Week 7 / 10 Feb 2025
In feudal Japan (1185–1868), a rōnin (/ˈroʊnɪn/ ROH-nin; Japanese: 浪人, IPA: [ɾoːɲiɴ], ‘drifter’ or ‘wandering man’, lit. ‘unrestrained or dissolute person’) was a samurai who had no lord or master and in some cases, had also severed all links with his family or clan. A samurai became a rōnin upon the death of his master, or after the loss of his master’s favor or legal privilege.
In modern Japanese, the term is usually used to describe a salaryman who is unemployed or a secondary school graduate who has not yet been admitted to university.
Questions I’ve been thinking about for making decisions:
- What would I want to do if I knew it’d work out?
- What would I want to do even if I knew it’d fail?
- Is there something that’s on both lists? (Where both the happy and sad case are desirable, for whatever reason) - If yes: This is something that I must really do. Figure out a way to try it out and see what happens.
The latest decision I made with this was the choice to take a gap year (more of Apr 2024 - Jan 2025 but you get my drift) to just do whatever I felt like doing that wasn’t employment.
Sort of life update: I am taking on my very first SWE intern role at a startup. I’m nervous but hopeful.
Week 5 / 27 Jan 2025
Links
Spectrum IEEE - How Claude Shannon Helped Kick-start Machine Learning
I first encountered Claude Shannon recently while reading the Nand2Tetris textbook.
- Drawing an overly close equivalence between human intelligence and AI risks succumbing to a functionalist perspective, where people are valued based on the work they can perform. However, a person’s worth does not depend on possessing specific skills, cognitive and technological achievements, or individual success, but on the person’s inherent dignity, grounded in being created in the image of God. […]
- Furthermore, there is the risk of AI being used to promote what Pope Francis has called the “technocratic paradigm,” which perceives all the world’s problems as solvable through technological means alone In this paradigm, human dignity and fraternity are often set aside in the name of efficiency, “as if reality, goodness, and truth automatically flow from technological and economic power as such.” Yet, human dignity and the common good must never be violated for the sake of efficiency, for “technological developments that do not lead to an improvement in the quality of life of all humanity, but on the contrary, aggravate inequalities and conflicts, can never count as true progress.” Instead, AI should be put “at the service of another type of progress, one which is healthier, more human, more social, more integral.”
Scott Stevenson - Lego Mindset vs Woodworking Mindset: On Accepting Toil
I am under the English Channel. I’m squinting into the sun. I’m in Swindon. I’m on my way back from the Arctic. I am putting myself at risk of being sued. […] I’m back in the UK, and I have a cold.
I got a copy of Elements of Computing Systems for free. Delightful :)
Week 4 / 20 Jan 2025
Links
surma: Ditherpunk — The article I wish I had about monochrome image dithering DeepSeek-R1 has performance on par with OpenAI o1 and is MIT licensed - thank u Dominika for always having the coolest links
82mhz writes: I will never need to buy a new computer again - saw this because arne shared this on mastodon
- good for me to keep in mind when I get the itch to ‘upgrade’ - my laptop is working fine! I don’t need the physical and financial resources that a new laptop would demand.
- The original implementations from the 80s were programmed with Microsoft Basic and outputs were drawn on a Canon X710 plotter.
issung posts a whole bunch of photos from late 90s-early 2000s LAN Parties on Internet Archive
Yapping
I attended a police auction preview. It’s something that’s been on my list for a long time, but I only got around to it this week. The preview was held at the Police Logistics Base at 1 Hemmant Rd, which is a place that I never knew existed until this week. It’s in Paya Lebar, down the corner from where City Plaza is. The assets are handled by ST Logistics, but the auctioning process is outsourced to a company named AssetHub (which is really what it says on the tin - a hub for assets). The preview allows you to take a look at what is on offer – in particular, I wanted to see Lots 60-62 which involved cameras with vague descriptions on the catalog such as “Black Leica Camera” (which was really a Leica C, and not the M-series).
After circling about the rack where the cameras were, I accosted one of the hapless ST Logistics staff members on duty and asked a burning question: Are any of these assets confiscated from the landmark $3B money-laundering case?
To my great disappointment, he says no – most of the items here are actually unclaimed items from the Lost & Found department. I don’t know how people lose Raymond Weil watches, DJI ultralight drones, cameras and jewellery, but this is apparently a thing that happens. And it happens often enough that there are enough things to organise into lots for auctioning. I ask him if these previews are largely attended by regulars and he says that he’d estimate about 80% of the turnout to be regulars. There are two guys walking about with loupes and flashlights to inspect the jewellery lots. They are regulars.
One of the Jewellery Guys says that they come here regularly to look for jewellery and watches, but they’ve given up on selling watches directly to consumers as it’s too much trouble to provide a warranty and deal with the cost of servicing it if it spoils within the warranty period as the cost of servicing may sometimes exceed what they make from a watch. Instead, they sell to watch dealers directly.
While leaving, I see a bunch of bikes that are also part of the auction and realised that these bikes used to belong to the Singapore Police Force. The SPF logo is scratched out on most of the bikes, but enough of it remains to identify these bikes as decommissioned police bikes.
Week 3 / 14 Jan 2025
This year, instead of having multiple posts for weeknotes, I have decided to make one megapost that I update periodically. The point is to have a record of roughly what I was thinking about in each week / links I liked. Also, one md file is really easy to process if I ever want to do any kind of NLP or glorified regex mumbojumbo on my own writing (e.g. ripping every link I mentioned in 2025 and categorizing it). Also, reading this requires periodically coming back to check on this post like a Tamagotchi, which will hopefully decrease the (already tiny) potential readership of this post, and hence encourage me to censor myself less.
Links
Nice explainer from Evan from SF Compute about creating markets for GPU compute
Anker introduces a solar beach umbrella that uses perovskite cells - thank u Dominika for introducing me to this :D
Solar panels that layer a perovskite film on top of a silicon base can absorb even more light — perovskite cells can be optimized at the blue end of the light spectrum while silicon cells harvest energy on the red end — to reach solar conversion efficiencies of up to 43 percent. s
User Robert Evans on bluesky:
after days of grifters and con men at CES we stumbled upon the booth for VLC. they were all dressed as wizards and told us, “we have nothing to sell, we just decided to show up”. i told them I’d been using their software to pirate media for 15 years and they said “keep doing that”
Music
Caroline Polachek covers Radiohead’s True Love Waits
Magdalena Bay - That’s My Floor
Text
Flabbergasted and deeply upset to learn that trying and facing rejection is actually a skill that needs practice. I tried a whole bunch of new things in 2024, but they didn’t involve the potential of rejection or shame. It made it easy to roll up to things like silversmithing workshops or start growing mushrooms out of nowhere, but I hadn’t tried much in terms of risking the chance of rejection.
This kind of avoidance is probably why I find the job hunt so difficult, and it’s made even harder by the fact that I would like to find a startup to join. This involves having a degree of agency and Teflon-like resistance to rejection that I don’t recall having besides that one time in 2018 where I was fuelled by a great deal of delusion. Right now, I spend ages trying to solve data structure and algo problems marked “Easy”.
When I feel silly, I think back to 2018-19 and how brave / borderline delusional I was. Tapping into a sense of unbridled agency, if you will. In that era, I got to work a shift at Native. I mostly remember faithfully cutting lotus leaves into circles - the bar has a sustainable / regional produce sort of slant, and they were using lotus leaves as coasters as they’re hydrophobic. I got to try this out because I randomly sent them an email that basically said “I’m probably kind of useless but am looking for work and used to work as a waitress at Ce La Vi. Can I come and try it out?”. I ended up being bad at it as they didn’t call me back for another shift. Looking back, I was probably way too anxious and nervous about work, which probably prompted the bartender who was stuck with managing me to say:
Look, you’re scared about only being here for a few hours but it doesn’t matter. By the time they step in, the customer’s only been here for 30 seconds. You will always be more experienced than they are. You need to chill out.
(or something along those lines.)
I think about this encounter whenever I have to do something where I feel too new and unqualified (e.g. teaching for Code in Place).
Another time where I felt like I really tried was when I walked in to the law library to ask if they wanted to hire people to shelve books. I had heard through the grapevine that they were paying students $12 an hour to do this, which sounded quite lucrative. She said no – they had enough help already – but I could leave my number and she’d text me if anything came up.
Whenever people make offers like this, I usually think they’re half-joking and won’t ever actually contact me. But she did, twice – once for a summer job at the campus Central Library (which I ended up not doing as I was interning), and another time she didn’t even text. She just called me on Whatsapp, then sent me the email of a professor from the public policy faculty. She said he was looking for a Research Assistant (What does he do? I don’t know, you have to ask him). I don’t know how effective this is as a lead generation tactic, from the professor’s point of view. I eventually got the job. To this day I strongly suspect that I was the only person who was aware of it.
I try my best to lean back and lean into that version of me. Something prospecting, curious, unfazed.