Idea Garden
Thu, Jul 24
A short rant about writing unclearly/using roundabout notation, arguments, or just not optimizing for clarity.
The horrible things that it does (that often go unmentioned):
1. Entrench negative precedents for the rest of time (think bad unit notation, bad names)
2. Reduce the number of people that can understand your writing (duh)
3. Prevent other people from building/criticizing your ideas, since they are afraid to admit that they don't understand them
4. Make your area of thought an "ivory tower" for pretentious people that want to sound like they know something you don't
5. Turn valuable people off from your discipline because it is boring, difficult to understand, and mentally taxing (in a bad way)
6. Make future debates confused, stunt growth, make the simplest results complex, and result in diverging opinions rather than converging ones; language game.
All because you were too lazy to simplify, simplify, simplify. All non-verifiable academic journals (which then judge on the "seeming significance" of your writing) are like poison; they rot everything they touch, since they do not optimize for clarity. It is not only a burden on yourself, but a burden on anyone that interacts with your ideas going forth.
Sun, Jul 6
Anything physically possible (or even, perhaps, logically possible - that is, embodying no logical contradictions) is worth thinking about for the purposes of a thought experiment. Why, if it will never happen? Because good knowledge is not about things in particular, it is about things in general. That means that any rules that we apply to our everyday lives must also be tested on every possibility, even if they are "out of distribition." That is why thought experiments are so useful.
It follows that "this is a meaningless hypothetical since it will never happen" is not an adequate response to a hypothetical. However, "despite being interesting, this hypothetical should not have any probabilistic weight attached to it for the purposes of some goal", is.
Fri, Jun 27
An idea about analogy-making:
We interact with very curious ideas in pure mathematics and in theoretical computer science. They are abstract, yet seem very pure. They are orders of magnitude simpler, in the information-theoretic sense of the word, than people, trees, friends, emotions, etc.
Why is it, then, that understanding math can be difficult, whereas understanding things like someone else's emotions is so easy that a two-year-old can do it? It's interesting, isn't it, that math is often considered hard when the hardest mathematical concepts can be encoded in a billion times less space than the messy, imperfect domain of, say, the intentions of others.
We can understand people almost effortlessly, among other complex concepts like food, promises, weather, and payment, because these ideas (or analogous ones) had evolutionary pressures imposed on them. As such, they are relatively "low effort" encodings in the mind. Whereas other ideas - exponential functions, complex numbers, entropy, etc - had little to no significance whatsoever, despite being information-theoretically simpler. We naturally do our "computation" - our thinking - in the lowest effort language possible. That means that we must use analogies, often imperfect ones, to even have a chance at understanding more abstract concepts. They help us translate to a lower-effort encoding.
One example is Dennett's so-called "intentional stance" - that is, the understanding of various processes as having wants and desires. The ball "wants to" roll down the hill. Hydrogen "wants" an extra electron. Balls and hydrogen items don't, of course, "actually want" anything, but we naturally think of them as such, since it lowers the mental effort of reasoning about them in our brains.
It follows, then, that the ideas we reason about (at least, the instantiations of them in our minds), are not actually as pure as we think. They have taken on forms that we can parse; forms that are often imperfect. I think it's an interesting philosophical question whether or not this is the only way we can reason. Quantum physics is hard not because it is complex (its laws can be written in just a handful of euqations), but because it is in a realm where it is difficult to find useful analogies.
Melanie Mitchell notes in Complexity that Newton was criticized for not giving a "mechanism" for gravitational force - that is, it didn't explain gravity in terms of anything we would find simpler (i.e. objects, particles, invisible strings, etc etc). This is what we look for when we ask "why" - an observation explained in terms of simpler constituents. But these kinds of mechanistic models can only be mechanistic to an extent; eventually, there may not exist an accurate low-effort encoding in our minds that can be reached through analogy. In this case, despite things still being information-theoretically simple, we have to go in blind - physics does not bend to the whims of neurobiology. At this point, Feynmann might just tell us to sit down, shut up, and calculate.
(turns out he did not actually say this... though I think he probably would have)
Tues, Jun 24
We are curious in two ways. The first, to gather data - to learn what is where and what is what. The second, to develop explanations - to understand why things are the way they are. I am biased towards the second one. That is where beauty lies.
Elegant explanations are one of the most beautiful things in the world. Why? Because they lessen the cognitive burden upon the mind. They are, at a physical level, a way to be more energy-efficient.
Sun, Jun 21
Does being secure in faithful love require irrationality?
These two statements are highly unlikely to be true together:
To be faithful, you must love them for some indescribable quality or combination of qualities that is, in a way, arbitrary; that is, you love them not for some combination of outward qualities (there will always be someone with better qualities), but for something about them specifically - something you cannot rationalize to others, except in the context of their relationship with you.
It follows, perhaps, that to love, one can be fair, or faithful, but not both. To answer the question - it can be rational in the sense that you love them based on something to do with you - namely, that they love you. But unconditional, faithful love - love independent of self, of relationship with you in particular - is irrational. And it is supposed to be.
The brain has two evolutionary imperatives; one is to be as efficient as possible (i.e. not use too much energy) and the other is to be as useful as possible (saving the body from danger, doing important planning, etc etc.)
If you draw these things on two axes, you get what I like to call the "quadrant theory of doing stuff."
On the horizontal axis we have "exertion": not exertion in the sense of challenging your ability, but rather the sheer mental energy expended to do something. On the vertical axis we have "perceived utility." I say perceived because it is a proxy: a shortcut-of-sorts that the brain uses to gauge true evolutionary utility - that is, the actions that will maximize your chances to pass on your genes (through survival, reproduction, etc).
| High Exertion | Low Exertion | |
|---|---|---|
| High perceived utility | Flow | Entertainment |
| Low perceived utility | Drudgery | Boredom/Rest |
High perceived utility, high exertion:
- doing a fulfilling job or hobby
- learning out of curiosity
- going on a "hero's quest"
High perceived utility, low exertion:
- scrolling social media,
- watching a TV show, gossiping,
- doing something fulfilling but not difficult.
Low perceived utility, high exertion:
- homework, monotonous jobs, what we call "busy work"
Low perceived utility, low exertion:
- meditation, doing nothing, resting
In these four quadrants, our minds constantly experience an attraction towards the top-right corner. We detest the bottom left; we would rather be anywhere else. As such, we are pushed to the bottom right or top left. I'd of course rather do a fulfilling job than a monotonous one, but I'd also rather do nothing at all.
We all acknowledge that copious consumption of social media feeds does not typically have a substantial positive impact on our lives. Why do we still do it, and why does it have a high perceived utility? Because the shortcuts in our brains are imperfect out-of-distribution; that is, very different from the savannah where our genes were curated. This is why some behavior that would have resulted in increased survival chances a dozen millennia ago (gorging on a high-fat, high-sugar diet in order to avoid starvation, for example) no longer serve us well. So it is with perceived utility.
The things that entertain us are often pleasurable because they imitate something that gives us a higher chance of survival. Often, this is information about our social environment. Our brains are wired at a very fundamental level to enjoy drama TV shows, funny videos of people embarrassing themselves, or reading about the latest trends done by people you have never met and never will meet. This is because our brain takes a shortcut; instead of meticulously calculating true utility, the instinctual mind often relies on its hard-coded evolutionary heuristics. Of course watching an important stranger embarrass themselves on the internet is not useful to your life, but in the jungle it likely would have been (likely due to the smaller group structure - someone you see once you are likely to see again). But sometimes we take the longer path; looking in the future to manually increase perceived utility of things that have true utility. We can force ourselves to do things that are not immediately fun because we know they will pay off in the future. This is what we call self-control.
To get yourself to do something you don't want to do, make it feel either important (increasing perceived utility by manually adjusting it to reflect true utility), fun (increasing instinctual perceived utility), or easy (decreasing exertion). If you can't do one, it will never be done.
A rule for life: despite your mind instinctually pushing you to the top right, try to spend as much of your time as possible either in the bottom right or the top left. Do fulfilling, difficult things or nothing at all. Rarely are things both low in exertion and high in utility. Our evolutionary heuristics are quite gullible.