[gmd63]

Much of the restricted depth of this nature is a consequence of deliberate obfuscation or neglect by the people who are involved in the doing, such as trade secrets, or simply choosing not to write about or share the actual things that impact their craft.

You can easily accumulate depth of knowledge from reading in areas that are well documented.

[basil-rash]

How can you claim your knowledge of the field is actually deep, if you’ve never even done anything with it? Knowledge without application is nothing, imo.

True “depth” comes from the know-how acquired in digging through all the minutia nobody else though to document, on your way to producing a new creation nobody else thought to build.

[DavidPiper]

While I think there's some truth to yours and some of the sibling comments here, I think a lot of the discourse around documentation is actually just motivated reasoning.

The trope that software projects are poorly documented is so pervasive as to be saying nothing at this point. Despite extremely useful and celebrated documentation projects (MDN and Python off the top of my head) that dramatically accelerate other peoples' learning and productivity.

But when the topic is raised, the response is always "But we don't have time for documentation".

And when that time is given it becomes "It will be useless or out of date soon anyway, just go read the code" (As if there is nothing between non-practical high-level docs and per-line code comments.)

And when forced it becomes "Fine, but it won't be any good" or "Other people won't read/benefit from it anyway".

And when it's done, often as little effort as possible is put into it.

All of which smell to me like cover stories for: "I just don't want to do it."

[ijustlovemath]

In the case of Python, I do think the approachability of the documentation slipped after they migrated to the 3.x site, and I think this is somewhat reflected in the search engine rankings.

[bdjsiqoocwk]

> Knowledge without application is nothing, imo.

Supremely narrow minded. Knowledge has many facets.

[basil-rash]

Claims without support are nothing too.

In other words: example?

[bdjsiqoocwk]

Ok here's an example.

Henry Heimlich used "the heimlich maneuver" for the first time when he was 96 (in 2016), 40 years after inventing it, in the "senior home" where he was staying. But according to you, having invented was nothing in the first 40 years, because he didn't actually use it, never mind that it was used by others and taught as a life saving technique.

You're just beathtakingly ignorant of how progress actually happens. It's like your understanding of the world comes exclusively from thinking about the world in abstract hypotheticals, rather than interacting with it, which is ironic coming from someone stating that only action matters.

[basil-rash]

You’re so caught up in insulting me, you never stopped to think of I’d consider inventing a novel medical technique an “application” of knowledge. I would.

If all he did was read books about things other people did, he certainly would not have invented that.

Your following attempt at armchair psychology is humorous, but not much beyond that I’m afraid. Don’t quit your dayjob.

[bdjsiqoocwk]

My dayjob is as psychologist. No joke.

And my specialty is spotting morons on social media.

[bryanrasmussen]

>You can easily accumulate depth of knowledge from reading in areas that are well documented.

Given that specific knowledge relevant to a field may be of a highly specific and hard to understand nature it increases the risk that any attempt to understand that knowledge by simply reading it will fail due to simple misunderstanding of what one reads.

In documenting things there are always points in which documenting minute details of a thing starts to detract from the purpose of documentation, that is to say the more in depth and detailed one documents the less readable the documentation becomes, therefore one leaves out things that should be easily understood by others when trying to use the documentation to actually work in the field or will quickly be imparted by other practitioners in the field if it is one with easy access to others.

Documentation by its nature is aimed at everyone, but there may be particular things that would be obvious to many people but not some specific person, and that specific person when reading the best documented guides to the area of knowledge will still not be as knowledgeable as they believe, because everybody is different.

Very many areas of knowledge have specific relation to things that people do with their bodies, martial arts, sex, cooking, etc. etc. In such cases there is of course muscle memory, thus no matter how precise and painstaking the documentation will be in these areas you will not be as knowledgeable as one that builds up muscle memory in the field by doing if you rely on only reading the documentation.

I could go on, but given my point about minute details it might be self-defeating.

[callalex]

> or simply choosing not to write about or share

The default action for an autonomous entity or system is to do nothing. It’s really odd to phrase doing nothing as a choice. Anything other than doing nothing is a choice. Especially when it comes to work, most humans just don’t want to bother spending time thinking about work outside of paid work time.

[burnished]

You can also choose not to do something.

[Der_Einzige]

This x1000!

I'm so tired of (mostly boomers) talking about learning how to work on cars like it's easy. Sorry guys, but the companies who make those DIY repair manuals that you guys keep talking about using to learn wrenching don't seem to make stuff for any modern vehicles. Don't believe me? Go check out how pathetic their selection is right now: https://haynes.com/en-us/

It is nigh impossible for a non-car person to learn how to wrench without direct literal hand-holding from those who do.

It's also lead to the mechanic industry being FAMOUS for scamming grandmas, mothers, US soldiers, and other captive audiences. I think the only groups with a worse reputation are lawyers and car salespeople.

[qup]

Luckily, "those who do" now have youtube channels with videos for almost any car job you need to do, with the specific make and model.

The boomer way wasn't that easy, but it actually finally IS easy to learn to work on cars.

There's also this, I can recommend it: https://www.howacarworks.com/

[Der_Einzige]

Nope, most of the stuff I want to do on my 2017 RX do NOT have videos on YouTube. Good luck changing the spark plugs yourself including with YouTube. The back 3 will be impossible

[qup]

Oh you're right then, I withdraw my point. It's impossible to learn to work on cars.

[Der_Einzige]

It's impossible to learn to work on modern cars without help, and the idea that things are similar enough from the old cars that it carries over to modern cars is simply not true for a whole lot of vehicles. You really do need videos for each and every vehicle.

Watch even experienced mechanics works on new cars. They tend to bumble around for awhile.

[rnewme]

Not sure why you got down voted but I agree.

[MichaelZuo]

There isn’t a need to read anything beyond introductory materials to acquire depth, if your competent enough.

In fact that’s what I would consider the critical dividing line between a regular genius and a bonafide super-genius. Someone who almost supernaturally acquires expertise/intuition/depth/etc. with very little visible effort.

[neilk]

In my experience, this is more of a fictional trope than a reality.

There aren't any Tony Starks who become experts in thermonuclear astrophysics overnight.

Richard Feynman was as close to this trope as you can get in reality, but insisted that his reputation for being able to solve difficult problems was due to having a "different box of tools" than others. And he obtained that by studying rather obsessively, well beyond assigned texts.

I have been lucky enough to know a few people who also might qualify as geniuses able to produce miraculous results. One of them decorated his laptop with the logos of defunct computer companies of the 1950s-1980s. He drew a lot of inspiration from papers and books that few others have read in thirty years.

[ioblomov]

I’d assert that, rather than being fictional, it’s just exceedingly rare. While Feynman, Einstein, von Neumann, et. al. are indisputably geniuses, they all got relevant graduate degrees before doing their best work. The self-taught Indian mathematician Srinivasa Ramanujan, however, was invited to Cambridge University on the merits of notebooks he developed in isolation after reading a few mathematical texts…

https://en.wikipedia.org/wiki/Srinivasa_Ramanujan

[neilk]

The original claim was not that geniuses and savants exist. They clearly do!

The original claim was that a genius should be able to easily and quickly attain mastery of any topic, given a grounding in the basics. Like, you give Mr. Super Genius IQ a book on quantum chemistry and the next day they’re putting things together so cleverly they are on the precipice of new discoveries.

Feynman actually did this a couple of times! But I think that just shows the “unreasonable usefulness” of mathematics, as well as his own spirit of curiosity. (And at least one time that we know of, he faked it by reading a colleague’s research before he presented it. Not to take credit, but just as a prank, so he could interrupt and predict every conclusion before the presenter finished.)

[Der_Einzige]

Ramanujan is particularly interesting because he repeatedly claimed that divinity gave him his results.

I'm an atheist but when someone like him tells me that god gave him all of his equations - I'm not going to respond with "well ackshully". I'm going to ask them how we can make it easier for them to listen to god.

[MichaelZuo]

How would you confirm either way?

Assuming you don’t have a direct line to divinity yourself, you would need to have that profound genius in the first place to evaluate…

[rnewme]

What does being competent enough mean?

Do you have an example of bonafide super-genius?

[aerophilic]

Story from a friend who once was offered the first PhD in Computer Science.

He once mentored a guy who was so ridiculously good at writing programs, that he wrote out (once) a 5 foot stack of punch cards (back when that was how you would code), to create a program that was needed by the company for some purpose.

It worked, flawlessly, the FIRST time.

This was his MO.

Once, he wrote a program for an internal client, and it got shipped. They ran the program and ran into a problem. My friend told him to debug it for them… he replied “I don’t know how”. Up to that point in his life, He had never had to debug a single piece of software…

My friend helped him debug the program: Turns out, the problem was not an issue with his code, the customer had given him the wrong spec for a critical interface… and that was the only reason it had not worked the first time.

There are truly people that qualify. The only note I will make is that generally if you are a super genius in one dimension, you likely have something you are absolutely terrible at in another. Hopefully it is in a dimension that either doesn’t matter, or you have enough complementary people around you to mitigate it.

[rnewme]

I don't buy it. Every supposed super genius I ever heard about in the end had a tactic to their work. Even the high iq people have to have an approach to harness the intelligence.

[caminante]

Great at writing programs, but doesn't know how to debug?

You're pulling our leg or in denial.

[aerophilic]

This was 30+ years ago… I don’t know the details, but trust the source. The source used to teach at Carnegie Mellon. The point my friend made to me was that this guy was so good he never had to debug a program after he was done.

FWIW, apparently back before Intel released a particular 4 bit processor, this guy made an emulator and compiler for the chip so they could start writing code in anticipation of its release.

Once again, not direct experience… but trusted source.

[caminante]

> this guy was so good he never had to debug a program after he was done.

Every CS department has these legendary tales that grow more hyperbolic over time (See "never" above.)

They are fun to tell!

[samatman]

Even introductory materials cover the difference between "your" and "you're" rather thoroughly.

[MichaelZuo]

So?

[doesnt_know]

What you're describing as a "super-genuis" is just someone that thinks they know it all after reading some introductionary material. They don't know what they don't know so they think they're "done".