I have an exam in eleven hours and a quiz directly following it.
This week was a mix of me trying to catch up on a few weeks of material, talking through the concepts with Claude, and trying to take notes that I could later come back to read. I never do that last part.
I’m pretty confident I can pass both of these, but I’ve just been in a really low energy mood all week knowing these exams are situated between me and a good night of sleep come Friday.
The day started a bit late, waking up around 10am after accidentally sleeping in. I didn’t really have any classes to go to, so it was easier to shower, put on some clothes, and go to Starbucks to get some work done. I ate a really good bagel, which might surprise you, but in the same way people are ashamed to say they like Starbucks coffee the best (it is the best), they also refuse to admit their bagels are pretty okay. There was a small urge to buy another but the idea was shot down my brain.
I’ve been working out of a Starbucks or this other local coffee shop for a few hours everyday this week (and the week before it) but spent pretty much all day in and out of coffee shops for the first time in a while.
I knew if I sat at home there would be too many distractions. Twitter, my roommate, books, the internet, my phone, other apps on my laptop. I don’t usually get distracted and waste time but it’s much easier to stay disciplined working somewhere that isn’t the desk in your bedroom. The first study session of the day didn’t go horrible, but I spent most of this time replying to messages I’d let build up instead of studying, which wasn’t the best idea.
I took a lot of notes (obviously handwritten) on the Roman Empire’s economy and learned that there isn’t a lot of data we can use to prove most of what we know. The empire grew to be massive, but crumbled just as quickly. I’m not sure when this happened because I haven’t finished taking notes yet, but for the time they were an empire, the Romans seemed to be doing quite well.
The only real “data” we have that confirms this comes from measuring the distribution of fish factories across the empire, pottery manufacturers, varying locations of wine sourcing, shipwreck frequencies, and something to do with discarded animal bones. This was less interesting to me than it was a massive breath of fresh air. I’ve spent so much of this week just trying to make sense of the theories before I could really take time to read the history.
It’s my second-to-last semester as a student, and if there’s one thing I’ve learned, it’s that economic formulas (within papers or larger theories) are really overcomplicated. Most of these can be broken down into more simple explanations of a person or group wanting something, and the numerous variables that go into this calculation.
Everything is a utility function and these are stacked, combined, compared, or manipulated in some way to describe an economic situation. Even if most of what I’m reading about doesn’t ever play out in reality, it’s a much easier major than anything in STEM. “Studying” economics gives me more time to write on Substack and talk to anonymous internet characters on Twitter.
Each time I tried to open up a PDF and read from top to bottom, I got really confused over the graphs. I’m not saying I was completely unaware of what was being shown to me, but when you’re in a studying for an exam mindset, there really isn’t any time to sit back and genuinely examine the content.
You wouldn’t expect it because of how easy it is to read most graphs online, but there’s an unspoken rule that your economics paper won’t get published by the forces of academia unless there are no fewer than three really over engineered graphs for readers to ponder.
To get my point across, I googled “really complicated bad graph” and the following image popped up - it’s meant to give you an idea of what I saw anytime I tried to analyze a graph:
Economics isn’t super hard, but it’s a lot tougher when you’re being tested on it the next day. Sometimes I like to read about these empires and estimations of their GDP or living standards, but other days it’s really a chore to get excited about it.
Things aren’t all that bad.
They’re letting me bring a cheat sheet but this doesn’t help much, considering the formulas are pretty basic and the tougher part is cramming maybe 200-250 pages of information into my brain. I’ll still be bringing one except there isn’t any normally sized paper in my apartment and it’s too much of a task to get up and walk to the store now. I’ll just have to write really small on a sheet of paper maybe 3/4 of the accepted size.
The second coffee shop was a much more relaxing experience and gave me unlimited refills on their black coffee, which is probably a normal thing at coffee shops but came as a surprise to me since I don’t visit many coffee shops. I did more reading here, and when I was through some longer PDFs, it was time to read some non-school stuff I’ve been wanting to check out. This won’t be in any particular order.
I’d been meaning to check out near’s post on personality basins for months now. The decision to read it today came from a post I read the night before, describing the ability for LLMs to veer off into RLHFd personality basins. When I went to find the tweet for linking purposes, I suddenly couldn’t locate it, but I know for a fact it comes from this one account that draws really interesting visualizations of LLMs - we’ll be reunited again one day.
Near discussed how we are kind of like LLMs too, in the sense that our upbringing and physical reality RLHF us into different ways of being. I knew this was true, but the parts about altering your personality basin or shifting into another were really interesting. A good friend of mine wrote a tweet the other day about our neuroplasticity, which is relevant and can be found here.
I read a good post on less wrong about integrating data, the different methods of cleaning data, and why it’s all so complicated. The best part of it was this explainer on how Palantir works and why they’re so good at what they do:
To be honest, I’m not sure why I read so many random posts online. Like I said on here yesterday, most of what I read doesn’t offer me any advantage in writing about crypto - and that was previously the entire point of this blog! I guess this comes from a really unexplainable urge I’ve been feeling lately to learn as much as I can about everything I can before I graduate. This has resulted into a really large digital library of sorts in my Obsidian app that will probably keep growing until I’m forced to clean it up. Which reminds me of an article I read earlier today about reformatting Obsidian for productivity, if that’s something you’re into.
I also skimmed the beginnings of this 129 page, free PDF by Josh C. Baez titled ‘What is Entropy?’ and kind of confused myself. I’ve never taken a physics class officially, but have spent a normal amount of time to gain at least a baseline understanding of physics. As a quick aside, I’ll say that physics is much more enjoyable than math, even if I’m solving word problems (no idea if that’s what they’re called in the real world) in a math class - a good physics problem feels like a puzzle, which is kind of fun but usually never results in me solving it. I got pretty tripped up reading about entropy, but someone out there might gain some new knowledge from it.
There was a short amount of time spent reading a post about minerals and mining, which I really enjoyed. Not a lot of people know it, but I’m pretty interested in rocks and mining sometimes. Over the summer I visited Utah and bought an Ethiopian Opal for about $30 - parts of it look like the opal from Uncut Gems. Reading about the mining industry is fun because it’s an area of business nobody really thinks about.
There are entire countries - like Australia - who pretty much rely on rich mineral deposits for a large portion of their GDP. It says on Google that mining makes up roughly 13.6% of their GDP, which is pretty impressive considering how long people have been brutalizing Australia’s land. The post I linked is really good and I wish the author had kept writing these.
Lastly, I’ll share a post I’ve been taking notes on from nfx, which covers startups and incumbents in the “age of AI” as some people are calling it. I’m writing a longer post about this idea and how Y Combinator has a pretty impressive batch of companies despite what you might have heard on Twitter, but the general idea is that nothing is sacred and software is going to continue eating the world.
I read a bunch of other things too, but these were a few of the ones worth highlighting. After the coffee shop, I ate, went home, went back to the coffee shop, ate ramen, bought some kombucha, and ended up in the lobby of my apartment. It’s time to go off and finish studying, but hopefully someone enjoyed reading this and can make use of the links.
Talk to you later.