2021 into 2022 - A Year in Review
life learning01 Jan 2022
18 minute read
Table of Contents #
This post is a sequel to last year’s post, my recap of 2020.
If you want to talk to me about any of this, chat, email, or have any recommendations on anything, email me or DM me on Twitter.
Doing yearly recaps helps me observe my progression in many areas, and keeps me accountable for next year. I share things I’ve done, learned, read, and often sprinkle in some introspection on recurring motifs that have animated my year. I also included a best of twitter.
Like last year, I also include miscellaneous content recommendations, and a recap of changes to my to my public wiki.
Learning / Knowledge #
Note: I’m tracking what I learned this year by simply looking at the version control of my personal knowledge base.
In this section I explain what I learned, what I planned on learning, and also made a little recap of some new additions. This is a general overview of ideas and subjects that I’ll develop in other sections.
What I wanted to learn #
From last year’s post (grade out of 10 for my progress):
Get better at:
- Haskell / Rust (7/10)
- Competitive Programming (7.5/10)
- Natural Language Processing (7.5/10)
- Turkish (4/10)
Learn more about:
- Writing Well (9/10)
- Learning how to Learn (9/10)
- Mathematics (I have an eye on a few books I’d like to read) (8.5/10)
What I’ve learned #
- philosophy: in the last year of high-school where I am the core curriculum has 4hrs of philosophy a week, and I’m really liking it. More in the books section.
- natural language processing and AI: I wanted to learn more about this, and I did. I had been reading Prof. Jurafsky’s book, but decided I need something that had more balance between theory and practice (the book was very dense and useful, but too removed from practice for it to stick with me). I did the fast.ai Deep Learning for Coders course, and would like to do their NLP specialization this year. Thoughts on the course here.
- education: have started researching and writing my thoughts on education systems, will write a post on french education soon.
- meta: finished the great coursera learning how to learn course
- Haskell / Rust: finished the Get Programming with Haskell book. I now have a solid base of knowledge of the language, but lack experience or practical projects.
- physics: self-learned some more classical mechanics, and I have quantum mechanics books on my reading list
- competitive programming and algorithms: did quite a bit of competitive programming, although less than last last year. I got in the final of the national computing olympiad and attended a one week camp this summer. I learned a lot, but don’t think I’ll devote much more time to these competitions.
- cybersecurity and capture-the-flag events: capture the flag competitions are events where teams compete to solve hacking puzzles. Our team has been way less active due to the end of lockdown, but we still competed and I learned quite a bit. (we ranked #2 at trollcat CTF 2021, and #9 at UmassCTF). I competed in the [selection for the French national cybersecurity team], and ranked #18 in the junior category. I almost hit the cutoff for the next stage, but will need to learn more cryptography to really get to the next level.
- turkish: I felt duolingo wasn’t effective and just stopped. I need to find mechanisms / people to learn with.
- mathematics: I’ve self-learned quite a bit here, and will develop in the books and events sections.
- rationality && effective altruism: I knew nothing about this! Am discovering the community, and it’s been great!
- my constrained learning (what I learned for school)
- lots of history / geography, interesting case studies on russia / china and the 1929 economic crisis.
- generic math & physics curriculum
- french and english litterature (see books)
- ACT practice to get into college
- etc…
What I want to learn #
This is intended to be a list of things I want to learn, and I know I won’t do many of them.
- NLP: go deeper, learn more about transformers and text generation / question answering, ideally do the fast.ai NLP specialization
- mathematics: statistics (useful for above), continue what I’ve been studying in linear algebra, and pursue interest in cryptography. I also need to persevere in competitive mathematics, and practice for olympiads coming up soon. Also want to learn more game theory.
- turkish: set up habits with family to speak turkish (hard), maybe find courses I can attend, or people to talk to
- physics: review notes on classical mechanics, and learn QM and quantum computation (read quantum country)
- education: I need to learn by writing, and publish the drafts I have lying around on this.
- Haskell / Rust: implement practical projects
- cybersecurity: do more solo training, and motivate my team to do some competitions. I need to get better at saving the insights or learning I do during events.
- economics: I might look at marginal revolution university
Projects #
What I wanted to do #
From last year’s post (grades on how well I did):
- Consolidate Archivy and stabilize the project (10/10)
- Work on a project related to content discovery/consumption and the current state of finding content online. (7/10)
- Implement a project for a solar energy idea I’ve been having (0/10)
- investigate compression of software packages specifically in Pypi (0/10)
- maybe write a simple, minimal newsletter manager centered around links (0/10)
What I did #
Archivy #
Archivy - my open source knowledge base software focused on customization, good search and integration with the web, has progressed well, and most notably:
- bookmarking and saving web content is now easier and you can specify rules for certain websites, thus providing your own customizations to archive specific websites.
- I added a more lightweight setup where Elasticsearch isn’t necessary for search
- implemented bidirectional links between notes, ie you can link to other notes inside your notes and then see what’s linking to what.
- added image upload / downloading images from archived webpages
- general improvements to general user interface and design, for editing, configuration and search.
I worked on a few new plugins:
- static site generator to take your personal knowledge base and host it publicly online. I use this for my wiki.
- firefox integration
Also improved the following ones:
- archivy Hacker News downloader - better error catching / caching, now works with docker.
- archivy Pocket downloader - same as ^^
Archivy is also now on the AUR and in NixPkgs.
In terms of open source community, I’ve been very happy about contributions and especially appreciated @edditler’s help, and @clemux’s support.
Espial #
Espial is an English word that means discovery.
I have been working (in private, for now) on a project named Espial, to enhance and augment organization in knowledge management software, using Natural Language Processing. I have a blog post on the way about this, but in the meantime, here’s some of the stuff I’ve shared online:
Here's a teaser for one of the features of my project: this knowledge graph was auto-generated by my software, where the red notes are notes and articles of my knowledge base, and the white are "concepts" -- ideas or themes detected by my program. pic.twitter.com/shyr2wB9qW
— Uzay-G (@uzpg_) November 11, 2021
There’s also this thread, where I downloaded all of the articles I had liked and asked people to give me links they had read. For each link, I fed it into Espial and then combed through the results to see if it found similar, relevant posts, which I then shared. (Spoiler: It worked!):
I wanna run an experiment - if you reply with an article you liked, I'll send you an article (or several) loosely (maybe very loosely) related to it.
— Uzay-G (@uzpg_) December 19, 2021
Other Projects #
Made a discord bot analytics platform, dabbled in synaesthesia with a project to combine poetry and music, worked a bit on AdiosCorona (illustrated articles on the COVID pandemic), and thought about decentralized content discovery.
What I want to build / do #
- launch Espial, build Roam and Archivy integrations
- improve Archivy, democratize use to non technical users and investigate federation.
- improve music organization on spotify, ideas similar to this thread.
- Abstract goal: find entirely new domains / directions I want to build towards
Writing #
I published 6 blog posts this year, for a total of ~8k words, compared to ~11k in 2020 and ~9k in 2019. I have way too many drafts I didn’t publish for lack of time / direction, and a lot of my content is just posted on my wiki, where I’m less stringent on producing good content. I need to fix this, and have some posts that are ready to publish.
Thanks to @kewbish for giving me feedback on what I have been writing.
What I wanted to write #
From last year’s post:
- Learning Haskell: Perspectives coming from OOP and quickly highlight the very neat ideas Haskell introduces that are pretty much non-existent in more traditional languages (didn’t do)
- A Case for the Web: I’ve been seeing a lot of posts about fundamental problems with the Web and its implementation. These bring up a lot of good points but I think many of the ideas that make the web so powerful are left ignored or presented as a lesser trade-off. (didn’t do)
- Overview of how I organize/process new knowledge (done!)
What I wrote #
- my last year’s recap
- social overfitting - a metaphor between the positive feedback loop of online production and AI overfitting.
- How I manage my Knowledge
- An article a day for 2021 - an article on community genius, and a description of a challenge I had done to send my class an article each day in 2021. In the end I didn’t last the whole year, but had a good run.
What I want to write #
- public book reviews: I need to figure out how to do these well. Currently have drafts for 3 book reviews in my files (The Selfish Gene, The Unconsoled, and GEB)…
- an overview of the french education system: I’ll explain its structure & specifics, and tie it into a general comparison with other education systems.
- My vision of how AI and web content can change knowledge management. I have a written draft, and am just waiting to launch my related project before I share (Current title: A New Approach to Knowledge Management, using Natural Language Processing and the Internet’s abundance of content)
Other Goals #
Past #
Last year’s post:
I’d like to do some more climbing and see some of my close friends I’ve never met in real life (if covid allows). (no climbing, but lots of hiking and running, and I did see one!)
I’m also hoping to obtain an internship this summer (in research or industry), which should be a fun and educative experience if everything pans out. (nope, did other fun stuff)
I look forward to also getting into writing poetry as a creative outlet, as I really enjoy producing the odd poem here and there. (did this + lots of private ones)
Goals for 2022 #
- get into a university I’m happy with (have been spending lots of time on college applications)
- learn more piano
- go to more museums and concerts, if COVID allows
- general possibility expansion, as I discover new paradigms and ways of thinking or creating, as I have in the last two years.
Miscellaneous #
Twitter #
I started actually using the twitter account I had created in June 2020. Twitter can be a very high-value social space, but I need to be careful not to overdo it or procrastinate. Have been meeting and talking with cool people, like judah. I wrote ~400 tweets.
I tried predicting how many followers / following I’d have in a year. As of writing this, I am at 106 followers and am following 305 accounts. I’m 95% confident I’ll have between 140-5K followers next year, and 50% confident I’ll have between 200-900. These predictions are just fun, I don’t care much about this metric.
My “best” tweets: #
- I do some introspection and try to present who I am, my values, and my projects
- Predict your followers / following thread
- I recommend similar articles to ones commenters shared, using some AI and human curation
- Book review of Kazuo Ishiguro’s “The Unconsoled”
Camps Attended #
SPEAR - Summer Program in Experiential and Applied Reasoning #
This program is similar to SPARC. I met some amazing and kind people. Gavin Leech and Chana Messinger actually pushed me to start Espial! SPEAR had similar effects on me as the Google Code-in community (described in last year’s post): expanding my knowledge / belief in what I could do and what was possible. I was exposed to many different interesting communities, like the effective altruists focused on how to best improve the world, and the rationalist / longtermist communities.
Mathematical Summer in Paris #
Mathematical Summer in Paris was a great experience - it’s a week-long, international camp for 16-20yr olds, and its focus is higher mathematics. I learned TONS at this event, although it was more breadth than depth. Notes are here. Especially liked the talk on different types of geometry entitled “On a woodcut by MC Escher”, by Yves Saint Aubin. Great talks on cryptography, Diophantine equations, persistence theory and quiver representations (this was confusing but amazing), and Natural Language Processing. If I can, I want to do this next summer.
Random lists and updates to my knowledge base (articles, quotes, words, etc…) #
- New beautiful words, discovered a lot this year.
- my digital attic of great articles. I haven’t updated it in the last ~2 months, will get to that soon.
- Quotes
- Poems I enjoyed
- began indexing some markers for the music I like / want to share here - work in progress.
- decided to start adding recipes
Books read #
Non-Fiction #
- The Theoretical Minimum: Classical Mechanics - Leonard Susskind [physics]. Notes here, great introduction. (10/10)
- Get Programming with Haskell - Will Kurt [functional programming]. Notes. Also great. (9/10)
- Doing Good Better - William MacAskill [effective altruism]. (9/10)
- Finite and Infinite Games: A Vision of Life as Play and possibility - James P Carse. I loved this book and its presentation of “infinite games”, when people contribute to something that will outlive them, and ideally last forever, like culture. (9/10)
- Gödel, Escher, Bach - An Eternal Golden Braid by Douglas Hofstadter (2021). Dense book covering many intriguing topics in math & CS. Hofstadter has amazing skill at finding links and motifs in wildly different notions, and tying them together into an impressive book. [mathematics, logic, and cognition / AI] (8.5/10)
- Etonnement Philosophique (Translates to “Philosophical Surprise”) - Jeanne Hersch. In progress, great summary of philosophy (9/10).
- Stranger in the Village - James Baldwin. Great argumentative essay on race in America, studied in school. (9.5/10)
- Working in Public: The Making and Maintenance of Open Source Software - Nadia Eghbal. In progress.
- Linear Algebra: Ideas and Applications - Rchiard Penney. In progress.
- Schools on Trial: How Freedom and Creativity Can Fix Our Educational Malpractice - Nikhil Goyal. Too melodramatic in criticizing current education without much nuance, although it does present new, interesting countercultures in education. [education] (5/10)
Fiction #
- Exhalation & Story of your life, both short story collections by Ted Chiang. Absolutely stunning, these stories are moving. (10/10)
- Ubik by Philip K. Dick - Pretty mindbending and interesting twist on the whole “(distorted) perception of reality (what even is reality?)” theme. (10/10) [scifi]
- La Peau de Chagrin (The Wild Ass’s Skin) - Balzac. Really made me think more about the balance between my desires, and what was actually right for me and wouldn’t damage who I am. Beautifully written (review and quotes here) [genre: french “roman”] (10/10)
- The Unconsoled - Kazuo Ishiguro. The perpective created in this novel is stunning and confusing. It was a weird, bizarre, frustrating, but highly rewarding read that you follow through the eyes of the main character, a renowned pianist visiting a bizarre European town. Instead of any sort of linear narrative, the story sets up the narrator in a setting magical, not in the sense of fantasy or any actual “magic”, but in the way suddenly time and space just fit or extend through the narrator’s journey, to his advantage and encumberment. Ishiguro constructed so many deeply interesting and creative characters who revel in this chaotic and airy writing, through the eyes of a narrator who embodies all at once a stranger and a friend. (9.5/10) Notes
- The Anomaly - Victor Miesel - discovered this because it won the french literary Goncourt prize, as a science fiction novel. Solid book, some beautiful passages and a general theme / plot that I found satisfying. Some descriptions / moments seemed a bit unnecessary sometimes, and although I understand the change in tones (going from eloquent prose to almost prosaic conciseness) was intended, i didn’t always like it. More comments here. [scifi] (8.5/10)
- Peace - Gene Wolfe. Confusing, and complex work about a man near death who revisits his entire life, dashing from event to event in a fantastical way, without any respect for time. Beautiful (8.5/10) [scifi]
- The Handmaid’s Tale - Margaret Atwood: describes an authoritarian dystopia founded around puritan principles in a world with low fertility; where a subset of women called “handmaids” are forced to bear children with men they don’t know. Well written, and unpredictable. (8/10) [scifi]
- Wide Sargasso Sea - Jean Rhys. Beautiful passages, this novel rewrites Jane Eyre’s story but takes the point of view of the “crazy woman in the attic”. Powerful discourse on patriarchal forces and colonization, love. (8/10)
- Les Animaux Denatures (You Shall Know Them) - Vercors. Interesting philosophical perspective on the issue of what makes us human, after a new species with human characteristics is discovered and exploited. (7/10)
- Howard’s End - EM Forster. This novel is long, winding, and sometimes airy, but it is a beautifully written story of class issues in 19th century England. (8/10)
- The Glass Menagerie - Tennessee Williams. Beautiful play (9.5/10)
- Le Rouge et le Noir (The Red and The Black: Chronicle of the 19th Century) - Stendhal. Frustrating read, as Stendhal’s character is full of contradictions and yet keeps us enticed such that we still read the whole novel. (7.5/10)