Why bother to remember when you can just use Google? | Hacker News Reader

Why bother to remember when you can just use Google?

arstechnica.com

30 pointsshawndumas25 comments

Chris Crawford, the author of the brilliant (for its time) game "Balance of Power," wrote a book about the game and his design process. In it, he pointed out that the simple facts encodedc in the game--such as the GDP of Ghana or the number of troops in Mali--were unimportant, but the relationship between the facts, the cause and effect between actions countries take and the facts, those were the important things to understand.

Thsi comes up all the time when discussion programmer intervies. No, there is no point in asking if somone has memorized something they can find in Google. But their is a point in asking questions to elicit their understanding of the relationships, correlations, causes, and effects between the facts. Being able to Google the answer to a relationship is not the same thing as understanding a relationship directly.

The problem with "known" facts is they often extreamly context specific. There are actually a wide range of interview questions where I would much rather have someone say "I don't know" than say a correct answer without aproprate qualifications. Because even something as basic as how large is an int is can quickly lead to a bug when someone assumes the usual rules still apply.

> Because even something as basic as how large is an int is can quickly lead to a bug when someone assumes the usual rules still apply.

Would someone actually ask this and expect a numerical response? The only correct answer is sizeof(int), right?

The sizes of primitives in C are implementation defined, but the sizes of primitives in some other languages (e.g. Java) are defined by the specification.

That's true, I was purely thinking in C there.

In C, it still applies when you use the stdint.h sized types, such as uint32_t and uint64_t.

This seems like a case where recognition matters more than memorization. On occasion, it can come in handy to memorize that on most modern 32-bit systems, INT_MAX == 2147483647. More often, it can come in handy to see that number or something close to it pop out in an unexpected place and immediately think "huh, that's (close to) INT_MAX". The same thing applies to 4294967295, 1048576, 65536, and the ASCII codes of various common characters.

Wow! I'd completely forgotten about this game. Thanks for the reminder :-)

"The discovery of the alphabet will create forgetfulness in the learner’s souls, because they will not use their memories; they will trust to the external written characters and not remember of themselves. You give your disciples not truth but only the semblance of truth; they will be heroes of many things, and will have learned nothing; they will appear to be omniscient and will generally know nothing." -- Socrates, in Plato's "Phaedrus"

It seems that they tested the remembering on data that is uninteresting or/and useless to the subjects. It makes perfect sense not to remember such information and even more so when one knows it's googleable.

On the other hand (insert anecdata disclaimer here) I have found out that Google does not prevent me from remembering things like git options or obscure facts I had to dig up as ammunition in flamewars I participated in.

We remember what matters to us. For example, ideas, concepts, and feelings matter because we shape everything around these things. But, how much do facts matter in comparison? To remember facts, we have to memorize the information itself rather than the metadata (information about information).

It is all about leverage, we can leverage more information by storing this metadata in memory. Additionally, metadata is a lot smaller in information content than the information itself. (it takes up less space in our minds)

Because then you are an expert in information retrieval, not the subject matter. You have no intuitive grasp of the material. You have no way of building on top of the material. No way of innovating, no way of growing.

On the other hand, expertise in information retrieval is a very valuable skill that allows one to penetrate (if not cover) spheres of knowledge previously considered the domain of experts or connoisseurs.

I remember my introductory writing class in college where the professor emphasized how important it is for students to know how to use external sources, libraries, indexes, catalogues, etc. It's funny that nobody would ever question whether using libraries would negatively impact our brain. Why then, in 2011, people still ask this question about Google? Answer: gerontocracy is afraid of being replaced by a younger, smarter generation.

Of course diversity of sources is important, but then any smart search engine user has figure that out already.

I know Django pretty decently, but I haven't touched Rails in years. If someone asked me to write out how to do something basic in Rails, I would have no idea what words to write. But what I would understand is the process (what I want to achieve and what theoretical tools I would use to achieve it), what to search for, and what documentation to read, and I would have an answer pretty quickly given those resources.

I find this emphasis on simple information storage and retrieval way to prevalent in our schools. Courses entirely based on reading and regurgitating facts made up a sizable amount of my education, including college. After a while it becomes hard to care, since the random facts are worthless to remember, sometimes even when going on to the next level course.

Transactional memory. Why bother memorizing something if you know how to access that information anytime when needed? I seem to recall Sherlock Holmes saying something about how he tried to forget everything that was unnecessary, so as to preserve his mental capacity for other things (ex. how when Homer learns something new he forgets something old).

It is far more important to have critical thinking, than good memory. Memory can be incorrect or incomplete and without critical thinking, can go unchallenged or unimproved.

The article and the comments here make some good points but my question is whether rote memorization of facts is beneficial in other ways, e.g. it may strengthen critical thinking unrelated to the memorized facts. It's intuitive to think that remembering a list of countries, their GDPs and mortality rates would help you understand the correlation functions between those variables but it may also help your brain to form similar but not identical groupings and associations.

I seem to remember reading that some educators are reemphasizing rote memorization because they found other beneficial side effects. Unfortunately, I didn't memorize it enough to make my point here, hmmm, time to Google...

I'm not so worried about the present - I am 34 years old and expect the web to be around for the rest of my career. Consequently, if I can't recall facts today, who cares? From a professional point of view, I'm worth much more if I can think critically about facts than if I can remember them.

However, I am very worried about the future. If we don't practice information retrieval, our ability to retrieve information should theoretically dim. When I am in my 60s/70s, am I going to be at increased risk for Alzheimer's disease (or Alzheimer's -esque symptoms)?

How do you know what/how to ask? I'm learning all the time, wide range, and I often have to learn more, or ask people who know more, just to know enough to ask Google. What good is Google if it can't help with the furthest reaches of your knowledge, if you can't form questions in a way it (or any other search engine) can answer?

But you still need to remember the text you wanna search for :P

Not always. Sometimes Google has the 'Did you mean: x', so you don't even have to do that.

Then the kids are right, textbooks are just for losers. Losers of memory, which in the real world gets so many the "loser" tag.

An interesting experiment I'd like to see would be a cataloging effect of transactional memory. For instance, if our brains are adapting to the fact that we can look up information, could we possibly be improving the breadth of knowledge we are aware of.

As an example, say we have X objects. Of X objects, there are Y properties. There are also N folders. We have two classes of subjects participating in the experiment. Subject class A is tasked with memorizing the Y properties of X objects. Subject class B is tasked with memorizing which folder of N the Y properties of X object is located.

If for example, X = 100 and Y = 5 then N = 100. An example result for A subjects might be an average of 5 objects where each of the 5 properties were memorized correctly. [That's 5 properties x 5 objects = 25 memory units]. The question I have, is if B subjects would remember an average of 25 objects stored in which folders correctly, or might they remember less/more correctly? If the Y properties vary in type (from color, to weight, to appearance, etc) do they require more memory power than a consistent X is in N folder memorization?

If it turns out you can memorize more about where information can be found, than you can about the actual information then it would seem that [for so long as the information can be looked up] it is more practical to catalog that information exists and where it can be found, than to engage in memorizing the information itself.

Then the idea of "being a jack of trades, master of none" has a slightly different caveat. In that one could be more-so a jack of all trades than one could master a single trade. Ie, the sum of knowledge held by the jack is greater than that held by the master.

Yes, the abstract thinker does hold onto more actual knowledge then the concrete detail thinker. Breadth really does outway depth in terms of true thought, but in the mechanized specialty world of now limiting the brain to be concerned with only a certain set of repeatable, hard to perfect or continually repeat procedures is what is rewarded for most people. We don't pay you to think, ya know ;)

Reply on news.ycombinator.com