Civilization Is Built on Code (2017) | Hacker News Reader

Civilization Is Built on Code (2017)

nautil.us

96 pointsdnetesn32 comments

There's a lot that can be unpacked here -- several accurate observations, but in my opinion the author stretches the metaphor too wide to accommodate a narrative, then offers an underwhelming conclusion in turn.

They simply mean recipes, instructions, laws; anything that communicates instructions of desired behavior. Even they acknowledge that there's a wide spread of optionality, from good-faith advice to laws backed by the treat of state violence, but at this point the metaphor ceases to be useful. Humans have always been telling other humans what to do, but civilization is built from the combined survivors of thousands of years of experimentation, adaptation, governance, and passing down good advice. Saying it's built on 'code' is no more revealing than saying it's built on actions, or the shoulders of ancestors, or imitation, especially when you expand definition of code to encompass any sort of prescriptive (or descriptive!) record about behavior.

When I think about treating laws and recipes as code, I think of the difference between absolute machine computation vs a process that collaborates with human judgment. The "stupidity" of computers' literal-mindedness is something that non-programmers just don't seem to understand. Some recommended things to consider:

- Aristotle: laws are for the general case, judges for the particular. This gives the saying, "Hard cases make bad law." It's good to have rules, but keep people in control.

- Michael Polanyi: not all knowledge is "formalizable". Actually recipes are an example of his: two cooks following the same recipe may give drastically different results.

- Roger Penrose: If the human mind were a computer, why does it seem not bound by the constraints Gödel's theorem implies for a formal axiomatic system?

I wonder if the author's book engages with any of these people?

The first quote makes sense, the latter two are at best mistaken.

> Michael Polanyi: not all knowledge is "formalizable". Actually recipes are an example of his: two cooks following the same recipe may give drastically different results.

... or rather, our formalizations are lossy for the purpose of their usage, lossily interpreted by their users, or simply not good.

Recipes are a good example of a lossy formalization, lossily interpreted (very much on purpose, to keep the recipe a reasonable size and to put only a reasonable amount of effort/time into cooking for the circumstance).

They are in no way indicative of our inability to formalize (and indeed changes due to environment, whether they be ingredients, humidity, temperature, etc. are in fact encoded in some more sophisticated recipes, or more commonly in baking).

> Roger Penrose: If the human mind were a computer, why does it seem not bound by the constraints Gödel's theorem implies for a formal axiomatic system?

It's 'extremely hard' to show that the mind would somehow escape being bound by this (a very incomplete reading as to why can be found at [1]).

This is like suggesting that the mind isn't subject to complexity theory by observing that we can factor 'large numbers' in reasonable time.

[1]: https://thefutureprimaeval.net/godel-vs-the-human-brain/

I recently finished reading Snow Crash (1992), which touches on this topic in an interesting way. The general premise without giving too much away is that ancient Sumarian was made up of a specific structure that acted as a kind of assembly language for the human brain, enabling self-executing code like "How to Bake Bread" or "How to Plant a Field". A villager could go to the "source code repository" aka "the temple", get a recipe for some function they wished to complete (or needed to be completed by the village in general), and execute it by having it read to them by the priest.

This was a way to bootstrap human society, ultimately giving rise to free thought and "higher level language" that was not bound by self execution.

It was a pleasant surprise and I found myself enjoying it immensely, even with a bit of suspension of disbelief.

I loved Snow Crash ( https://www.amazon.com/Snow-Crash-Neal-Stephenson/dp/1491515... ) too.

Something I find interesting is that the instructions described in the book, the Sumerian 'Me', actually existed. They were instructions on how to do fundamental things, managed by the priests and which the Sumerians believed came from their gods and goddesses. The concept of neurolinguistic programming by the Sumerians is a nice fiction, or at least there's never been any evidence of it.

A good book on it is Sumerians: Their History, Culture, and Character, by Samuel Noah Kramer. https://www.amazon.com/Sumerians-History-Culture-Character-P...

An online version can be found at his university: https://oi.uchicago.edu/sites/oi.uchicago.edu/files/uploads/...

Interesting bits around the Me are around page 116.

Very cool, thanks for sharing.

There's a cool comparison between english and the shawnee language:

https://en.wikipedia.org/wiki/Linguistic_relativity#/media/F...

The Shawnee language example is indeed imperative, like low level computer code.

The problem with treating laws as code - which seems to be a popular idea these days - is that by trying to do so, we'd be ignoring lots of hidden complexity. All humans have brain with more-less the same architecture and the same "firmware", we also share stupid amounts of common context (all those years spent constantly learning and observing). All of that lets us skip stupidly high amounts of computation when dealing with each other. Things like trust, reputation, or just two people looking at a rule and both knowing what it really means and was intended for (and knowing the other party knows it too) - all of these are huge efficiency gains. I don't think it's wise to try and throw it all away, because trusting another human suddenly went out of fashion.

I'm pretty sure the author meant "structures & processes" not "code".

It seems you may be correct in what was intended:

> The word “code” derives from the Latin codex, meaning “a system of laws.” Today “code” is used in various distinct contexts—computer code, genetic code, cryptologic code (such as Morse code), ethical code, building code, and so forth—each of which has a common feature: They all contain instructions that describe a process.

[from the article's third paragraph]

Yep - I feel the article addressed that point fairly well, explaining where the word code came from, how it is used today, and the specific take on it they used when writing the article.

There is a feedback mechanism between civilization and a societies willingness to commit to code. There is nothing automatic here. People can reap the benefits if they commit, but they don't always commit. We tend to associate the early civilizations with great monuments, but they key difference between early societies and advanced ones is the willingness to commit to process. In different ways, this is a point made by the Legalist tradition in China, the Executionist movement in Poland during the 1400s and 1500s, and then the "liberal" tradition during the Enlightenment in Europe during the 1700s. And it is no coincidence that these periods coincide with the peak of Chinese civilization, and then the peak of Poland's great Commonwealth, and then later the peak of Europe's success.

Regarding Poland:

"A purely political reformist movement had come into existence at the beginning of the 1500s. In spirit it was very close to the Reformation, since it placed the accent not on innovation but on stricter observance of the law, on weeding out malpractice and corruption. It was known as “the movement for the execution of laws”, or simply as the “executionist” movement. One of its first preoccupations was that the law itself should be codified and published in clear form, and as a result much groundwork was done in the first half of the 1500s, culminating in legal reforms passed in 1578 which fixed the legal system for the next two hundred years [[ till the death of Poland ]]"

http://www.smashcompany.com/philosophy/poland-was-shockingly...

I don’t find this to be a compelling theory. I’m currently reading “James Madison: A Life Reconsidered”, and it strikes me that at the beginning of the Confederacy it was surprisingly difficult to get states to actually comply with treaties or laws made by congress. That they were able to actually get the Constitution ratified is truly amazing.

Fast forward to today, and some of the same problems persists - however the level of respect for procedure seems to be similar.

"it was surprisingly difficult to get states to actually comply with treaties or laws made by congress"

Yes, exactly, and so they lost the war. It was the more process-oriented North that won the war. Despite the fact that Lincoln violated habeas corpus, the Northern territories otherwise continued to follow the law, raise taxes, obey the draft, obey the courts, and get organized enough to win. The South never had that kind of organization, nor any commitment to building a society where everyone would follow such a process. Indeed, one reason that rebellions tend to fail is they are made of people who want to break the rules, the very rules that might allow their rebellion to succeed.

This was long before the civil war. And my objection was to the idea that advanced countries somehow subscribe less to process - I don’t see compelling evidence that this is true.

You meant the Articles of Confederation period - he mistook your use of Confederacy for the civil war period, most Americans will.

Fair enough :) The Articles of Confederation obviously resulted in a (smaller) Confederacy in the 18th century, but as a Canadian I’m less familiar with the popular parlance that would be taught in American schools.

Sorry, I did misunderstand your use of "Confederacy" although you are right that the USA started off as a Confederacy.

But the point still holds. The notion that a loose collection of small states could survive had been brought into contempt when Britain defeated the Dutch states. Many people argued that the North American colonies were making a mistake by imitating the loose structure that had already failed in the Netherlands. And it also failed in North America. A better organized Britain was able to defeat the Dutch. A badly organized Confederacy almost collapsed because it was unable to manage its debt. A better organized nation under the Constitution was able to pay its debts and start the process of building a cohesive society.

I'm generally critical of anyone who wants to generalize much from history, since history is full of so many contradictory trends, which play out simultaneously, and what actually happens tends to be like the waves we see at the surface of the ocean, which are always the coincidence of several waves crossing the same point at the same time, and often going in opposite directions. But that process-oriented societies have tended to beat less organized societies seems to be a thesis that can withstand a lot of scrutiny.

I agree with most of what you're saying. I think I actually may have misinterpreted this line: "We tend to associate the early civilizations with great monuments, but they key difference between early societies and advanced ones is the willingness to commit to process."

I initially misread that as "earlier societies were more process-abiding than advanced ones", but on reflection you were likely stating the opposite. I shouldn't comment before I've had a cup of coffee in the morning ;-)

Sorry, that might have been badly written. When I write on my blog I normally hire an editor to catch those kinds of sentences, which can be misunderstood. Maybe the safety net of having an editor has made me lazy. When I write a comment on a place like Hacker News, I should really re-read it to make sure it is clear.

As an aside, just grabbed a copy of your book - looks like an interesting read, and I can definitely relate :-P

Thank you. I hope you find it both humorous and also educational.

Is there a link between morality and adherence to process? For example, if you know something is wrong but you do it anyway, are you going to then generally avoid process because you have an internal conflict?

Civilization absorbs technology.

"Learning dominates when technologies are new and not very useful for broad human purposes. If you want to explore a coral reef in 2017, buy a plane ticket to the tropics. If instead you buy “FishWorld, a fully immersive underwater experience for Oculus and Gear” you will learn more about virtual reality technology than you do about reefs. Now cast your mind forward: in 2037, you will not play with or think about “virtual reality”. Either it will be defunct, or it will be so useful that we speak of an evening at home exploring coral reefs. The technology will fail and evaporate, or it will succeed and be absorbed, but either way, it will disappear."

http://www.sampenrose.net/civilization-absorbs-technology/

Books have been around for quite a while, yet we still generally say we've been reading about coral reefs rather than exploring them.

I've seen plenty of people who use both types of phrases interchangeably.

Point valid; video-chat has been around for a while too but in-person conversations still is a higher-quality experience.

When I was in high school we watched about coral reefs thanks to "Laserdisc". VR is the new "Laserdisc".

>> As code advances, higher-level technologies feed on more fundamental technologies in much the same way more complex organisms feed on simpler organisms in the food chain.

This is why it's a bad idea to export your lower level technologies to other countries. You're still dependent on them, but lose control.

Civilization is built on communication.

More specifically, the ability to record ideas.

It's one thing to simply communicate, but to be able to record ideas and effectively enable cross-generational communication is extremely powerful. Native Americans, for example, had verbal communication and some limited written communication, but much of their knowledge has been long lost because it was passed down verbally. Of course, the United States did a good job at erasing their history by exterminating the native people, but you can still look to the Aztec civilization as a contrast in how recording ideas had an impact in how much we understand about their history.

Top clueless moron technotopian article of 2018.

"How did we humans manage to build a global civilization on the cusp of colonizing other planets?"

No, we are on the cusp of climatological, ecological, environmental and economical disaster.

Always have to laugh a little at self-aggrandizing nonsense.

Without physical hardware, there would be nothing on which “code” could be built.

Atoms came first. Bits followed. History, people.

Reply on news.ycombinator.com