Time feels both short and long, in the moment and in hindsight. It always seems short on one scale, and long in the other. Either the days fly by, or the months do.

Well-spent days are ephemeral, but stretch on endlessly in the rear-view mirror; wasted hours seem to drag on forever, but hardly make any memories.

This is because we don’t understand the length of a span of time directly; we think of times as "short" or "long" based on what we perceive as _happening in_ that time. The more events our brain knows about from a period of time, the longer it feels like it was.

![](https://brennancolbergdotcom.s3.us-west-1.amazonaws.com/meta-images/writing/perceived-duration.jpeg)

When dopamine-pumping headlines bombard you online, the onslaught seems endless: your brain is busy reacting to tons of novel information, so time creeps by! But you won’t remember most of what you see, because it's ephemeral garbage. So, when you look back later, there’s only the blank space of an empty afternoon.

In contrast, when you focus deeply on one thing, it’s all your brain grasps, and the time seems to vanish in a **poof!** But putting so much thought into something builds a strong memory, which might last for years. Months filled with productive days feel longer in your rear-view mirror than the ones you frittered away, because their memories are still around—while wasted time leaves no lasting impression.

It’s a striking paradox: the shorter your days seem, the longer they’ll last. What's the best way to have memories of years chock-full of rich, deep recollections? Do whatever makes the hours fly by.

the paradox of perceived duration

*or,[^1] "Why people who think that long words are inherently bad and should never be used—because they might not make sense to other people—are wrong."*

A common piece of writing advice is to "avoid using long words." According to this line of thought, good communication expresses thoughts completely in simple words, instead of showing off the author's vocabulary—and those who use big words flatter their egos at the expense of getting their point across.

When obscure words are used, fewer people understand them. But this can be for one of two reasons: because they're being used by the author to sound more intelligent (which is bad), or because they serve as efficient references to obscure abstractions (which can be good). Some things *are* complicated, and require precision: in such cases, sesquipedalian communication is indeed effective.

---

Look at any academic discussion of science, or at the debates of engineering professionals: no layman has any hope of comprehending the strange words flying around. But this makes sense, because jargon serves to *compress* ideas which would otherwise take a long time to explain. A programmer can spend paragraphs describing how fast their sorting program runs, or can say "in `O(n log n)` time" and be done with it. Nobody without any computer science knowledge can tell what the latter means, but that's usually okay—whoever they're talking to should know, or can ask them what they mean.

But often big words are used to puff simple ideas up, so that they seem more impressive than they are. This is how most people—including me—learned to write: they're given a shallow topic to write a long essay on, and after writing down everything they know, they pull out the thesaurus to meet the requirements by restating their ideas in all sorts of different ways. It's most common in school, but these bad habits trickle into the real world too—and are what the common advice tries to warn against.

Good usage of weird words *compresses* ideas; bad usage *obscures* them.

---

Complex words in a proper setting correctly signal to observers that they are not knowledgeable enough to follow along—and this is fine, because it's true. The reason such language is used is *to compress* communication with those who *do* understand, though; confusing normies is a side effect!

But out of that comes danger. Just as vulnerable prey mimic the decorations of their more dangerous cousins to avoid getting eaten, people who want their ideas to be accepted without argument will also use complicated words in the hope that you, uncomprehending, surrender to them (just as you would yield to someone who's above your level). This strategy, an insidious version of the appeal to authority, is called "Eulering"[^2].

How can you tell one type of obscure communication from another? By trying to understand, and asking clarifying questions. Anyone who understands the big words they use—and means what they say with them—will be able to uncompress their speech: to give you a longer and more familiarly-worded explanation. When a charlatan or poser is using big words, in contrast, asking questions will force them to either (1) backtrack on what they said, (2) give circular definitions, or (3) contemptuously deny your request, implying or saying that you're simply *too far* below their level to understand.

Unfortunately, it can be quite hard to judge someone if you cannot interact with them. If someone leaves an incomprehensible Reddit comment, how can you tell if (1) you're uneducated in the area, (2) they're crazy, (3) they're deliberately messing with you, or (4) they're correct, but just a bad writer? Sometimes you can't; you just have to rely on your gut, or be content with uncertainty and move on.

---

Long words are not inherently bad; in proper usage, they serve to compress, rather than to obscure. But since compressed language is often hard for the layperson to understand, those who seek to dazzle you into submission will sprinkle their language with long words to obscure their arguments, hoping to be accepted on faith and authority rather than from a place of mutual understanding.


[^1]: The intentionally incomprehensible title is [a joke](https://www.merriam-webster.com/dictionary/sesquipedalian).
[^2]: ["Getting Eulered,"](https://slatestarcodex.com/2014/08/10/getting-eulered/) Scott Alexander (2014).

against antisesquipedalianism

A young Elon Musk, deciding what to do with his life, famously reflected on which upcoming technologies would most impact the future of humanity[^1]. His list? "The internet; sustainable energy; space exploration, in particular the permanent extension of life beyond Earth; artificial intelligence; and reprogramming the human genetic code." [Do](https://en.wikipedia.org/wiki/Zip2) [any](https://www.paypal.com/us/home) [of](https://www.spacex.com/mission/) [those](https://www.tesla.com/about) [sound](https://openai.com/about/) [familiar?](https://neuralink.com/about/)

Ever since reading about this exercise, I've tried to do it for myself, seeing the value in such clarity of vision. My early lists were topped by education: I wanted to start undermining Prussian schools by providing a better solution, so that my future kids will have a better childhood than me. I've been [working](https://celium.colberg.dev) [on](https://joinweekdays.com) [this](https://synthesis.is) for four years now, and feel that my goal is on track. So the list has regained relevance: how do I want to use my life? What will I be working on in a decade?

Here are my current interests. I'd feel blessed to have a big impact in any of them. 

## Nanotechnology

The vision of atomically-precise manufacturing explored by Feynman[^2], Drexler[^3], Stephenson[^4], and Storrs Hall[^5] fascinates me. Diamond skyscrapers! Sapphire rocket engines! Custom fabrication of anything we can imagine, at an atomic scale and out of vats of lone elements! Moore's Law, but for atoms! The idea of descending the Feynman waldo path[^2][^5] and having molecular-scale self-replicating nanofactories that can then do practically anything... it titillates me to no end. As far as I can tell, the challenge remains untackled; how cool would it be to be the one to do this? With enough funds and dedication, I could probably make it a reality.

## Abundant [Atomic?] Energy

Fission, fusion, everything: bring on the nukes! Any really "cool" technology, to be abundantly useful, will need energy to be a few orders of magnitude cheaper. So we as a species _really_ need to get back on the Henry Adams curve of exponentially-increasing energy production, and nuclear sources seem to be the obvious path forward here[^5][^6]. (If you think about it, isn't the fact that the energy we need to light and heat our homes _doesn't_ have a negligible cost kind of ridiculous?!) Fortunately and unfortunately, it seems like this problem is being tackled already [by](https://firstlightfusion.com) [a](https://www.cfs.energy) [gigantic](https://www.tokamakenergy.co.uk) [fleet](https://www.helionenergy.com) [of](https://generalfusion.com) [fusion](https://tae.com) [startups](https://phoenixwi.com), at least one of which will probably succeed: bad because it's too late for me to make a big difference, good because that's a key prerequisite to a lot of other cool shit (see above and below) I can work on instead.

## Hovercraft

This one is mostly from Clarke[^7]. He thinks that hovercraft could economically replace trans-oceanic shipping, with the added benefit of no longer caring about the difference between going over land or water (except to avoid cliffs). The reduced friction between vehicle and ground that hovercraft bring would also eliminate a lot of our infrastructure problems: if all you need is a flat surface that you barely impact, who cares about maintaining expensive highways or the like? I'm still unsure about the affordability math (why aren't shipping lines already doing this?), but am designing and building a small hovercraft myself to learn more. This interest is almost purely for the "haha that would be COOL if it works" factor.

## Education

I mentioned it before, but this is really such an all-encompassing category. What should children do? (Run around, hurt themselves, and read.) What should teenagers do? (Follow their curiosity and use their knowledge to make changes in the world.) What should young adults, and older adults, do? (The same thing.) How can libraries and social structures be improved to emphasize eternal exploration and edification? There's so much room for improvement, and it's _such_ an obsessively thorny and fractal problem; despite what I said earlier, I probably could spend my whole life in it.## Competent Governance

I don't really know what this looks like, but I'd love to help fix society's decision-making incentive structures. I doubt this means "running for office"; it's more like experiments with sovereign cities, digital nations, and so on. How do we attract the most competent people into the high-leverage roles of politics? How do we speed up government decision-making, and give bureaucracies a sense of upside perspective? How can we persuade officials to optimize for something other than simply reelection? What would it take to make corruption vanish because it's a bad strategy, not just out of fear?

## Space Colonization

Pretty self-explanatory. God willing, within my lifetime humanity will have transcended the petty bounds of our home planet and begun to reach to the stars. First Mars[^8] and the Moon, then other planets and the void itself[^9]... I would be loath to see all this happen without pitching in at all. Perhaps I'll be a colonist, perhaps an industrialist, perhaps just a propagandist. Whatever form it takes, I'd love to be involved.


[^1]: ["The Cook and the Chef,"](https://waitbutwhy.com/2015/11/the-cook-and-the-chef-musks-secret-sauce.html) Tim Urban (2015).
[^2]: ["There's Plenty of Room at the Bottom,"](https://nanoparticles.org/pdf/Feynman.pdf) Richard P. Feynman (1960).
[^3]: ["Engines of Creation,"](https://amzn.to/3r9cWqD) K. Eric Drexler (1986).
[^4]: ["The Diamond Age,"](https://amzn.to/3v2ov46) Neal Stephenson (2000).
[^5]: ["Where Is My Flying Car?,"](https://amzn.to/3LGGtjk) J. Storrs Hall (2021).
[^6]: ["The Education of Henry Adams,"](https://www.gutenberg.org/files/2044/2044-h/2044-h.htm) Henry Adams (1904); chapter XXXIV.
[^7]: ["Profiles of the Future,"](https://amzn.to/3JheCVh) Arthur C. Clarke (1968).
[^8]: ["The Case for Mars,"](https://amzn.to/3DKKoZO) Robert Zubrin (1996).
[^9]: ["The High Frontier,"](https://amzn.to/3x9rNVS) Gerard K. O'Neill (1976).

my musk list (2022)

Occasionally, an idea pops into my head to do something—but I'm too busy to do it right away. So I tell myself I'll do it later, and move on. But recently I've noticed something funny: these things I say I'll do "later" almost never get done.

I think it's because we all live in the "now." Now is "now"; yesterday was "now"; tomorrow will surely be "now." And "now" is not "later": never do I sit down, say to myself "ah, now it's _later_, gotta to do all those things I put off!" and then do them.

I can follow through on _specific_ things I plan for the future—like sitting down to write and edit this essay—if I say "I'll sit down at 4 PM to write." Why? Because 4 PM will eventually roll around, I'll notice that "it's 4 PM now," and I'll feel cued to follow through. And if I can't find a precise time to do something, it's pretty safe to assign it to a specific day, week, or month—because that broader target will eventually be "now," and I'll deal with it then.

But punting todos off to an undefined "later" is like kicking them into the abyss; if they ever get done, it'll be because they resurfaced somehow and got placed into a more specific queue. If I actually want to do something, but not _now_, I have to pin it to a precise future time.

The opposite is true too, and a fun trick: if I feel an urge to do something that wouldn't actually be a good idea, I can make myself a confident promise that I'll do it "later." "Of course I'll buy a new car. Just... not now. Later." And bingo—I can feel all the anticipation and dopamine and giddiness that comes from a big purchase, while saving myself some hassle and leaving my bank account intact. (I can also use my money multiple ways simultaneously in this manner: I can buy a Tesla _and_ a new computer _and_ some fancy books "later," and still have all the money left over!)

"Later" doesn't exist. Be wary of this, and abuse it where you can.

"later" doesn't exist

College is a game. In playing it, you can optimize for many different outcomes: status, income, free time, pleasure... or learning. The latter is what I aimed at before leaving; here is my advice—gleaned from both success and failure—to anyone seeking a similar course.

## 1. Never attend lectures passively.

Impersonal lectures are categorically inferior to their own recordings. You can watch videos anytime, or multiple times; pause them or back up, in case you missed something; or speed up or slow them down, to match your comprehension rate. So you should never waste your time by letting someone talk at you when you could watch a recording instead—unless you wouldn't actually watch it, and need the accountability.

Aspire to never take classes with a passive lecture at their core. If you can't avoid some, skip the lecture—and learn from the textbook, slides, and recordings instead. If they take attendnace, figure out how to fake it, or take the hit to your grade. But whenever you can't get out of attending, do your utmost to engage: sit in the front row, do live exericses, volunteer answers, and ask about anything you are curious about or don't fully understand. Seek first to not waste your time; but once it's taken from you, get everything you can out of the situation.

Instead of sitting through big lecture series, take small and interactive discussion-based classes to get your credits, and read the books that lectures are based on to get your content. That way, you'll constantly alternate between ingesting new ideas in bulk and digesting them in a give-and-take with other people's models of the world. Take in then process information like you ingest and digest your food; our minds and bodies are best nourished on the same principle.

## 2. Use the library.

Reading is the best way to learn a course's content—but there is a lot more than assigned readings out there to absorb. If you're briefly curious about a topic, or some quote stuck out to you and you'd like more context, or you want to learn something that no small courses are offered on... just find a book in the library! You have access in under 20 minutes to virtually anything in the record of published knowledge. Why _not_ use that as much as humanly possible?!

My university had over 9 million books in its libraries, and no limits on how many I could check out. So I took library use to the extreme, checking out a copy of virtually every book that even vaguely interested me and piling all of them (up to 200 at times) in my dorm's closet. With this "Random Access Library"™, even that 20 minute delay between impulse and investigation was eliminated. (I'm not sure I recommend doing this; it was great for me, but your priorities and mileage may vary.)

But that's not all! Libraries are also fantastic places to work; they're quiet, magnificent, and varied. Get restless in one place? You can take a 5 minute walk—without even stepping outside—and sit down somewhere else to continue. You can pick your ambiance: dead silence deep in the stacks, white noise by a ventilation duct, regular chatter in a café, or the occasional creaking of wooden chairs being pulled around in some reading room.

Working in a university library is a good way to stay humble, too: there's nothing like desperately racking your brain while surrounded by hundreds of thousands of books to remind you of the limits to your comprehension.

## 3. Talk to interesting people.

The people who seem most interesting are either very similar to you, or very different. In either case, you can learn a lot by having deep conversations with them, in a manner that not much else compares to.

Similar people hold up a mirror to you: you can see yourself in how they think and act and interact with the world, which gives you the opportunity to decide who you want yourself to be. My biggest personal changes have always come when I recognize myself in the flawed behavior of other people: "Wow," I'm able to tell myself. "Is _that_ how I seem to others?" And once I've learned that about myself, I can change. Lean into the good parts, try to fix the bad parts, and come out a different—better—person.

You can also hone the specifics of your knowledge in conversation with others who understand the same things. By talking with others of similar but slightly different beliefs, you can challenge and refine the precise details of your ideas—in a way that's impossible in isolation. When you're the only person around who knows what you're talking about, who can tell you when you're wrong?!

Extremely different people are interesting for the opposite reason: they make you question and defend your core assumptions and values, the bedrock on which all your other knowledge is based. If you're wrong there, it's _really_ important to know, and it's hard to know our assumptions or values until we come across someone who rejects them. It can be shocking to hear the earnestly-expressed ideas of extremely different people, so I relish sitting down with them: either I come away with a better understanding of why I believe the things I do, or I'll be corrected in my errors. There's no way to lose.

## 4. _Do_ things with your knowledge.

As you learn, it's important to poke at your knowledge to ensure it matches reality and doesn't have any gaping holes. Do this by _using_ your knowledge: take tests, write explanations, persuade people, run experiments, make projects... whatever you're most inclined to do. By measuring your understanding against reality, you can find and patch its imperfections; the more you do this, the better. No specifics truly matter, just the action.

Schools try to trick you into doing this by giving you homework, making you take tests, and occasionally requiring a guided project. But any assignment you get in class is fundamentally constrained: it has to be evaluated and given to hundreds of people. This means most are rote, formulaic, and general; assignments famously feel like chores. So it's tempting to generalize, as many people do: "these assignments seem pointless, so practice must be pointless." Or worse, "so learning must be pointless." No! It's just that schools are in a tricky structural position; they're doing the best they can, only cannot help but fail.

That's because learning is inherently personal; it's about what _you_ know, a quest to improve _your_ understanding. So you should prod at the edges of your knowledge yourself: write an essay about something you barely get, test your physics skills against practice problems, code a personal website to learn how the Internet works. Practice pushing on your comprehension, and interfacing with reality there to be sure you're on the right track.

And you're in luck: although homework isn't the best, college is a goldmine for these kinds of learning opportunities! Join a club to learn debate; work in a research lab to see the cutting edge of science; start a blog to see how well your ideas stand up to online criticism; spend a few weekends in a makerspace to learn how to use a 3D printer. At large universities, there will be a resource for _anything_ you're interested in—so seek them out. Don't just take classes.

---

That's my best advice for learning at college: avoid passive lectures, read a lot, have good conversations, test your knowledge. Now here's the real secret: this advice is _general!_ You can read and talk and make things _anywhere;_ you don't need to pay $50k/year. And you can do these things your whole life long; learning needs not stop with a diploma, or when you "grow up."

In fact, I would argue that if you solely want to learn, today's colleges are not the best use of your money and time. Why not rent an apartment, get a part-time job, and spend all of your time reading and talking and writing and building anyways for a few years? You'd have much more freedom, you won't be in debt, and four years of diligently making things and showing them to people will certainly make you employable (or give you the income to not _need_ a job). If you like the campus or people, you could even live near a university and hang out there!

College might still be a good choice if you're optimizing for something other than learning. Do you want to go into academia? Do you lack the motivation or self-control to learn on your own? Do you need a visa to not get kicked out of America? Are you trying to build an elite network at a top university? Then college might be for you. But if you just want to learn, avoid it.


how to learn at college

**Trust** (confident dependence upon the fulfillment of promises) and **honor** (the quality of being trustworthy) form the difference between barbarism and civilization. Without them, humans inevitably devolve into backstabbing.

---

If I say that I’m going to do something, it’s important than I then _follow through_ with my promise. **Your level of trust in me is how confident you are that I’m going to deliver when I give you my word.** The more trustworthy I seem, the more likely you are to trust me—to _depend on_ me following through, acting as if it’s all but certain to happen.

So **trust is Bayesian**, built up with experience: the more I have previously done what I said, the more likely you are to trust me with higher stakes. And vice versa! If I’m consistently flaky, you’ll be skeptical of taking risk by believing my word.

If two people have trusted each other for years and over dozens of interactions, without issue, they can depend on each other without hesitation: **proven trust is the _foundation_ of strong relationships.** The same goes for organizations: I’ve received my package every time I ordered something on Amazon, so I don’t worry _at all_ about clicking “order,” as I might for a one-off order from some random other website.

**The value of longstanding trust is made most clear when it’s lacking.** Whenever I ask a stranger to do something for me, I’m wary—who knows if they’ll follow through? What if this new website just takes my money the first time and runs, never sending the thing I bought? When I don’t have a prior relationship with someone, there is uncertainty and _risk_ involved in dealing with them, so it’s extra costly! I have to spend extra time and energy planning for the case where they don’t follow through, double-checking that they’re proceeding on their task, just _worrying_ in the background in general.

**Communities are valuable because they enable trust formation**: their members have a reputation with each other, a shared sense of their trustworthiness. If you screw over a member of your tight-knit church, everyone else will hear about it, so you stand to lose many of your _other_ relationships! Because betrayal has such a high cost within communities, members can trust each other more than strangers, even without a prior relationship. Are you really going to risk losing the trust of our entire community to get one up on me in a random transaction?

Many million- and billion-dollar deals in the tight-knit worlds of finance and business are made with a handshake, and only formalized later; this is why. Sure, if conditions change, someone who regrets their decision _could_ back out: but word of this dishonorable act would spread to everyone they deal with regularly, nobody would trust them anymore, and their career would be _over._ It’s hardly ever worth sacrificing your entire livelihood to make a single quick buck.

**Nevertheless, some people _are_ untrustworthy**, and some people _do_ betray each other at will. But as they do this, they inevitably drive away any trusting people around them, and are left in a cesspool of other reprobates. **Such people cannot trust each other—and they know it.** So their days and efforts are spent plotting to backstab others, and trying to avoid being taken for a ride themselves; this arms race of guilery ensures that anyone who focuses on anything but such cunning is swiftly ruined. With their entire focus on these zero-sum games, there is no room for productivity, for improvement, for development; rather, what dregs of value exist ceaselessly change hands until they are consumed or destroyed.

When promises are kept, minds are freed from endless scheming, and can move on to greater things: to learning, to relaxation, to mastery, to wisdom. **It is trusting minds that have discovered and made abundant the fruits that humanity enjoys.**

trust is the bedrock of civilization


I do tasks on three timescales: they either get done immediately, after a shamefully massive delay, or not at all. There’s little in-between.

The latter cases are some of my biggest sources of shame in life. I pride myself on being as reliable and honest a person as possible, so it eats me up to betray someone’s trust for week after week after month… what does it say about me when I’m being trusted and completely fail to deliver, despite many promises?

Ironically, this instinct and identity are what _cause_ those massive failures: when I’ve been procrastinating on something, I can hardly bear to think about it—and so it’s excruciating to _stop_ procrastinating long enough to deal with it. It’s a tragic positive feedback loop: the more I slack, the harder it is to stop.

I call this phenomenon the “Procrastinatory Doom Loop (PDL)”™:
1. I commit to someone that I’ll do something “soon,” or by a deadline.
2. I don’t promptly follow through; instead, I do other things with my time.
3. As time drags on, the task pops into my head. But I’ve been procrastinating on it, so thinking about it brings to mind my irresponsibility, and makes me feel bad! So, to avoid cognitive dissonance, I push the thought away.
4. The deadline passes; I haven’t followed through. The person that’s waiting texts me, asking what happened. I, awash with shame, reply something vague about how it’s “taking longer than expected,” and promptly shunt thought of the painful topic back to my deepest mental recesses.
5. More time passes; it still isn’t done. When I think about the task, I feel a wave of self-revulsion, and move on quickly. (If completing the task would require focus, I’d need to confront this failure with my full mind for hours at a time: _unbearable!_ So the bigger the task, the more it slips.)
6. Repeat.

PDLs can end in two different ways: on an energetic day, I muster up the courage to face my fears… or I quietly abandon any hope of following through and gradually forget my commitment.

The former is usually helped along by credible threat of severe downside if I keep procrastinating (hi, IRS!). The latter is much more common in personal affairs.

(I’m not very good at avoiding them yet, but at least the problem is clear!)

Procrastinatory Doom Loops

I’m super excited about hovercraft right now. I want to build one.

During my reading session today, I finally cracked open [Clarke’s “Profiles of the Future”](https://amzn.to/3lJvVp2)—which is a _gem!_ It’s what I hoped [“Engines of Creation”](https://amzn.to/3dsfb14) would be (but failed to deliver): a rolling series of extrapolations about future technologies that are physically possible. All following two fantastic chapters laying into people who doubt the potential of future technologies because they haven’t yet been proven out, of course. I’ve only really gotten through the “technology” section, where he makes it clear how not-a-fan of cars he is, proposing instead a combination of local moving sidewalks, mid-range hovercraft, and long-range VTOL or rocket travel. All founded on physical possibilities, with some economic extrapolation (why submarines or hovercraft will dominate long-range shipping, etc).

I’m most excited about the picture he paints of hovercraft (or GEMs, as he calls them: Ground Effect Machines). These craft threaten to obsolete both cars, ocean shipping, and some long-range travel. Hovering 10 feet off the ground—and being more efficient with scale—they can cruise over any flat surface: ocean, roads, plains, you name it. They have almost no limits on speed because of their frictionless air bearing, a pretty safe failure mode (at least at low speeds) by simple deflation/floating, and ultimate path flexibility. Highways would be made obsolete, or at best a convenience; casual drivers would be able to go basically anywhere they can see; ports would be made obsolete because cross-ocean freighters could just continue onto land; roadway construction and maintenance would be made much easier from the significantly lower point pressures (a “car” would have its mass evenly spread over its area, instead of it being concentrated on less than a square foot of the roadway); and so on.

Their commercial viability presents a technical and economic problem: how much does energy cost, and how much of it can you store on the vehicle? I think that the energy costs are the biggest limiting factor, from his account; he mentioned that gasoline hovercraft are already in wide use in various applications (military, remote terrain access/maintenance, etc). He essentially projects that “modern” hovercraft (1960s) won’t ever become too prominent, but with the advent of dense batteries and effective electric motors, they’ll start to dominate. I wonder what the limiting parameters are, and how close we are today—with battery and electric car technology—to meeting them? Certainly the dynamic would change soon with power generation like [Helion](https://www.helionenergy.com)’s… a shipping-container-sized 20 MW fusion plant could probably serve as a decent engine for a massive freighter, right? Would it be affordable? I should do some research and calculations.

This is something right up my alley: way back in the day, I even made a primitive “hovercraft” in my parents’ garage! It certainly wasn’t very effective, with a ground clearance of zero (and a constant tether to corded power). But I’m at least vaguely familiar with the technology, and know the thrill of balancing on a pocket of air. Given that there’s at least a small chance that consumer hovercraft would be viable with today’s technology, I should look at it and see if I can build one. (Having a solid income stream helps to overcome any financial burden to experimentation, if that remains the limiting factor!) Even something with short range to drive (float?) around the roads of Corvallis could be fun.

I mean… what’s the specification for that? I need to be able to go ~15 m/s, to “brake” by having powerful acceleration/deceleration fans (a sudden stop in 3 seconds = 5 m/ss), to hold myself and some groceries or Home Depot parts, to not get annihilated should some car hit me? To tolerate 6 inch bumps in the road? To have at least ~3 miles of range (safety margin)? To not be obnoxiously loud? It sounds doable, I think?

One major difference from a car is that this vehicle will very much travel in a non-steering manner: without wheels to steer the momentum, slowing will require precise acceleration opposite the velocity vector. It won’t be too hard for people with video game familiarity (especially with spacecraft), but the idea of “turning” doesn’t really exist… which will be interesting. I’ll have to develop some kind of spacecraft-like thrust control software, which can move me in two axes and rotate me in the third. And a few pre-specified subroutines, key of which will be “STOP” (to accelerate opposite velocity, as mentioned before). That’ll require keeping track of position and velocity, too… lots of fun sensor/measurement problems.

Another difference is that there won’t be many physical issues to indefinite acceleration: the only real variable there is air resistance (|v|^3), without any sort of friction or heating or internal shaft pressures to worry about. I imagine that—at least on surface streets—speed will functionally be limited by acceleration potential, for emergency stopping capacity. In other words, safety will dominate physical constraints.

Which is interesting… because in theory, given that most of their energy is spent keeping themselves above the ground, hovercraft should want to go as fast as possible to be more efficient. Their maximally efficient speed will be at marginal balance of wind and time: when does it cost more to maintain a speed (fighting against air resistance) than it does to keep hovering more during the extra time it takes to _not_ travel at that higher speed? There’s an elegant formula here, and the balance depends almost exclusively on (1) how aerodynamic the craft is, (2) how efficiently it can push itself laterally, and (3) how efficiently it maintains its air cushion underneath. I don’t know the precise numbers, but I would expect the balancing speed to be well above that of cars, especially for bigger vehicles. This is because air resistance scales with the cross-section, so transports can be made longer [in the dimension parallel to velocity] essentially for free; at least relative to lateral resistance. Longer vehicles will still have to hold the extra length up; I imagine effort-to-hover scales with weight and (area or perimeter)? Cross-ocean transports may well end up looking like long, thin missiles, or bullet trains, if they optimize for speed. Although… if hover-effort scales with perimeter, the extra perimeter-for-area of high-aspect-ratio shapes might cancel right back out? Lots of experimentation and math to be done here.

My brain is racing over physics and engineering problems in a way it hasn’t done properly since high school robotics. I _have_ to do this: to design, build, and program a local-range hovercraft in my garage. It seems so fun, and will flex my problem-solving muscles in a non-abstract domain for once (I don’t want that capacity to totally calcify)! And who knows—if the technology proves out at a non-obscene cost, there might be some commercialization potential. Even if there isn’t, it’ll be a great hobby for a few months 😊

I'm excited about hovercraft (1/?)

Much of the metric-checking that I have the impulse to do this morning—and previously would have just gone and done—doesn’t really make any difference. So instead of doing it, I’ve been sitting around, thinking about it.

Checking my finances; why do that? They’ll be there just the same, and everything is even on autopay, so I could (and probably should) go months without looking. Checking my investments; why do that? Similarly, I’ll gain or lose money just the same, because my habit is to leave them alone. Checking my Twitter; why do that? I know that two more quotes have been posted since I last checked (the bot I set up yesterday is well-tested), that there are enough left in the queue for at least two or three days, and that they’ve probably gotten some likes. Checking my website analytics; why do that? All I would get is dopamine, because all the statistics are accessible later; people are seeing or not seeing my writing just the same, regardless of my awareness of their actions.

This habit of mine—to _check metrics_ over and over—is just wasteful. **It uses up my time and my limited mental energy, giving me only hits of dopamine, and affecting no outcomes whatsoever.** The proper approach is to produce and blindly publish, checking back only when I actually want to get meaningful feedback, so that I can adapt my behavior. Meanwhile, I should use my hours and brain cycles meaningfully: thinking about and understanding the world, touching it with high amounts of leverage in premeditated ways, enjoying life, and working on more things (writing or otherwise) that push the boundaries of my capabilities.

I think of Jordan Mechner’s experience publishing video games in the 1980s, as accounted in [“The Making of Prince of Persia.”](https://amzn.to/30teGkz) In a pre-Internet age, he made one game at a time—then only tracked its performance by occasionally seeing charts in newspapers, and by the royalty check he received every month! How great a contrast to today: if I were to make a game with my current habits, I’d be checking its metrics every three hours. But refreshing wouldn’t make it succeed more.

I also think of [“The Goal,”](https://amzn.to/3HgP0YV) which argues against automation for its own sake. If the mechanization of a process does not reduce the amount of resources spent on it, why should it be done? If I check my finances every day, why even bother having them on autopay? The benefit of autopay is that it allows me to _never check_ my finances, instead only analyzing them every few months. This frees up my mind; I never have to _think_ about money and how it’s flowing. But automating it all only to impulsively refresh Wealthfront… well, that doesn’t really have a point, does it?

A corollary: **automation is useful to exactly the extent that it eliminates problems, making you think less about something.** If a program doesn’t free up time, money, or headspace on net—only sucking up resources—it should be abandoned. (Goodbye, “smart”homes!) And the elegance of its implementation is irrelevant: the only things that matter are
1. how quick it is to get working,
2. how effectively it will continue working into the future, and crucially
3. how effectively it solves a meaningful problem.



checking metrics doesn't change anything

**[10/12/21] Clarification: when I talk about "tools for thought" below, I am _solely_ referring to the note-taking apps that rally under that banner, not [the more general theme](https://numinous.productions/ttft/).**

I was thinking about “tools for thought,” the rallying cry of the modern note-taking app proliferation. Are they _really_ helpful in their goal? It seems that Roam and its ilk basically just add tagging features in a giant nested list… but other than that, are pretty bog-standard ways to type on a keyboard.

So are they better than any of the other ways to get information out? I can type on a keyboard here in Ulysses, or in TextEdit, or in Apple Notes, or in Pages, or in Google Docs, and wherever I type they’re still the same words and they still convey the benefit of getting my thoughts out of my head. So would a pen and paper, a pencil and paper, a wax tablet and stylus, or even a stick and some sand (just ask Archimedes and Tesla).

It seems the only feasible benefit this new generation of note-taking apps can bring to the table is increased _searchability,_ memex-style. (Note that this could also be accomplished on a set of text files… so they’re really just fancy editors on top of search, at the end of the day.) But I write a lot, and I think a lot, and the _act_ of writing generally matters much more to me than any recall of it afterwards. It’s the getting-it-out-of-my-head that counts.

Sure, there are benefits to searchability: track your thoughts over time! Find an old note that comes randomly to mind super easily! Catalog and publish your ideas! But I don’t find any of those points compelling _as a tool for thought._ The best tool for thought is a habit of thinking, and of writing: you need to _do it_, there’s no cheat or tool that will synthesize and analyze your experience for you. Reading truths that the machine figures out for you wouldn’t be too much different from reading a book, right? It still needs to sink in; you still need to think and grapple with and understand whatever the idea is.

That’s why I’m skeptical of “tools for thought.” Thinking is simple to do, but hard—and we’ve done it for millennia. How could some modern crutch change the equation all that drastically, summoning thinking-points out of nowhere? What really matters is putting in the reps, not recording your workouts better.

It’s possible that a “tool for thought” that focused on _getting people to think more_ (and more deeply), perhaps by consciously causing its users to develop an addiction, might actually be useful. (Note productive \> consumptive focus.)

the paradox of perceived duration	May 4, 2022
against antisesquipedalianism	April 13, 2022
my musk list (2022)	April 6, 2022
"later" doesn't exist	March 30, 2022
how to learn at college	March 16, 2022
trust is the bedrock of civilization	January 10, 2022
Procrastinatory Doom Loops	January 4, 2022
I'm excited about hovercraft (1/?)	December 7, 2021
checking metrics doesn't change anything	November 13, 2021
tools for thought	November 11, 2021

Writing