Every complex idea has a million stupid cousins | Hacker News Reader

Every complex idea has a million stupid cousins

apxhard.substack.com

108 pointspaulpauper42 comments

Complex idea - we're running out of easy to reach oil, the amount infrastructure, labor, and energy required to extract a barrel of oil out of the ground will eventually make it counterproductive to drill new wells.

  Strawmen used to cut the idea down:
    Oh, you're talking about climate change
    The price will just go up
    Oh, that's just politics, you don't know the truth
    Technology will solve the problem
    etc., etc.

We here at HN are specialists at picking apart examples. It's a feature when it comes to weeding out con artists and crackpots. It's a bug when it comes to actually discovering new things that haven't been communicated perfectly.

We need to keep this in mind, and try not to be too blind to new facts or ideas.

You articulated it perfectly. I felt exactly what you're describing now while reading the comments to a post about mindfulness a couple of days ago, but I didn't know how to phrase it.

Thank you!

This. The other issue with peak oil is that people have been predicting it since 2005 (if not earlier) and the general picture has been, and still is:

    * we don't fully know when oil production will peak, because it's based on human behavior.
    * when we do pass peak oil, we won't know for a few years that we have. (It's possible that the 2019 local max was "the peak"; it's also possible that it was related to COVID-19 more than oil scarcity.)
    * very few of the issues caused will be attributed to "peak oil" at the time. Instead, they'll be blamed on state collapses, policy failures, economic depressions, and resource wars that will all be indirectly related to peak oil. 
    * peak oil is not necessarily a bad thing (although it probably will be, because the global system is already fragile and nearly universally despised) if we can find replacements in time--unfortunately, we're decades behind where we should be with nuclear energy, which for all its flaws is nevertheless the greenest energy source we've found thus far.

To the outside observer, "peak oil" seems like one of those things people have been worried about forever but that has not yet come to be and thus "never will".

This reminds me of how some of my friends criticize “woke” (i hate the pejorative use of this term) because they’ve seen too many stupid, twisted versions of originally noble and intelligent ideas.

It’s an unfortunate truth that most people who hold the “right” beliefs do so mostly by accident. This works at the surface, to some degree, but breaks down whenever reasoning from principles is required. We should try not to forget that it’s also true of ourselves.

This is an excellent article, and it really reinforces the importance of always trying to genuinely and honestly steelman new ideas.

It's the same with feminism, pacifism, communism, veganism, etc.

I have around me numerous activists about various topics I'm not invested in: feminists, communists and vegans, mostly. Most of them don't try to guilt trip me or convince me of anything. They just do their stuff.

Once in a while, one of them do, just like once in a while, a doctor makes a medical error or a plane crashes.

But we tend to only remember this one, and associate the movement with it.

Is trying to convince you in-and-of-itself a mistake? Or just the approach used?

If you're not careful with this it will turn out to be a general purpose cheat code whenever someone criticizes an idea that you don't want criticized. Just like how crackpots reply with "You just don't understand it" when someone criticizes their theories.

Even if "Woke" ideas aren't originally stupid and twisted [citation needed], the woke ideas that are in power today are. Communism might have been "noble and intelligent" in 1840, but good luck convincing those living under it in 1970 of that.

I’m not talking about communism, I’m talking mostly about diversity and inclusion stuff.

But yes, you could apply the same to communism: just because a lot of people are saying stupid things about communism doesn’t mean that communism is bad. Communism could be bad or good, but it should be judged on the merits of its strongest conception.

And I really don’t think steelmanning arguments is a slippery slope: if anything it’s an uphill battle that requires constant effort.

My comment shouldn’t be intrepreted as serving as a shield against criticism, it should be interpreted as motivating a larger degree of good faith effort at the personal level.

My point is that this all irrelevant when ideas pose a threat to people. Ideas have real world consequences. It might seem easy for you to approach communism with good faith because it's a harmless ideology in your time and place, but those who lived a terrible life under it might disagree, and you can't really blame them. It's a bit rich to demand good faith out of people towards their aggressor.

Also, good faith is a 2-way street. If a ridiculous drawing of a horse is defended by otherwise intelligent and honest people, yeah I might be interested what do they have in mind that makes them say so.

If a ridiculous drawing of a horse is defended by a bunch of rage zombies who know no other disagreement strategy except censoring, bullying and otherwise exercising coercion, I'm guessing that whatever they have to say is not that high IQ or high utility. What's the point? Ideas are cheap, whatever intelligent or useful things those modern Puritans have to say have likely been said a thousand thousand times in much better and articulate forms.

I don't think anyone is suggesting that we need to tolerate aggression, or that we must entertain stupidities. The main message and lesson to take from this post is rather to not let the stupidities stand in the way of learning.

The point is that if you pick any complicated topic, and try to learn about it by listening to proponents at random, 19 times out of 20 [1], you're going to land on a stupid explanation that doesn't even make sense. If your goal is to learn [2] then you should continue keeping an open mind until you land on that 1 out of 20, higher-quality version of the thing, and then give that one a chance. If the highest-quality version of the thing makes no sense, then you may be justified in discounting the idea. [3]

This notion might seem obvious to some people, but from my experience, I can say that not everyone finds this obvious.

[1] I just made up these numbers for demonstration purposes.

[2] There are other goals in life too (eg.: surviving, escaping oppression, etc), but this is one possible goal and it's the one we're focusing on in this context.

[3] But also, statistically speaking, it's very likely that we just don't understand the thing, so this is why I often like to say that one should be more skeptical of one's own understanding than of others (again, in contexts where the main goal is learning).

This is what happens to children. They do something they think is fun and their parents tell them not to do it. They ask why not? The parents either give a nonsensical half-answer because they don’t know how or don’t want to give a full answer (and aren’t sure their children could understand it), or they give a non-answer like “because I said so.” The child quickly gets the take-away message that parents are arbitrary assholes who have control issues, so they learn to just disregard everything their parents say. At least that’s what happened to me and every person I grew up with.

I realize that children don’t always have the mental skills to understand everything the way adults do, but there has to be a better way to handle these situations than they are currently being handled.

So... the example the author provides is super contrived. There's no problem with that. But I would expect a contrived example demonstrate the point without obvious holes. So I'm going to poke an obvious hole which IMO extends to overall idea the author is presenting.

Claiming that a horse is majestic is not an "idea" in the "product idea" or "feature idea" sense.

But let's assume for a moment that someone imagines a majestic creature in their mind. The person sees this beautiful majestic creature in their head. They can't wait to tell everyone about this beautiful idea. And when they decide to share it... they share an objectively ugly ass stick figure drawing...

Do you see the problem?

How one presents an idea is critical in validating it for a broader organization or society. If the person with the idea can't tell the difference between an ugly ass stick figure drawing and the majesty in their mind then how are you supposed to trust their vision? Their understanding of majesty is obviously disconnected from reality.

Okay let's translate this to something more real world.

Suppose you're a principal engineer or architect or whatever those people are called these days. You come in to work on a Monday and a mid-level engineer asks to meet you because they have a great idea. They were so excited that they even created a little demo over the weekend to show it off.

The demo goes horribly. Every step of the way something breaks. The engineer can't clearly describe what the idea is as they stumble through layers of complex problems and solutions. The meeting ends and you walk away confused and mildly annoyed at having your time wasted.

Perhaps the engineer did discover something interesting. But if they're not able to distill a demo into what makes their idea awesome then it would be irresponsible for you to blindly trust that the idea would work in practice. Now obviously in this context "idea" doesn't mean "use this function instead of that function and you get a 10x perf boost". That's simple. I'm talking about something more complicated like a new paradigm or way to architect your systems.

My point is this. Presentation matters. It doesn't need to be pretty. You don't need to be eloquent. But when you present an idea it needs to do an exceptional job of demonstrating itself to peers. It's not the job of your peers to unpack your brilliance - that's on you.

I’ve been pitched ideas many times, some were good and some were not so good. What I liked to see was a clear presentation because that usually meant clear thinking behind the idea.

A heuristics based approach like yours is very underrated.

>> when you present an idea it needs to do an exceptional job of demonstrating itself to peers. It's not the job of your peers to unpack your brilliance - that's on you.

For my sins I'm not just a programmer, but I also need to (literally) sell my work to other programmers.

Over many years I've come to appreciate the idea of selling benefits, not features. Yes, utility is important, but ultimately people want benefits, not feature.

"saves time" is the benefit, "integrates with xero" is the feature.

Focusing on the selling _while_I am doing the creating, helps - I'll be writing the docs, and it's just complicated to explain, so I'll go back to the creating to make it easier. I'll be prepping for sales, and I find the rough edges, so it goes back for more polishing.

Selling ideas is the same. These days I can do what I like - the track record helps - but in the beginning I had to do a lot of work preparing the idea "for sale" to my bosses. If I can't sell it to them, how can I sell it to someone else. Equally, I had a bunch of bad ideas - most I filtered out myself, others were filtered for me.

I still tend to have a "big picture" in my head, of how things will plug together, but I seldom sell the big picture - I find utility for each part, make them successful, and at the end pull the threads together to show what I had in my head all along.

This is a good point. On the other hand, drawing a compelling picture of a horse is a hard problem. I could do better than the stick figure but even my best attempt would end up looking pretty awkward and weird compared to a real horse.

I would need a lot of practice and a lot more detailed knowledge of horses to be able to communicate effectively.

This gets me thinking about how it is a lot easier to become convinced of an idea than it is to share it in a convincing way. It's much easier to follow along with a good explanation or argument than it is to recreate one after the fact.

It takes really deep detailed knowledge of a topic and strong communication skills to be able to convey a complex idea in an accurate, understandable and attractive way, especially if the audience is not interested in investing a huge amount of time on a deep dive into all the details.

I can see how it's a lot easier for short-handed versions or simple-but-wrong versions to spread, especially when someone without a deep understanding is trying to share what they've learned from an expert.

Yea, it's tricky.

The conception of an idea and the sharing of that idea could be completely different skills in some cases. I'm reminded of the classic Jobs/Wozniak duo that we often see repeated in the tech industry. There's definitely something there.

I was put off by the fact that the author apparently has examples, but refuses to share any, because they're "infohazards". Instead they advance an argument by analogy.

If this effect, memetic inoculation, is so prevalent, we should be able to find some ideas that we recognize as good today, but which took far longer than they should have to gain traction because people in the past had memetic inoculation against their simplified variants. I fail to see how that would be an infohazard.

> the author apparently has examples, but refuses to share any

Evolution.

Mic drop . . .

I assumed the author was talking about politics. It's natural for complex political ideas to be strawmanned by their opponents, and I've heard of strawmanning working as a sort of brain vaccine against better versions of arguments before.

> I've heard of strawmanning working as a sort of brain vaccine

Where, and from whom?

There was a paper or article which was shared on HN not too long ago... unfortunately I don't have the link saved and I'm not even sure how to search for it.

Edit: found it... Looks like it's just an opinion piece and not some piece of research [1].

1. https://www.slatestarcodexabridged.com/Weak-Men-Are-Superwea...

I think the author are making a good description of what we all can observe in our times. It's the inclination people have to prefer, or even want, simple answers to anything. Just look at how media is handling the complexity of things. They always look for experts who are willing to give short and simple answers. "complicated ideas take more energy to transmit than simple ones" he writes, and therein lies the problem.

The article is itself is a simple explanation to a complex mechanism in human society where people establish truths that are tangible but not (necessarily) true. This is what fuels opinions and even cause actions. The simpler the explanation the faster it spreads. And it spreads among those of us who are too lazy to get acquainted with a complex issue on topics we do not know much or even are ignorant about. I say “us” because even though I feel I am quite well informed and have a high endurance when it comes to acquire information, I know I am still totally ignorant on many topics that are outside my interest and competence, or doesn’t show up in my daily information flow. I think we all are, and that we therefore carry a lot of “stupid cousins” in our mind that we even might be spreading around without thinking about it.

The world and our society is complicated and it is very easy to simplify to try to understand why things are as they are. Within this simplification process we obscure the reality more than we enlighten it, and even establish totally wrong ideas of the state of things. When someone who really knows the complexity of the horse tries to explain to us that we haven’t got the whole picture and therefore have gotten it wrong, we do not listen. We all know how “the horse” looks like, why listen to experts who have actually knowledge about horses? This unwillingness to listen to experts is best seen in politics, and best in highly polarized societies. But it’s really everywhere.

  s/complex idea/reality/
  s/stupid copy/model/

After these substitutions it becomes clear where the author is mistaken.

"Reality has a surprising amount of detail." That's why we need an abstractions - models ("stupid copies").

We may have many different stupid copies, depending on a questions we want to answer, or problem to solve.

Every stupid copy is wrong, but some are useful.

Even a stick drawing of horse can be useful, if all we want to do is count the number of legs the horse has.

While the point you're making in your comment is very valid, I really don't think "models vs reality" or "all models are wrong but some are useful" is what the article is about.

1. Maybe. But if the article is about communicating ideas, then I don't understand why to use a horse as an example?

horse != idea

Idea is by definition something abstract and/or intangible.

2. IMO, complex ideas by definition are bad ideas. Good ideas are simple, but they might be contrarian, seemingly impossible and/or non-obvious.

EDIT: found a really stupid idea: a USB stick horse ;)

https://www.amazon.com/Horse-Shape-Flash-Drives-Memory/dp/B0...

This comment seems almost right. The author implicitly declares the existence of "memetic inoculation" as something they're describing, although they didn't specify cases where it becomes difficult to separate a bad idea from a bad description of an idea.

As someone else pointed out, crypto is an example where there are obviously good and bad ideas: using cryptography for distributed and decentralized contracts is a at the very least quite intriguing. Using crypto for a scam is quite a bad idea.

But, then there are ideas that are really good, but who's outworkings are usually described poorly. Take Emacs as an example. "text editor that does everything, including email, accounting, browsing the web, file management, and much more" sounds like a bad idea to anyone with a modicum of appreciation for the Unix Philosophy. "An app that makes programming and using practical text-oriented, highly-integrative applications in Lisp" is a better description of what Emacs really is.

Notice that in the former case, crypto and scams using crypto are two distinctly different ideas. In the latter case, we have two descriptions of essentially the same thing, which was an implementation of the modern idea of Emacs, which is factually a Lisp interpreter with built in text display. Those two descriptions could in theory be the basis for creating different apps, and they'd probably look quite different.

This is precisely the phenomenon that happens in HN with cryptocurrency-related topics. The stupid cousins of crypto (scams, ponzis, pump-and-dumps) have activated a heuristic in an intelligent body of people, leading to a blanket dismissal of crypto products.

Others (like myself) get defensive when their convictions are dismissed out of hand. And hence crypto/blockchain threads here tend to be the most polarizing.

Can you point me to the legitimate usages of cryptocurrency?

“a system for electronic transactions without relying on trust.“

- From the original Bitcoin paper.

There is always some measure of trust involved in any transaction.

The really hard problems of trust in transactions aren’t really about actual transfer of currency but stuff related stuff like can I change my mind, am I getting what I think I paid for, etc.

Bitcoin, at great resource cost, tries to solve the problem of trust that most people don’t actually have.

As the list would take me ages to write down, here are a few obvious ones:

- Global financial settlement, including multi-party and Very Large sums, in under an hour for a few dollars.

- Collateralized permissionless loans

- Global Open-source computing in a shared execution environment

- Proof of ownership that's quickly verifiable

- Insurance settlements

There's a limitless amount of bespoke use cases as well, as we'll continue to see pop up over the next decade.

I read the article and I kept looking at the horse drawing and thinking: I'm sure I've seen really majestic yet astonishingly simple drawings of horses somewhere. Then I remembered - the White Horse of Uffington[1]

There's a second one in Kent, carved into the North Downs at the turn of this century[2]. Personally I don't think it has the same majesty as the ancient one.

[1] - https://en.wikipedia.org/wiki/Uffington_White_Horse

[2] - https://en.wikipedia.org/wiki/Folkestone_White_Horse

A lot of religious ideas fall into this basket, particularly their manifestations in the USA. You could make a strong argument that American televangelists and their ilk killed Christianity in Western Europe.

Looking at Europe's history of Christian-on-Christian religious wars - century after century after century... And at the blind eye which major "Christian" churches in Germany turned to the Holocaust. And the systematic horrors (espec. child sexual abuse) in many Church-run organizations. And increasing education. And the ...

I'd say that the established Christian institutions of Western Europe mostly killed themselves, with no American televangelical help needed.

The "Christian" church in nazi germany was a state sponsored group that got processively diluted down to emptiness.

One of Hitler's goons stated that in the utopian future the churches will not be granted licenses to rebuild their churches inside the city or anywhere. The nazis fully intended to get rid of christianity and hitler was to be the head of a new religion, nietschze style.

The real Christian churches were persecuted by the Nazis inside Germany. The nazis sent out some officers to the monastry on Mt Athos (orthodox monks) to take home relics. The officers that arrived at Mt Athos decided not to ransack the place and just hung out instead.

Philosophers commonly bridge a lot of the gaps between dumb ideas and sophistacted ones with very sharp sentences and poetry.

It is possible to compress complex down into smaller packages, but very few people do it well. Often working out the complex idea is hard enough without having to express it poetically.

I humbly submit another hypothetical that might help explain this idea:

A drawing of a plane and the claim that “this thing can fly”.

Unless you capture extremely subtle aspects about the materials and curvature of components, the depicted plane almost certainly won’t fly.

Really glad I read the whole article to get the punchline. Almost bounced because the stick horse seemed like a dumb example - must’ve seen it’s cousin before

The even more complex version of the title would be "Every complex idea has an unbounded number of less complex variants"

Check out "The Righteous Mind" by Jonathan Haidt.

So is this guy's idea the stupid cousin or not?

Reply on news.ycombinator.com