That Sounds Smart | Hacker News Reader

That Sounds Smart

aaronsw.com

71 pointsrsaarelm67 comments

> But the larger consequence is that if you’re smart the world doesn’t seem very complicated. This might seem obvious, but the obvious thought is rather different. The obvious thought is: The world doesn’t seem complicated to smart people. But this isn’t what smart people actually think. They think the world isn’t complicated, period.

Woah, woah, woah. "The world" is a very large thing indeed. You'd have to be pretty stupid to think "it isn't complicated, period." Simplifying insight is often hard won. And some things are just irreducibly complex. To think otherwise smells like Dunning-Krueger.

Give me an example of an irreducibly complex thing :)

You.

Fail, please refer to Evolution.

Evolution is just a way of explaining how someone came to be, not how they work. Being a good materialist I do think that people are reducible, but we certainly haven't been able to figure all of the steps out yet.

* The Belgian question.

* Quantum mechanics.

* The sovereign debt problem.

http://en.wikipedia.org/wiki/Irreducible_complexity#Response...

Also, I would think that exptime-complete problems we could call irreducibly complex :)

My first thoughts strayed to computability theory though.

Oh yeah, that's why the phrase "irreducibly complex" popped into my mind. Whoops.

Just to be clear, I definitely wasn't referring to the intelligent design argument. I meant the phrase strictly in the "make everything as simple as possible, but no simpler" sense. The simplest possible correct explanation of some things is still extremely complicated.

Next month's weather.

Why people hate.

I think Bertrand Russell put it best when he wrote, "[t]he fundamental cause of trouble in the world today is that the stupid are cocksure while the intelligent are full of doubt."

I have found the smarter (and in the majority, not all) someone is, the more likely that person sees the complexity of things.

That reminds me of Richard Feynman, who tried to keep a list of everything he didn't understand. Smart people pay more attention to what they don't know than what they do.

"Real knowledge is to know the extent of one's ignorance." Confuscius

I think there is a difference between "complex" and "complicated". It might require more effort to delve into the details of complexity, but to the extent that you can recognize where the effort needs to be applied to develop a reliable understanding, it's not "complicated" in the sense of being incomprehensible.

The author mentioned that "normal" people will often presume that someone who sounds complicated is smart, but I've seen the converse bias among smart people: they expect things to make sense, so when an explanation seems too convoluted, they tend to dismiss it as bullshit.

But, I suspect, for most people the world is a strange and mysterious place, governed by principles they do not understand, which affect them severely but cannot be controlled, only coped with as best as possible.

The world is a strange and mysterious place, governed by principles we do not understand, which affect us severely but cannot be controlled, only coped with as best as possible. People that think otherwise might not be as smart as they consider themselves to be.

I do agree with the overall point. Jargon can be used to simplify communication by using a language all parties understand, or to mask the fact that the speaker has no idea what they're talking about. But I take issue with the argument that everything can be described in a simple way. To quote Feynman, "If I could explain it to the average person, I wouldn't have been worth the Nobel Prize".

> [Smart people] think the world isn’t complicated, period.

um.. what!?

The world is a complicated mess in just about every aspect except in those axiomatic systems we developed to create simplified models -- but the map is not the territory. The world is by far larger than those specialized areas we're good at.

"As the circle of light increases, so does the circumference of darkness around it." -- Albert Einstein

Sure, but I think that there was an implicit assumption that the person under consideration was explaining something within their area of expertise. Otherwise the article would be claiming that a smart person can explain anything simply, which amounts to equating "smart" with "omniscient".

The other assumption I made about the article was that it was discussing a useful heuristic, not a 100% reliable test. The inherent complexity of the world is one of the things that can make it give the wrong result, but not the only one.

Perhaps that is true, but I feel that if the blogger really understood what he was talking about, he would have been able to phrase it in a way that wouldn't cause any confusion. ;)

if you genuinely understand something — really, truly understand it — then it doesn’t seem complicated and you can explain it rather simply

Uh, no? Some things are inherently complex. Try explaining how the Y combinator works (not what it does - why it works) to a layman in simple terms. I don't think it can be done.

OK, let's try.

I think it can be done in 40 minutes, one hour tops. 15 minutes to explain what a function is:

  f1 : ℝ  -> ℝ
  f1 = x |-> a × x + b

  f2 : (ℝ × ℝ)  -> ℝ
  f2 = (x,  y) |-> x + y - 1

  f3 :  ℝ -> ℝ  -> ℝ
  f3 = (x,  y) |-> x + y - 1
  (I fear curryfication can't be skipped)

  f4 : (ℝ -> ℝ) -> ℝ  -> ℝ
  f4 =  f      |-> x |-> f(x)

  f5 : ∀ a, ∀ b, (a -> b)  -> a  -> b
  f5 =            f       |-> x |-> f(x)
  (Polymorphic types may be left out, but I think
   they're important)

Then 5 minutes to explain what a recursive function is:

  fac : ℕ  -> ℕ
  fac = 0 |-> 1
        x |-> x × fac(x - 1)

Then 5 minutes to present a fix point combinator…

  fix : ∀ a, (a -> a) -> a
  fix =      f       |-> f (fix f)
  (or we could just say that a is ℕ)

…and implement the factorial with it.

  ffac : (ℕ -> ℕ) -> ℕ  -> ℕ
  ffac = f       |-> 0 |-> 1
         f       |-> x |-> x × (f(x - 1))
  
  fac : ℕ -> ℕ
  fac = fix(fac)

We can do a step by step, normal order evaluation. If any question is asked, just say that evaluating the arguments first just won't work here.

Then, maybe 5 more minutes to explain the concept of self reference without actually naming oneself:

  ~PM(~P)

Where you assume that for any "printable" string X, PX is true, ~PX is false; that for any X such that XM(X) is printable, PM(X) is true; and that any printable string is true, or just don't have the form PX or ~PX. You will note that ~PM(~P) is true and not printable, because it basically says "I am not printable". This is an anonymous self reference.

Finally, we can go on and present the Y combinator itself. It does the same thing as the "fix" combinator, we just have to show that's because there is an anonymous self reference somewhere. I think it can be done in 10 minutes.

Phew. Done. And I didn't use any knowledge that a high school kid shouldn't have. Did I miss something?

> Did I miss something?

Just the part about explaining it to a layman in simple terms..

Of course, the raw text above won't suffice. However, you will note that the notation I used is learned in high school.

To me high school mathematics are quite accessible, though I understand that most people forget it as soon as they can. So, I just assume I will have to start from kindergarten arithmetic. And I had, in fact.

I once explained functions to a 14 year old girl. I went as far as high order functions (function composition, I think), and she understood. I checked, she could answer my questions correctly. She was bright, but still. I think I was able to explain functions in rather simple terms.

Now, if I can explain functions to a 14 year old girl, the rest is certainly doable.

"Of course, the raw text above won't suffice."

http://annieinfinite.com/wp-content/uploads/2009/10/then-a-m...

Phew. Done. And I didn't use any knowledge that a high school kid shouldn't have. Did I miss something?

One thing. How long it takes most people to understand a concept. It took me much more than 15 minutes to understand what a function is in high school, and I have spent far more than 15 minutes explaining it to other people. And it definitely took more than 5 minutes to understand the concept of a recursive function.

And you are assuming they have knowledge of concepts like string and whether or not they are printable...

Now, understanding a concept and grasping it to the point it becomes part of yourself are not the same thing. I don't expect people to remember the whole thing in less than an hour. Just to understand it, and being able to read supporting material without drowning.

A string is a sequence of characters, and "printable" is some arbitrary property. I would of course say as much in my real explanation.

Understand that my comment was just a sketch. I wonder If I shouldn't actually try to flesh it out. A video, maybe…

Please do. I want something to point people to.

"And I didn't use any knowledge that a high school kid shouldn't have"

...a high-school kid that has studied haskell maybe.

Either you're being arrogant or you are very lucky to have gone to a school which teaches rudimentary type theory.

I once explained it to a literature PhD, weak at math, who'd taught a course on writing under censorship. I told a short self-referential story whose structure mirrored the Y combinator's -- about tricking censors charged with banning self-referential stories.

(Yes, many things are just complicated. But confusing explanations usually could be improved, just as confusing code usually could be lots clearer.)

"Smart people actually say things that are very simple and easy to understand. And the smarter they are, the more clear what they say is. It’s stupid people who say things that are hard to understand."

This is just one of those ideas that is good in theory, but simply isn't in line with reality. A smart person can make things as simple as possible, but they're not magic. You can't take a complex idea and magically make it simple.

  People who truly understand their subject should have no
  trouble writing for a popular audience.

This is false for at least two reasons.

1) People who truly understand their subject know they can't write for a popular audience without handwaving, partial truths and actually getting the complexity of the subject across. They may not want to write in that way.

2) The ability to 'think like a layman', properly estimate their background and slowly build up a story that they can follow without getting lost is a skill entirely different from 'knowing your subject'. It is prepostorous to suppose every intelligent person has that skill. It's a typical case where a lot of people mistake intelligence for skill and it's insulting to (technical) writers that have actually put effort into learning how to write.

All in all, I doubt whether 10% of all people that 'truly understand their subject' can write for a popular audience.

Right. I like the sentiment of the post but it's definitely not a black and white case of:

1) You are stupid if you are hard to understand 2) You are smart if you are easy to understand

They say the best teachers and coaches (i.e. the ones which are easiest to understand) are the ones who had to work the hardest to find success in their craft. Conversely, the most naturally gifted of people sometimes don't understand how amazing what they take for granted is and can't pass on this knowledge. There is also a lot of real world contradictions to this (I'm thinking sports here) so yes, there are a lot of shades of grey.

Yes, there are so many fallacies in that essay. It shows another way to pretend that you're smart: appeal to your audience's vanity by suggesting that you share a rare resistance to manipulation by intellectual intimidation.

Another fallacy is that smart people are more likely to be found behind profanity-laced rants. In my experience, people resort to vulgarity as a gimmick intended to convey a message that isn't adequately conveyed by their content itself. The emotional, gimmicky nature of it is the reason that it tends to be negatively correlated with logical quality.

University academics can be a good example of this. Sometimes experts in a given field can be terrible lecturers/writers, yet no-one would doubt their abilities.

The author blatantly ignores modes of communications.

A discussion of two specialists' will be completely impenetrable to a layman. However, this fact says nothing about the specialists'. An 'intelligent' specialist should be able to communicate ideas from his field to a layman, but the efficiency of this communication will be much lower than of the discussion from the example above.

I agree, not to mention, some subjects are too complicated to explain to a layman, a lot of mathematical concepts (just as one example) require certain non-trivial fundamental knowledge in the layman before you start, if these fundamentals are lacking, you are stuck.

I think the author is just projecting.

1. There are plenty of smart people that are bad communicators.

2. Life is compicated and if you can't see the complexity you probably suffer from the Dunning-Kruger effect.

From article:

But the larger consequence is that if you’re smart the world doesn’t seem very complicated

This is only for the part of reality designed by humans. Other parts of it are complicated. Not realizing this is a sign of ignorance.

I think he's conflating "smart"[1] and "good at communicating".

[1] In fact "smart" should probably be restricted based on domain. You can be a brilliant violin player but that doesn't mean you'll be a brilliant molecular biologist.

I think you're conflating "smart" with "brilliant". I know several very smart people who aren't particularly brilliant at anything, and a couple who are brilliant at their domain, but really cannot be described as smart.

If they aren't particularly good at anything, how do you know that they're smart? They could merely sound smart to you because they seem (for example) articulate and hold similar opinions.

They're very good at many things, just not brilliant at any. They have fairly deep knowledge in a very wide variety of fields, they can grasp new concepts easily, they can quickly construct and understand both sound and intelligent arguments and are generally all round very intelligent people. They've simply taken a more jack of all trades approach to things, rather than focusing on becoming brilliant at something.

If you refuse to learn anything that isn't explained in a jargon-free blog post, the world will seem devoid of anything that can't be.

This definitely is a bit over-generalized, but I think it makes some good points. One of the things I admire most about a few of the smartest people I have met is their ability to explain things in a very simple way. Furthermore, they are able to break down problems into very simple solutions. I definitely have troubles with doing this. Many of my solutions are complex. With quite a lot of effort, I am able to simplify the solution bit by bit, but when I talk to one of these people, their solutions blow me away. I always walk away saying to myself, "Why didn't I think of that? It's so simple!". I hope it's practice, and if I get enough, I'll someday be able to do what they do.

Depends on the discipline though. Hacking will never be communicated accurately to non-hackers in the same way that jazz will never sound right to the non-musician.

I do agree that good communicators manage to somehow frame their thoughts in a way that encourages other people to think a little further, and realise some truth that was staring them in the face.

After all, peotry is about making new things familar and familar things new.

It's unfortunate that great communication is largely unscalable (and hence the reason why YC exists).

This is actually my own test of whether or not I've really grokked something:

If I can't explain it to someone such that she understands it as well, I haven't really understood that something.

Smart writing won’t be in formal and difficult-to-understand journal articles, but in the profanity-laced angry rants you’ll find on someone’s blog. That’s where the smart people are, even if everybody else just thinks they’re dumb.

The trouble with this principle is, by Sturgeon's law, you are forced to sift through the 90+% of profanity-laced angry rants that aren't worth anything to find the few that are actually worthwhile.

The good point I think the article makes before generalizing about smart people's qualities is the fact that in many cases, such as argument, there are people who will toss some concepts, jargon or ideas without fully understanding them and because of that can look to an outsider (maybe your boss) like they won the argument.

Of course, if the other person in the conversation is really into that concept, jargon or idea he can quickly point out its flaws. But what is most likely to happen is for the other guy to have to make some effort to remember what that concept was and having to put the pieces together. When he can actually identify what was wrong in the argument, the battle will be long gone and lost.

Meanwhile the other guy may have made his opinion prevail and gain something because of it.

Posted by me in his comments section:

"The feeling of understanding can, however, be deceptive. There are lot’s of high IQ, well read and honest (as in not consciously trying to mislead) people who disagree about solutions to problems they both feel they understand and have easily explained (tough often differing) mental models of. The issue of vouchers in the first comment is a good example. The “brighter” person does not seem to realize that the greatest benefit of vouchers (according to some, including me) is that they will give an incentive to innovation. This is something that cannot be adequately tested for in small controlled experiments. Your heuristic is great for detecting nonsense but it can’t help you find truth."

PS. I'm not discussing vouchers today.

Honest question: how many people here actually think they are dumber than the average person?

I think I'm dumber than a randomly selected sentient being (from the entirety of space and time).

What probability would you assign to that statement being true?

I am dumb. (Not even relatively to average person, even in absolute measure)

On average, no. In random scenarios, frequently.

What do you mean by "dumber"?

If we by dumb mean having a low IQ (and/or other correlated qualities) I think very few hacker news readers are under average.

I agree, mostly. Being smart and being good at explaining aren't the same thing, though. They may often go together, but it isn't necessarily the case.

People who are good at explaining are almost always smart, but a lot of smart people have trouble explaining things.

One common problem is smart people assuming a certain level of background knowledge, and therefore referencing concepts or using terminology that's not familiar to their audience. It takes some experience to really figure out what vocabulary is appropriate for a particular audience.

I've worked with a lot of fourth graders (10-11 years old). I've seen a lot of people try to explain concepts to them and fail, sometimes because they're using sixth grade vocabulary the kids can't follow, and sometimes because they're using first grade vocabulary and overexplaining things and the kids get bored. It takes practice to explain things at exactly the right level for the audience -- it takes a teacher who is both smart (having mastery of the material) and experienced (knowing what level of complexity is audience-appropriate.)

I remember, I used to be penalized in A'levels if I didn't give examples and analogies when I was explaining a concept. Actually the lecturer never gave any points for technical explanation. But this sort of learning came in handy years later when trying to explain the research I'm doing to my boss and people around me. I just think it is a skill that can be learned.

I came here to post "the exact same" comment.

The assumption of smart people being good at communicating the explanation very well, is not necessarily true. In my personal experience, two of the most smart people, I have come across are basically not the best in communicating. They are so sharp, that thing obvious to them, are not obvious to most of the people.

However, for most people, I find the observation to be correct.

"For every complex problem, there is an answer that is clear, simple--and wrong."

There are two ways to seem smart:

1) Over-complicate everything.

2) Over-simplify everything.

Aaron's essay describes the former. It, and this comment, are examples of the latter.

If somebody convinced my that the world isn't complicated, it would trigger a couple of warning flags of their own. Most likely, they fooled me.

> Smart writing [...] in the profanity-laced angry rants you’ll find on someone’s blog.

I'm thinking of Zed Shaw.

One thing I have noticed is people who do not seem smart to me are always quick to remind you of how smarter than other people they are.

You are smart if: you realise that formulating smartness is not smart.

Bugger.

Reply on news.ycombinator.com