5 ms·

There's too much information in the modern world to function without some degree of compression. Compare the journey of Newton inventing calculus to that of th

There's too much information in the modern world to function without some degree of compression. Compare the journey of Newton inventing calculus to that of the modern day student learning it in a semester. The student is presumably experiencing "compression", but the alternative is they spend their whole life merely returning the point that humanity has already achieved. If someone tells me they did a scientific study that proves that changing the color of my phone background will save some percentage of my battery life per day, should I spend the time to reproduce this experience without compression, or should I take this knowledge and use it to improve my life as I move forward? Sometimes it is of course dangerous to act on the "compressed data", you need to unpack it to either understand it or evaluate its truthfulness, but just knowing that doesn't magically grant you the time to uncompress all the data you receive, you plainly couldn't function in modern society that way. Instead we develop heuristics about which information we accept at the summary level and which we need to delve into more deeply. The alternative is never accomplishing anything because you're too busy re-doing all of human history.

nradov 1y ago
Up until maybe the 17th century it was at least possible for a single brilliant, educated person to know the entirety of knowledge in Western civilization. We're way past that now.
- Night_Thastus 1y ago
  I think it was possible for someone from then to think they knew all the knowledge in Western civilization. Even at that point though, it was far too much to keep all in one person's head. They'd only have a surface level familiarity with most topics at best. I mean even one topic like architecture would've been impossible. You'd have to go back a lot further IMO to get it to fit. Somewhere around the development of agriculture, maybe.
- barrkel 1y ago
  There's a lot of tacit knowledge in trades that was not documented, and in many ways can't be easily documented. You can't really document how to swim or play an instrument; you need to practice and train your body. Would you brilliant person from the 17th century be competent as a member of a tribe in South America or as part of a steppe clan in Asia?
  - nradov 1y ago
    Sure, I should have been more specific and limited the statement to academic or intellectual knowledge. Stuff like history, natural history, mathematics, astronomy, philosophy, major works of literature, etc. The point is that up until that time, a wealthy person who devoted themself to study could conceivably learn it all in a lifetime.
    - wredcoll 1y ago
      I've heard similar opinions before and considered them plausible in the past, mostly based on this idea that there were people who "invented" in multiple disciplines in a way that seems impossible today. However, now I wonder if that might be more an effect of the increasing level of communication among humans in general at the time, printing presses, steamships, postal organizations, newsletters, etc, all combined to cause ideas to spread much more widely and more quickly than they did in the past. Perhaps its more a function of being the first (or most popular) person to synthesize certain ideas based on being at the intersection of spreading knowledge.
dkural 1y ago
I think there is a spectrum of compression here, so I find myself agreeing with your overall point but with a caveat. Putting aside "experiences" - travel or emotion - for transmitting information, the level of "desired compression" is based on the background knowledge the receiver brings, and what they consider to be the irrelevant bits. I will say on historical mathematics - reading the greats shows you the road traveled, and often one realizes the unexplored paths along the way, that lead to brand new continents. This may not be as relevant if one is not trying to innovate. I would also say reading Newton as a "first book" is positively counter-productive to learning Calculus in high school.
- wredcoll 1y ago
  > I will say on historical mathematics - reading the greats shows you the road traveled, and often one realizes the unexplored paths along the way, that lead to brand new continents. This may not be as relevant if one is not trying to innovate. I would consider this a function of specialization, which is of course still good, perhaps even required in this modern world, but we can't specialize in everything.
  - dkural 1y ago
    Agreed! I often wonder how LLMs will deal with this eventually. It could be their biggest strength. Some people will want tweet-based responses, some will want book-length. It could generate the right level of content for an audience-of-one, knowing what they know, etc.
lapcat 1y ago
> Compare the journey of Newton inventing calculus to that of the modern day student learning it in a semester. The student is presumably experiencing "compression", but the alternative is they spend their whole life merely returning the point that humanity has already achieved. I wouldn't call learning calculus the compression of inventing calculus, any more than reading a novel is the compression of writing the novel. There's no continuing value in reinventing calculus or rewriting the same novel, but there is continuing value in learning calculus and reading the novel. Compression culture is refusing to do any work, demanding instead a mere summary of calculus or a novel. If you go to class for hours every week, read the calculus textbook all the way through, and do the homework, I wouldn't call that "compression" in the terminology of the article author. On the other hand, if you merely have a summary of calculus, you've actually learned nothing. You can maybe fake your way through a conversation about calc, sound intelligent without being intelligent, at least until someone forces you to do a calc problem and reveals your sham.
- j_w 1y ago
  Conflating the process of discovering something new and learning it after the fact feels like saying "why do I need to know arithmetic when I have a calculator in my pocket?" Anyone giving real critical thought to either statement knows they are not intending to achieve the same goal. They shouldn't be compared when talking about "compression."
- wredcoll 1y ago
  > Compression culture is refusing to do any work, demanding instead a mere summary of calculus or a novel This sounds rather tautological. My intent was to give an example of compression that is an obvious good. If we re-define compression to exclude that, pretty quickly we've just reduced the definition to compression = bad things. Bad things are, you know, bad, by definition, so this gives us no guidance for our future decisions. > more than reading a novel is the compression of writing the novel I was just having a conversation over in the jutjusu post about how and why you make commits, when you squash them, when you reorder them, and so on. What is a "pull request" if not a compression of the work I did to develop the required change? I even include a summary at the top so people don't have to read all of the actual code.
GuB-42 1y ago
I think another example of "compression" is car maintenance. Either you learn all there is to your engine, or you follow the manufacturer recommendations and follow the schedule. Sure, it is better to really know your engine, and you probably know someone who will tell you they know better than the schedule, and sometimes, they do. But maybe you have other things to do than to become a mechanic, so you just follow the schedule. It will cost you more, and maybe you could get a bit more longevity or performance by doing things your way, but it frees you for other things that may be more valuable to you than your car.
- wredcoll 1y ago
  50 years ago, people, on average, knew a lot more about maintaining cars. They had to since they broke down far more often. 30 years ago, people knew a lot more about maintaining personal computers. Knowing more things is always good, but there's finite time in which to learn things so you have to choose between them, as you say, it seems pretty rational to select the more valuable ones.
- mms0312 1y ago
  And if you add understanding of incentives to following the manufacture's schedule - even better. Manufacturer has no interest in your car lasting as long as possible, just enough to stay alive before the warranty expires. Taking this into account, do your oil changes earlier than scheduled and the car will cost you a lot less to maintain. I guess that would be somewhere in the middle between the two extremes
cal_dent 1y ago
Its interesting, to me anyway, that most people seem to be in agreement that the core problem is that there's too much information in the world. But the common solution from a tech perspective, and downstream to broader culture by default, seems to be to summarise. So that we can digest more (arguably just surface level) information with the same amount of time. Rather than just choosing what you engage in more consciously. Abundance of everything seems to win. Kinda like how, particularly since Steam, gamers just accumulate more and more games without actually playing most of them at all. Just because its there doesnt mean you need to have it.
theshrike79 1y ago
There was a podcast about this (in Finnish though) how there was a time in the world where you could realistically know _everything_ there is to know. You just had to be rich enough to not have to work for a living so you could just learn and study all the time. But it's been a long time since that was possible.

nradov 1y ago

Up until maybe the 17th century it was at least possible for a single brilliant, educated person to know the entirety of knowledge in Western civilization. We're way past that now.

Night_Thastus 1y ago
I think it was possible for someone from then to think they knew all the knowledge in Western civilization. Even at that point though, it was far too much to keep all in one person's head. They'd only have a surface level familiarity with most topics at best. I mean even one topic like architecture would've been impossible. You'd have to go back a lot further IMO to get it to fit. Somewhere around the development of agriculture, maybe.
barrkel 1y ago
There's a lot of tacit knowledge in trades that was not documented, and in many ways can't be easily documented. You can't really document how to swim or play an instrument; you need to practice and train your body. Would you brilliant person from the 17th century be competent as a member of a tribe in South America or as part of a steppe clan in Asia?
- nradov 1y ago
  Sure, I should have been more specific and limited the statement to academic or intellectual knowledge. Stuff like history, natural history, mathematics, astronomy, philosophy, major works of literature, etc. The point is that up until that time, a wealthy person who devoted themself to study could conceivably learn it all in a lifetime.
  - wredcoll 1y ago
    I've heard similar opinions before and considered them plausible in the past, mostly based on this idea that there were people who "invented" in multiple disciplines in a way that seems impossible today. However, now I wonder if that might be more an effect of the increasing level of communication among humans in general at the time, printing presses, steamships, postal organizations, newsletters, etc, all combined to cause ideas to spread much more widely and more quickly than they did in the past. Perhaps its more a function of being the first (or most popular) person to synthesize certain ideas based on being at the intersection of spreading knowledge.

dkural 1y ago

I think there is a spectrum of compression here, so I find myself agreeing with your overall point but with a caveat. Putting aside "experiences" - travel or emotion - for transmitting information, the level of "desired compression" is based on the background knowledge the receiver brings, and what they consider to be the irrelevant bits. I will say on historical mathematics - reading the greats shows you the road traveled, and often one realizes the unexplored paths along the way, that lead to brand new continents. This may not be as relevant if one is not trying to innovate. I would also say reading Newton as a "first book" is positively counter-productive to learning Calculus in high school.

wredcoll 1y ago
> I will say on historical mathematics - reading the greats shows you the road traveled, and often one realizes the unexplored paths along the way, that lead to brand new continents. This may not be as relevant if one is not trying to innovate. I would consider this a function of specialization, which is of course still good, perhaps even required in this modern world, but we can't specialize in everything.
- dkural 1y ago
  Agreed! I often wonder how LLMs will deal with this eventually. It could be their biggest strength. Some people will want tweet-based responses, some will want book-length. It could generate the right level of content for an audience-of-one, knowing what they know, etc.

lapcat 1y ago

> Compare the journey of Newton inventing calculus to that of the modern day student learning it in a semester. The student is presumably experiencing "compression", but the alternative is they spend their whole life merely returning the point that humanity has already achieved. I wouldn't call learning calculus the compression of inventing calculus, any more than reading a novel is the compression of writing the novel. There's no continuing value in reinventing calculus or rewriting the same novel, but there is continuing value in learning calculus and reading the novel. Compression culture is refusing to do any work, demanding instead a mere summary of calculus or a novel. If you go to class for hours every week, read the calculus textbook all the way through, and do the homework, I wouldn't call that "compression" in the terminology of the article author. On the other hand, if you merely have a summary of calculus, you've actually learned nothing. You can maybe fake your way through a conversation about calc, sound intelligent without being intelligent, at least until someone forces you to do a calc problem and reveals your sham.

j_w 1y ago
Conflating the process of discovering something new and learning it after the fact feels like saying "why do I need to know arithmetic when I have a calculator in my pocket?" Anyone giving real critical thought to either statement knows they are not intending to achieve the same goal. They shouldn't be compared when talking about "compression."
wredcoll 1y ago
> Compression culture is refusing to do any work, demanding instead a mere summary of calculus or a novel This sounds rather tautological. My intent was to give an example of compression that is an obvious good. If we re-define compression to exclude that, pretty quickly we've just reduced the definition to compression = bad things. Bad things are, you know, bad, by definition, so this gives us no guidance for our future decisions. > more than reading a novel is the compression of writing the novel I was just having a conversation over in the jutjusu post about how and why you make commits, when you squash them, when you reorder them, and so on. What is a "pull request" if not a compression of the work I did to develop the required change? I even include a summary at the top so people don't have to read all of the actual code.

GuB-42 1y ago

I think another example of "compression" is car maintenance. Either you learn all there is to your engine, or you follow the manufacturer recommendations and follow the schedule. Sure, it is better to really know your engine, and you probably know someone who will tell you they know better than the schedule, and sometimes, they do. But maybe you have other things to do than to become a mechanic, so you just follow the schedule. It will cost you more, and maybe you could get a bit more longevity or performance by doing things your way, but it frees you for other things that may be more valuable to you than your car.

wredcoll 1y ago
50 years ago, people, on average, knew a lot more about maintaining cars. They had to since they broke down far more often. 30 years ago, people knew a lot more about maintaining personal computers. Knowing more things is always good, but there's finite time in which to learn things so you have to choose between them, as you say, it seems pretty rational to select the more valuable ones.
mms0312 1y ago
And if you add understanding of incentives to following the manufacture's schedule - even better. Manufacturer has no interest in your car lasting as long as possible, just enough to stay alive before the warranty expires. Taking this into account, do your oil changes earlier than scheduled and the car will cost you a lot less to maintain. I guess that would be somewhere in the middle between the two extremes

cal_dent 1y ago

Its interesting, to me anyway, that most people seem to be in agreement that the core problem is that there's too much information in the world. But the common solution from a tech perspective, and downstream to broader culture by default, seems to be to summarise. So that we can digest more (arguably just surface level) information with the same amount of time. Rather than just choosing what you engage in more consciously. Abundance of everything seems to win. Kinda like how, particularly since Steam, gamers just accumulate more and more games without actually playing most of them at all. Just because its there doesnt mean you need to have it.

theshrike79 1y ago

There was a podcast about this (in Finnish though) how there was a time in the world where you could realistically know _everything_ there is to know. You just had to be rich enough to not have to work for a living so you could just learn and study all the time. But it's been a long time since that was possible.