Bounded-Error Log

When I was very young, I thought cooking was easy. I sliced plastic vegetables with a toy knife and then Velcroed them back together, ad infinitum. For at least some time, I wanted to be a chef when I grew up.

When I was slightly less young, I thought cooking was hard. My reference points were mostly (1) my parents, who seemed to know how to make a million different dishes in inscrutable ways without thinking, and (2) MasterChef contestants (who I assume were better at cooking than my parents because they were, well, on MasterChef) messing things up and getting kicked off the show.

Now, I think I probably elided some meaningful distinctions there in my youthful naïveté. Cooking food that will keep you from getting kicked off MasterChef is hard. Cooking edible food is easy.¹ Cooking storebought dumplings in particular is so stupidly easy it’s unfair. More generally, though, most recipes tolerate a lot of substitutions,² number fudging,³ and even straight-up skipping pesky instructions, like the ones in baking recipes where you mix two sets of ingredients separately in specific orders. There are reasons for those steps, but ignoring them and dumping everything into the same mixing bowl usually won’t make your results inedible. You can also just decide to omit ingredients you don’t like. Probably the least tolerant ingredient measurements in recipes are the measurements of baking soda or baking powder, which by the way are different things, in baking recipes. But otherwise you’d really be surprised how many corners you can get away with cutting — I’ve even completely winged one baking soda/powder measurement with decent results. I think this is especially important to know for people from technical backgrounds like me, who have an instinct to treat the numbers in recipes as precisely measured, painstakingly optimized choices to produce the best dish. They usually aren’t, and even if they are optimized for the recipe author’s palate, they probably won’t be optimized for yours.⁴ And they certainly aren’t optimized for any tradeoffs you might want to make between food quality versus the time and effort you’re putting into cooking. Make the tradeoffs you want. You’re not on MasterChef.

There’s some oft-cited psychology studies that suggest that once your salary goes above $75,000, additional money doesn’t make you happier. This sounds like a sage bit of life advice if it were true, the ultimate rebuff against excessive greed and materialism and sacrificing other things for a six-digit salary, but it overstates the case a bit. 80,000 Hours’ analysis of money and happiness is probably the analysis I’d trust the most here; I think it would be more accurate just to say that you get diminishing returns of happiness from salaries above $70,000.¹ Still, that was enough for me to decide fairly early on that I wasn’t interested in trying to get a high-paying job for its own sake, or in spending too much effort trying to invest my way to a fortune.² I wanted my job to be personally satisfying and good for the world, while paying enough for me and my family (current and future) to get by, but I planned to treat any additional money after that as little more than a bonus used for breaking ties.

I still mostly stand by that decision today, but over the intervening years I realized there were a whole host of reasons to want money that weren’t that selfish at all.

I can’t begin to count the number of times we were exhorted in high school to go to class. College is different, they said. Nobody is going to force you to go to class any more. This is what you came to college to do, what you paid so much time and money for. It’s on you to make sure you’re learning.

By and large I followed this advice, until I considered that I might have overcorrected given the exhortations. There are a lot of definitely bad reasons to skip class, chief among them being too lazy to get out of bed. There are also some non-obvious reasons to go to class, such as getting the professors to recognize you — this is a reason to go to office hours even if you’re not particularly struggling with the class, or if you know people who might be able to help you that you could ask more comfortably or more conveniently; professors who know you may eventually be able to give you career advice, research opportunities, or letters of recommendation. (Of course, you shouldn’t try to befriend professors purely for these selfish motives; they’re also good to know just as fellow humans.)

But there are also plenty of legitimate reasons to skip class. (It’s really unclear how many people out there need to hear this, but my past self did, and I want this draft out of the way-too-long queue of posts.) Here are some.

We interrupt the irregularly scheduled philosophical posts for some programming memes:

Over the last few days, the Internet has divided itself over what the value of the expression 8÷2(2+2) should be. Some say it should be evaluated as (8÷2)×(2+2) = 16. Some say it should be evaluated as 8÷(2×(2+2)) = 1.

At the risk of belaboring the obvious, the core dispute here is not really mathematical. There is not some sequence of mathematical operations that produces some number, where mathematicians disagree about what number it produces. Instead, this is a dispute about mathematical notation: what sequence of mathematical operations the expression corresponds to the way it’s written. Specifically, it is a dispute about whether multiplication written as juxtaposition (how “2” is written right next to “(2+2)”) has strictly higher precedence than division. It is closer to a linguistic or typographical dispute than a purely mathematical one, and the correct answer to the dispute is that whoever wrote the expression that way should learn to write math better.

This debate is not even new. The internet had fun arguing over 48÷2(9+3) and 6÷2(1+2), which are functionally identical ambiguous expressions, eight years ago. I don’t know why the debate is resurging now and why we still haven’t gotten tired of it.

But life is short, so since we’re here anyway, let’s make some additional memes.

Asking the computer

Some of my coworkers had the idea to ask some programming languages what the answer was. The results were underwhelming.

$ python3
Python 3.6.7 (default, Oct 22 2018, 11:32:17)
[GCC 8.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> 8/2(2+2)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: 'int' object is not callable

I hated timed essays in high school. It’s pretty clear if you skim that part of the blog archives:

My [SAT] essay got a 10 out of 12. It’s an essay I’d be ashamed of posting anywhere else; it’s disgustingly traditional and formulaic. […] This was simply because I knew that using my normal essay-writing mindset, I’d get maybe a 3, because I’d spend the first twenty minutes debating myself over which side I was on and rewrite the introduction ten times. Too bad. I wasn’t there to write a good essay; I was there to get a good score on the SAT.

The worst issue is that students do not need to give in-depth explanation of anything they learned. Due to its stringent time limit, the essay portion rewards quick reckless writing much more than deep thought. […] encouraging students to practice writing 25-minute essays in order to improve their college-bound skills is like encouraging people to play Grand Theft Auto in order to improve their driving skills.

I took the Grand Theft Auto analogy pretty far. I’m not proud enough of these posts to link them, but you can find them if you try. In short, past me thought timed essays rewarded writing too quickly, with a disregard not only for facts but for the opportunity to lay out your opinions and thoughts so you could clarify and revise them, which was actually the most valuable part of writing.¹ I conceded that the time limits made sense as a practical concession to allow you to test students’ writing skills fairly,² but I felt like there was still way too much emphasis on the raw speed.

This post is about things. You know, physical objects. Earthly possessions. The indispensable yet fickle chains that bind us to this plane of existence.

As a freshman I loved getting free stuff. I went into my first career fair starry-eyed, taking free pens and t-shirts and sunglasses and drawstring bags from every company that would let me; I’m sure many other MIT students have gone through the same ritual of passing. There seemed to be no downside. Besides, all the swag was probably small change for most of the companies anyway, due to VC funding or massive government contracts or whatever. So I might as well make full use of it.

Less normally (although I have no idea by how much), I also hoarded a lot of things that naturally pop up in everyday life, the stuff that would otherwise get thrown away: extra napkins, plastic bags, produce containers, boxes, those payment envelopes that come with magazine subscriptions and exhort you to Subscribe Now To 12 Issues For 20% Off!! At some point I kept the fortune from every fortune cookie I had ever eaten in a tiny resealable bag. The rationale was basically the same: they might be useful some day¹ and there was no downside.

I wanted to write a post after two years at college about everything I had learned. I didn’t, firstly because I didn’t make it a priority, and secondly because trying to write about everything I’ve learned at MIT over any nontrivial length of time is the kind of poorly scoped endeavor that I could never complete to my own satisfaction.

Two years came and went, and now it’s been two more years and I’ve learned even more things, not to mention, actually graduated. Jeez. I tried to self-impose a deadline for the big post, but it didn’t work out. There were still too many higher priorities, most of which were also natural consequences of graduating. I also couldn’t bring myself to cut anything, because unlike most of the stuff I haphazardly throw onto this blog, I can actually imagine an audience for just about everything I wanted to write about.

Finally I decided that I would break it into lots of small posts on specific topics. This way, at least the perfectionism can’t bleed between posts too much. The first topic I wanted to write about is simple, mundane, and also fairly limited in scope itself: how to choose your MIT username.

http://111.186.63.17/perf.data.gz

Environment: Ubuntu 16.04+latex

In this challenge, we get a gzipped file called perf.data and a minimal description of an environment. Googling this reveals that perf.data is a record format of the perf tool, a Linux profiler. Installing perf allows us to read perf.data and see some pretty interactive tables of statistics in our terminal describing the profiling results, from which we can see some libraries and addresses being called, but they don’t reveal much about what’s going on. One hacky way to see more of the underlying data in a more human-readable way (and to see just how much of it there is) is perf report -D, which dumps the raw data in an ASCII format, but this is still not that useful. (One might hope that one could simply grep for the flag in this big text dump, but it’s nowhere to be seen.) Still, from this file, we can definitely read off all the exact library versions that the perf record was run against.

I put this question in my FAQ, because at least two people have asked me this question, and that’s how frequent a question needs to be to be on my FAQ: I got an IMO¹ gold medal in 2012, as a ninth grader, and an IOI gold medal in 2014, as an eleventh grader. I could have kept going to either, or even decided to try taking the IPhO or something, but I didn’t. Why not?

The short answer: It was a rough utilitarian calculation. By continuing, I would probably displace somebody else who would gain more from being on an IMO/IOI team than I would. Besides, I wanted to do other things in high school, so I wasn’t losing much.

I think the short answer actually captures most of my thinking when I made the decision back then, and it’s not really new; I said as much at the end of 2013. But behind it was a lot of complex thoughts and feelings that I’ve been ruminating over and trying to put into words for the better part of a decade. Hence, this post.

There is a natural question that precedes the frequently asked one that I have never been asked, something I am now realizing I never honestly asked myself and never tried to answer deeply: Why did I participate in the IMO and the IOI in the first place?

(this post is sketchily backdated from February 2020)

Fourth year with ✈✈✈ Galactic Trendsetters ✈✈✈. Past: 2018, 2017, 2016, writing with Random in 2015). Evidently I was unable to find the time or motivation to blog about this, and I can’t remember any particularly good stories, so let me just slip this backdated list post in the archives.

The theme in a few sentences: The hunt organizers declare a new Molasses Awareness Day, but because they didn’t follow the proper procedure, throwing the world of Holidays into chaos, where “chaos” means molasses. This resulted in lots of puzzle-to-meta matching, since puzzles were in Holiday towns but metas were at lampposts between two towns.

Highlights of puzzles I solved:

• A Good Walk Spoiled (Holi): Happens to be about a topic I know well and find interesting. Fun steps. Justified use of substitution ciphers; the pasting into quipqiup was worth it.
• State Machine (President’s): A one-aha! one-step puzzle. Maybe more unfortunate because apparently this was not the only puzzle to use a specific aspect of a specific source material, but I didn’t know about that other puzzle so it’s fine.
• Clued Connections (Pi): I went to sleep near the end of this puzzle, so I didn’t see it through, but Floorpi was collectively called out by it.
• I Can’t Deal with These Endless Numbers (Patriot’s): Incredible.
• Loaded (Arbor): Simple and self-contained enough.