterly unrealistic?

In some cases, it may be better to just have less if it means the people involved won't meet the bar. I could be wrong, but by my reckoning, really bad therapists and cops can do way more harm than longer waiting lists or response times. But in many others, it seems likely to me that this causes just as many problems as it solves. I think bad teachers should be fired, but what if firing all the bad elementary teachers ends up with 50+ kids per classroom? Even great teachers would be badly handicapped by that. And lawyers... I mean, there already aren't nearly enough public defenders out there, let alone "good" ones.

https://twitter.com/DaystarEld/status/1551805259831336961

@ivan
Copy link
Author

ivan commented Dec 18, 2023

In fact, we attracted no attention from anyone outside the vehicle whatsoever. A stretch limo would have gotten us more attention than this Jaguar SUV bristling with more technology than the Russian embassy—its spinning LiDAR, cameras, radar … all failing to turn heads. We had no driver for God’s sake. Surely, that is worthy of a second take? Yet no one was jumping in our path to test the AV,  no moms throwing their babies to safety. I see a passenger in another vehicle pulling up next to us with a camera up... but they are watching something on it and paying no attention to the second biggest San Francisco tech story of the year (the first is ChatGPT, silly). All this brilliant technology we were riding in [...] was being ignored.

https://www.engineering.com/story/two-short-rides-and-its-obvious-waymo-is-way-safe

@ivan
Copy link
Author

ivan commented Dec 18, 2023

The missing stair is a metaphor for a person within a social group or organization who many people know is untrustworthy or otherwise has to be "managed," but around whom the group chooses to work by discreetly warning newcomers of their behavior, rather than address the person and their behavior openly. The "missing stair" in the metaphor refers to a dangerous structural fault, such as a missing step in a staircase; a fault that people may become used to and quietly accepting of, that is not openly signposted or fixed, and that newcomers to a group or organization are warned about discreetly.

https://en.wikipedia.org/wiki/Missing_stair

@ivan
Copy link
Author

ivan commented Dec 19, 2023

RDF-star extends RDF with a convenient way to make statements about other statements.

https://w3c.github.io/rdf-star/cg-spec/editors_draft.html via https://twitter.com/meekaale/status/1736908386694832271

@ivan
Copy link
Author

ivan commented Dec 19, 2023

Dopamine has never been so easy for a hustler

https://twitter.com/Kr3py/status/1736967148835332220

@ivan
Copy link
Author

ivan commented Dec 19, 2023

Itsa me hash! bcc2564db34a20315686e3857c85093eb14541b354db0a4b02c5ef3b205188cf Should I get mysteriously logged out of Manifold, the first person to guess this hash is prolly me and should be treated as me!

https://manifold.markets/JoshuaB

@ivan
Copy link
Author

ivan commented Dec 19, 2023 •
edited
Loading

I want to try a new mode on my new phone... it's called 'spirit mode', it actually takes a picture of everyone's kind of spiritual essence.

https://www.youtube.com/watch?v=FzRczjFxsHo&t=1m24s

@ivan
Copy link
Author

ivan commented Dec 20, 2023

-You can pet the cats and dogs
-You can read the thoughts of the cats and dogs

https://steamcommunity.com/profiles/76561197975630210/recommended/1475810/ via https://store.steampowered.com/app/1475810/Ghostwire_Tokyo/

@ivan
Copy link
Author

ivan commented Dec 20, 2023

This paper introduces PowerInfer, a high-speed Large Language Model (LLM) inference engine on a personal computer (PC) equipped with a single consumer-grade GPU. The key underlying the design of PowerInfer is exploiting the high locality inherent in LLM inference, characterized by a power-law distribution in neuron activation. This distribution indicates that a small subset of neurons, termed hot neurons, are consistently activated across inputs, while the majority, cold neurons, vary based on specific inputs. PowerInfer exploits such an insight to design a GPU-CPU hybrid inference engine: hot-activated neurons are preloaded onto the GPU for fast access, while cold-activated neurons are computed on the CPU, thus significantly reducing GPU memory demands and CPU-GPU data transfers. PowerInfer further integrates adaptive predictors and neuron-aware sparse operators, optimizing the efficiency of neuron activation and computational sparsity. Evaluation shows that PowerInfer attains an average token generation rate of 13.20 tokens/s, with a peak of 29.08 tokens/s, across various LLMs (including OPT-175B) on a single NVIDIA RTX 4090 GPU, only 18% lower than that achieved by a top-tier server-grade A100 GPU. This significantly outperforms llama.cpp by up to 11.69× while retaining model accuracy.

https://ipads.se.sjtu.edu.cn/_media/publications/powerinfer-20231219.pdf
via https://twitter.com/deliprao/status/1737206517990875629
via https://twitter.com/mayfer/status/1737311228442423642

@ivan
Copy link
Author

ivan commented Dec 20, 2023

An Arachnophobia Mode has been added to the accessibility options.

https://portkeygamessupport.wbgames.com/hc/en-us/articles/16304407055251

@ivan
Copy link
Author

ivan commented Dec 20, 2023 •
edited
Loading

If you don’t look, you won’t find.

Charlie Munger, https://www.thegoodinvestors.sg/category/what-were-reading/

Ask, and it shall be given you; seek, and ye shall find; knock, and it shall be opened unto you.

https://biblehub.com/matthew/7-7.htm
via https://deoxy.org/evasion/12.htm

@ivan
Copy link
Author

ivan commented Dec 20, 2023

Another childhood friend recently found an old diary where on one page he had literally just written the word "remember," over and over.

https://benjaminrosshoffman.com/childhood-memory/

@ivan
Copy link
Author

ivan commented Dec 21, 2023

  • The ability to do speculative execution within failure contexts, meaning you can try out actions without committing them. When an expression succeeds, the effects of the expression are committed, but if the expression fails, the effects of the expression are rolled back as though the expression never happened. This way, you can execute a series of actions that accumulate changes, but those actions will be undone if a failure occurs in the failure context.

https://dev.epicgames.com/documentation/en-us/uefn/verse-language-reference

@ivan
Copy link
Author

ivan commented Dec 21, 2023 •
edited
Loading

Certain apps have always gotten special treatment. If it’s big enough to mess with phone sales they’re allowed nonsense a normal dev would be permanently banned for.

https://news.ycombinator.com/item?id=38721034

It’s worth noting that use of NEHotspotHelper requires a special entitlement (com.apple.developer.networking.HotspotHelper) that you have to apply for, and presumably Apple won’t grant unless your app has a legitimate need for it.

That said, this maybe shows an incompatibility between Apple’s privacy strategy and “super-apps” like WeChat and AliPay. When a company shoves all functionality into one app, that app suddenly has all the entitlements, and it’s harder to tell when and how any sensitive data is being used.

https://news.ycombinator.com/item?id=38721115

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment