Hacker News Reader

Comment by loboman | Hacker News Reader

http://www.pangeaprogress.com/1/post/2010/09/einstein-edison... 'While in Boston, Einstein was subjected to a pop quiz known as the Edison test. (...) A reporter asked him a question from the test. "What is the speed of sound?" If anyone understood the propogation of sound waves, it was Einstein. But he admitted that he did not "carry such information in my mind since it is readily available in books." Then he made a larger point designed to disparage Edison's view of education. "The value of a college education is not the learning of many facts but the training of the mind to think," he said.'

View on HN

While context also has some part to play, in general I am with you in what you are trying to say.

BTW, are fictional characters counted as references? ;-)

http://en.wikipedia.org/wiki/Sherlock_Holmes#Knowledge_and_s...

From that article, In A Study in Scarlet, Holmes claims he does not know that the Earth revolves around the Sun, as such information is irrelevant to his work. Directly after having heard that fact from Watson, he says he will immediately try to forget it. He says he believes that the mind has a finite capacity for information storage, and so learning useless things would merely reduce his ability to learn useful things.

EDIT: Somewhat relevant (and OT) comic... http://theoatmeal.com/comics/tesla

obligatory debunking: http://www.forbes.com/sites/alexknapp/2012/05/18/nikola-tesl...

Debunking of the sorry attempt at debunking.

http://theoatmeal.com/blog/tesla_response

I know this is offtopic, but it's a great point... There's a paragraph that says "This is a fair point." next to it. It basically says that we all stand on the shoulders of giants - and that other inventions allow new inventions to be discovered/developed. This is the EXACT reason I disagree with most patents.

Apologies for the off-topic rant, but it's one of my passions. I'm most probably preaching to the converted.

If you're going to use Holmes as a reference, then you agree with Edison and disagree with Einstein.

Holmes was characterized as profoundly valuing facts (as well as methods of thinking), if and only if those facts were useful to crime-solving. He was noted, for example, for his monograph on tobacco-ash residues. That is, he knew so much about the details of tobacco ash that, upon sampling some ash found at a crime scene, he could infer many useful things from it. Another example is that he is so familiar with all the various mud around London that when he sees some dried, he can guess where it came from.

Holmes is, overall, preposterous. And his theories (the theories that Sir ACD put in his mouth) are almost entirely without evidence.

That said, theories without facts to work upon are like a level with no fulcrum.

I think there are two different kinds of facts:

1. Random facts that have no obvious connection to what you're working on. 2. Facts that come in use frequently, and form the basis of expertise.

I expect there were many facts that Einstein did have memorized, even though he could look them up.

Holmes is surely an exaggeration, but in many domains expertise is simply mastery of basic elements.

Understanding many little pieces lets you see the whole more clearly than someone who is missing parts. Experts have enhanced intuition.

This is a distinct type of knowledge from the depth of the pacific ocean. That kind of knowledge will not improve intuition if memorized.

Right.

I've said this before, and I'll say it again. Do not take advice from Sherlock Homes, because he is fictional and thus can be no smarter than his creator, a man who quite literally believed in fairies, and could not be persuaded that Houdini was not actually magic.

While I agree with your point here, I'd like to question the statement that a creation can be no smarter than its creator.

This may well be true if we're talking about real-time creation, but it's absolutely possible to put in more brain cycles into offline, pre-processed creation and get better results.

Right. There are at least two different types of 'smarter'. Thinking speed is one dimension on which a mind can scale, but not the only one. A moderately competent chess player may play at grand-master level if given centuries of dedicated time to think about each move, but no amount of time-per-move will make a chimp a good chess player.

Thomas Edison invented the cylindrical Phonograph, and movies were beginning to just show up in 1921. I'd imagine knowing the speed of sound - especially how it relates to syncing with video - was of major importance to him, much more than it was to Einstein.

Context is everything.

Regardless of the context, it's still a fact that that information is readily available in books.

Why is that so many people on this forum feel that the information pointed to by this link is of such great importance: https://gist.github.com/jboner/2841832

I don't know if that information is contained in a particular book, but it is certainly readily available all over the web.

Why do we care? Could Edison have had a similar interest, though focused more on his own business?

Separately, what good is knowledge of theory if you have no practical experience upon which to understand it? The candidates given this test were also supposed to be in line for the executive ranks. Having a general knowledge of geography, geopolitics and trade would seem to be fairly important.

It's also unclear from the article exactly how Mr. Edison interpreted or used the results. Frankly, it sounds like many of those quoted in the article were offended by being asked such "simple questions"... despite not being able to answer them. Perhaps Mr. Edison was interested in finding executives who didn't believe that common sense, mundane facts or important business details were beneath them and best left to others.

Those numbers are of practical importance to programmers in particular. They directly affect nearly everything we do to some degree. On the other hand, without knowing anything about you, I can pretty confidently guess that the identity of Leonidas and the process of tanning leather are not practical concerns for you at all.

Also, I would argue that memorizing those numbers themselves is unimportant and putting too much stock in them is a mistake (some are very likely a bit different on the computer you're working with than they were on the one used to compile the list — for example, hitting L2 cache is slightly faster than a branch misprediction on the i7 IIRC). The important takeaway IMO is the orders of magnitude at work.

If you were developing technical products in 1921, the process of tanning leather was probably very relevant, as were questions about the density of different kinds of wood, the sources and production of various metals, etc.

Just as a programmer should know the orders of magnitude of latencies, an industrial engineer in Edison's time should have a sense of the characteristics of materials he might be choosing between. For example, should the handle of a new tool you're inventing be made of leather-wrapped steel, or hickory (traditionally used for ax handles)?

The trick is not that the information is available but knowing where that information is, so that if needed, one could look it up.

I tend to want to remember the places (in this case, blogs, books, articles,etc.) where I could start looking to find a piece of information as more important.

A simple example of this is me wanting to learn Backbone.js. I have been following Addy Osmani's blog and know that he can be considered an authority on Javascript.

I am now following his Developing Backbone.js Applications book online - http://addyosmani.github.com/backbone-fundamentals/#mvc-appl...

I trust that I have picked one of the top people in this area to help me learn this. If I just Google, just the sheer amount of links to blogs, training sites, videos, etc. would be overwhelming. It would be likely that I would pick a resource that is not the best.

With the information overload that we have, you now stand out by being able to find (and apply) the information in a timely manner.

It's somewhat of a trick question, as the speed of sound can vary greatly depending on the material it's propagated through. In air, it varies according to temperature and altitude.

Despite the space station being at an altitude of 370km, I'm sure sound travels through its air at about the same speed as it would in the open at sea level. So it's not altitude but density.

Is it important in this context to know exactly what the speed of sound is in feet per second without looking it up, or is it more important to be generally aware that sound travels through air via compressed air waves and that it is quite a bit slower than light, the details of which can be looked up when necessary? Are these executives going to be tasked with making lightning-quick product decisions involving physics calculations on a daily basis?

I doubt Edison had a full understanding of the theory of operation for the phonograph he invented. I can't see how knowing the speed of sound in air could matter at all in relation to synchronizing audio with video, or Edison's motion picture camera. He could not broadcast sound from a speaker far enough for the sound to become unsynchronized with any video that was played.

This is undeniably true, but I think misses an important correlation: in my experience, people who are best able to find and access "minor facts" like this when needed are precisely those who are commonly considered "walking encyclopedias". I strongly suspect this was true of Einstein too, even if he couldn't pull up that one figure.

I think penalizing someone for not knowing any fact in particular is silly. But throwing a ton of questions like this at a candidate just to see how many they hit has, IMHO, more value than is commonly admitted.

This view is even more correct now that "what is the speed of sound" is only a google search away.

Maybe Edison only asked these questions because of how long it can take to look something up in a book.

For example, would you hire a programmer who couldn't answer:

- name one or two html elements and what they are for

- what is a for loop, and when would you use one?

These can of course also be looked up in books, but already knowing the answer to those questions (and many more advanced questions) in a lot of what you're paying for when you hire a programmer.

Well, let's compare "what is the speed of sound" and "what is a for loop". Both are domain-specific questions, one for physics, the other for programming. For both I'd accept an answer that describes the notion or idea behind the concept and not a memorised instance of it.

So, for "what is the speed of sound", I would accept an answer such as "The speed of sound is the speed at which a wave propagates through a given medium"; though I expect an actual physicist to involve molecules, springs and so on in his answer. What I would not accept is a string of digits. A string of digits shows you know how to remember a string of digits.

Similarly, for "what is a for loop", an acceptable answer is one such as "A for loop is a construct for bounded (at least in principle, but you can have unbounded for loops in some languages) iteration over a series of elements, either generated on-the-fly or from a concrete container". The analogue to a string of digits for this question would be to give the BNF definition of a for loop in C. I think you'll agree that knowing C syntax doesn't show you know how to program.

Given two candidates (for a C programming position), I would of course prefer those who know the syntax. Both because that's one less thing they need to learn before they can become productive, and because knowing C's syntax is indicative of spending a lot of time programming in C.

So to return to your comparison, if I were hiring for a position that required working math where the speed of sound was required... I would treat knowing that constant by heart indicative of their knowing the rest of that subject matter by heart. I'd want to test a lot of that as well. "What is the speed of sound?" would be like a fizzbuzz.

But any reasonable person should be able to answer the speed of sound question. After all, the rule of thumb for estimating lightning distance is that every 5 seconds is 1 mile. Reciting a definition is no better than reciting a string of digits.

Are you saying that people who do not know that rule of thumb are not reasonable people?

I obviously am.

Chalk me up as unreasonable then, as I had never bothered to remember that rule.

That rule of thumb only works in a very small number of countries ;)

> The speed of sound is the speed at which a wave propagates through a given medium

That's really the only 100% correct answer as well, since the speed of sound varies by temperature, humidity and a lesser extent, atmospheric pressure. If Edison expected a discrete number, that would be kind of disappointing.

Sure I would. Some time back html did not exist, and knowing that someone decided <blink> was a good idea doesn't matter. And for loops only exist in some programming languages and are not that fundamental even if common.

I wouldn't even think to ask those questions in an interview.

Reminds me of a similar story with Henry Ford, where he said he doesn't need to know everything since with the press of a button on his desk he can summon someone who can relay the information to him.

Henry Ford was notoriously ill-informed. He sued a newspaper for libel after they said he was an idiot. In the trial they asked him in what year the revolution that founded the country took place. "1812?" facepalm

edit: unfortunate typo

sued a newspaper for liberal? he really was an idiot.

And yet Ford was a fantastic innovator and worth $188 billion (adjusted for inflation) when he died! It really hammers in the point that general knowledge and intelligence are very different skillsets.

The Ford story has a superficial similarity to the Einstein one, but in reality they're diametrically opposed. Einstein _understood_ how the physical world worked so he didn't need to carry around unnecessary "information". Ford, at least in the story, has neither "information" nor understanding.

The distinction is similar to the distinction made in philosophy between "knowledge" and mere "true belief".

> Ford, at least in the story, has neither "information" nor understanding.

He does have information, he has information that some people around him will respond to his requests for more information if it becomes necessary to have it. The power of delegation at work.

Of course Einstein is famous for not knowing his own phone number in Princeton. He didn't see the point in remembering things he could look up.

I have to nitpick there. I suspect Einstein is actually famous for his Special and General Theories of Relativity, and for his amazing hair.

I completely agree with this, especially for us living now, since we have access to virtually any information on the internet. However, I can see why Edison would ask seemingly pointless questions. First, while that information is readily available in books, going to a library to find the book with the information needed was much more time consuming than pulling up Google, so there is the convenience factor of memorizing random tidbits of information. Second, and perhaps more important, is the fact that Edison was an inventor, hiring people to help invent things. As an inventor, one needs to have a vast array of knowledge in many fields, and then to be able to make connections between seemingly unrelated items/topics. In our case (hacker news readers), we seek to combine separate entities in ways that no one has done before, creating a new product (read: app/website). Making connections no one else sees is at the heart of entrepreneurship, and invention as well, so having that wide array of knowledge could be see as invaluable.

When I was in the Army the officer in charge of my section commented that another PFC was better because he read the software manuals so he could answer questions that came over the phone.

I said I memorize things that I can't look up easily.

In the original link, I didn't see any claim that Edison was looking for correct answers, or that Edison knew the answers. Only that he looked at the paper for a bit and then dismissed the candidate.

For all we know, he was looking for people who deliberately ignored some subset of questions as irrelevant, or who demonstrated some quality of judgment such as questioning the appropriateness of the quiz verbally before starting, asking to take it in a quieter environment away from the arguing people, helping mediate the argument instead, stopping and asking for their first few answers to be checked to see if there was any point in them continuing, or ... anything at all, really.

How I read this: Einstein would have approved of candidates using Google.

Reply on news.ycombinator.com