Expertise in the Age of YouTube(behavioralscientist.org) |
Expertise in the Age of YouTube(behavioralscientist.org) |
The majority of good polished videos have the least accurate or useful information. These I've surmised are people that would rather be into video production or youtube stardom than they would being actually good at the subject they profess. Alternatively, some of the worst grainy, shaky cam videos I've seen, usually on the jobsite, are some of the best practices, tricks of the trade or just informational rich videos you can get. These are people that are not into being youtube stars and are just part of the 'sharing community' as it maybe was originally envisioned. But as Rumsfeld said, there's known knowns, known unknowns and unknown unknowns, and when you're a homeowner watching how-to videos, its very difficult to know what you should know.... so most of them go by, instinctively, the quality of video production as their guide to what they should be doing.
Sure there are great craftsmen with great video quality and there are some really poor craftsmen with poor video quality but these are not the bulk of what's out there. What rises to the top of search offerings seems to be equally what's been marketed (poor craftsmen, high quality, "please like,share, and subscribe") or what's been organically liked (poor quality, high craftsmanship).
For the tradesmen out there dealing with homeowners, its now sometimes a bit of an uphill battle after homeowners have done their 'youtube' research and have determined how hard the job they've asked you to bid is going to be.
I explain it to myself in a similar way: people with nice, clean, heavy websites, sprinkling ads in between the text, are content marketers. Their job isn't to educate you. It's to regurgitate whatever bullshit they can find on topic[0] in order to create SEO and social-share optimized vehicles they can then attach advertising or affiliate payload to.
--
[0] - I worked alongside content marketers few jobs back. There was zero verification or care about accuracy of what was posted, and a lot of copy-pasting+rewriting of similar content from competition.
Quality significantly diminishes the moment you use your face in the video (because it's completely unnecessary when explaining how to do some series testing in calc for example)
Basic carpentry. I've built several pieces of furniture around my home including a pretty nice poker table (so, upholstery too). I built a 12'x6' shed in my back yard on a foundation that's not going to sink or slide over. Drywall and ceiling repair. Re-did a bunch of baseboard trim and only me and someone who knows what they're doing can find the mistakes. I know my limit though. I'm probably not going to attempt kitchen cabinetry or replace my roof for example.
Basic plumbing. Gutted and replaced bathrooms and kitchens on multiple houses. Repaired a busted spa back to like new. Re-did sprinkler systems. Cautiously working my way up to gas lines--don't want to blow up my home.
Basic car maintenance and repair. Everything from oil changes, brake pads up to replacing intake/exhaust systems, replacing a power steering system, light bodywork.
Metal and fiberglass work. I'm building a single engine kit plane in my garage. Online resources have been vital, and there's no way I'd have gotten anywhere on this project without them.
You need to have a good filter. You're right that the highly polished "Brought To You By Tool Company!" productions with the generic 80s porn music soundtracks are pretty bad, and the "Hey Guys! Like And Subscribe" bros are always the worst. Find the videos without background music and with bad lighting and sound quality where the person is actually grunting and doing the work in real time along with you. They're usually the best.
Easier'n most plumbing, I'd say. Run gas to stoves on three houses now, hardest part's piecing it together without needing three runs to the hardware store. Doesn't make me half as nervous as electrical. Or piss me off as much. Damn tiny spaces and too-stiff or too-bendy wires.
"on the jobsite" is probably the most important point, because then you get to see the actual process as it happens, instead of someone talking about it and a series of jump-cuts of them barely doing anything.
I feel like the quality-vs-accuracy thing happens with websites too; the highly-SEO'd flashy sites also tend to be the most devoid of actual detailed information, and personal/smaller sites have the most interesting detail, but unfortunately search engines almost always pick up the former.
Look at the news. The news with more spectacle tends to be of lower quality on average. Or anything that comes from Hollywood. Daytime and reality TV are extremely polished but mostly tripe from either an entertainment or informational perspective.
The more time someone puts into a "production quality" video, the more they're trying to sell you something, even if it's just their brand awareness.
Fancy graphics is by no means an indicator of good information, but neither are lectures.
It blows my mind that I can be taught programming concepts, say data structures, on a PDF. That is insanity, it is a total divorce from the idea of computer science to begin with, to hide data and information in a non-executable environment. Someone has the quote that "pdfs are where data goes to die". Give me the data structure, let me ask questions of it, hold it in my hands and inspect it at the pace of my own thoughts. This is not what material provided in "lecture" through "slide decks" provide.
The information density of lectures is also low, why are we anchored to the low bandwidth channel of voice (I could say the same about text writing or a textbook)? Give me the information, let me have it, then let's talk about it in organized discussion.
Agreed on everything else :)
If we take an example from programming:
A fundamental Python or JavaScript course would have massive audiences. As a result, some of the video author would made money from their videos and then keep polishing the videos.
In contrast, a in depth video guide on how to prove theorems in Coq would be very lucky if they have 10 audience...
Not like the Olympic games or video games, where professionals' performance are pleasant to behold. The problem of 'How to' video is it encourages shallow information because the real professionals' work are either too arcane or too boring for the audience.
Whatever information has been printed in ten million textbooks and distributed across ten thousand public schools by a legion of educators is at a stretch mostly accurate.
Our actual best understanding of the universe is whatever was just mumbled in a graduate seminar by a half-asleep emeritus faculty member who doesn’t really mind that nobody heard it, because they’re bright kids and one of them will work it out if they think about it.
Also, in general, the very best books can only be written by a subject expert who can coherently hold enough of the subject in his/her head to find out how to express it. It’s rare but exists (tannenbaums operating systems for example?)
Recently I have been watching a lot of blender tutorials and some people are much better than others at getting to the point and describing important details about what they are doing without losing the flow.
If the first30 seconds of the video consist of the person clicking seemingly randomly around in their UI while talking about what they are going to be talking about then I usually just stop watching and move onto something else.
This is a quote to remember. You nailed it.
That's the concept behind the "accountability partner": someone who is not an expert into what you're trying to achieve, but who bugs you regularly to make sure you're moving forward.
After searching for "accountability partner as a service" I couldn't find anything other than weird marketing websites that looked more than a little scammy, so I built a simple tool that sends me an email everyday for each task defined in the system, and that lets me record progress by responding to those emails.
The concept is akin to Jerry Seinfeld's productivity hack: "don't break the chain", meaning you need to do something about your craft every day. If you skip a day, the chain breaks and restarts at zero.
The effectiveness of this is surprising; the accountability system can't assess the quality of the work in any way and isn't even a person that I could disappoint, but having to respond to an automated email every day to log my progress and store ideas, etc., pushes me in ways I would never have expected.
The downside is that this means keeping a steady job is becoming increasingly more difficult. The moat that protected the job security of highly advanced professionals is rapidly deteriorating due to this democratization of information. Credentials mean less today than they ever have in history. Specialists will have to be on top of the ever-expanding spigot of new information in order to remain competitive and justify their fees in a time when anybody can just search Google/Youtube for the answer.
None of those would be problems if workers weren't financially dependent on their jobs, such as if we had a universal basic income. But since most of us our financially dependent on our jobs, we're basically in an arms race against each other to the bottom. Stay at the bleeding edge, or get replaced by a more financially desperate harder working 20 year old from an Asian country who learned everything you know from Youtube / the internet. Many will instinctively deny this now, but just watch how this continues to play out over the next 10-20 years.
For instance, my go to example are video game mods and ROM hacks. In the olden days, these communities grew quite slowly, and it took a fair amount of time for people to get to understand the workings of a new game's engine and how it could be built upon.
And you can see this if you look at Super Mario 64. In it early days, mods were really basic, and even the fanciest ones looked super primitive by today's standards:
https://www.youtube.com/watch?v=UgLc0rPRs3U
It then took years to go from stuff like that to more advanced original game mods like Super Mario 64 Star Road, and even longer before ASM coding became something many people understood and used for custom enemies, bosses, gimmicks, etc (like in Kaze Emanuar's current works).
Now contrast to Super Mario Odyssey. It took just months for mods to get made for that game, and under a year for fancy looking ones like this:
https://www.youtube.com/watch?v=ss7unbD4yHg
And even that's a slower example. Since then, we've gone from seeing modding scenes take years to develop, to seeing them a week or two after release, to in some cases seeing mods getting made BEFORE the game officially hit store shelves altogether.
Seriously, Smash Bros Ultimate got mods before the game was available to buy, all based on leaked copies posted online.
Outside of games, same thing happens in web development too. New framework pops up, professional looking sites/apps/whatever get made in the first week. It's gotten to the point that those jokes about looking for 'rockstar ninjas in [new programming language here]' may not be jokes anymore.
But yeah, the quality baseline goes up significantly quicker now.
Despite the problems platforms like YouTube cause (or enable), I certainly wouldn't want to do without it.
What is your preferred method of learning new tools, techniques, etc?
Roll forward a couple of more years and I'm working on my Ph.d. and have the thought that someone must have done more optimization on keyboard layouts: Meet http://mkweb.bcgsc.ca/carpalx. That inspired me to learn QFMLWY, something I have yet to do.
Roll forward to a few months ago and I'm experiencing some slight wrist pain; a coworker uses a Kinesis Advantage and recommends it. In true punk style (and unhappy with the Kinesis' outdated appearence), I endeavor to find a superior product. I stumble upon the Ergodox EZ, a fully-featured workhorse. I quickly become frustrated at my reduced typing speed and seek out a better solution. Cue Plover and the OpenStenography project. Learning stenographic typing is my next life goal.
This is a niche path involving one skill set: Improving the human-keyboard interface. And, it involved real-world interaction with people and the resources that people on the internet have provided. I'm convinced that either only in-real-life (IRL) or only online interactions would have inhibited my growth in this domain.
I'm sure that this same concept applies to every web development framework or conlang or beauty style or any human endeavor. As an aside: These ideas are also related to why I'm confident that an artificial general intelligence developed in the vacuum of the internet will not succeed relative to a competing technology that is able to benefit from both offline, interactive experiences and also online information trawling.
It may just be different audiences/userbases, idk.
For edu channels it seems it’s a balance of teaching core concepts and also teaching what’s “popular” to keep folks interested.
You have to make a cutoff at some point with caniuse, as I suppose to get 100% coverage we’d still be doing table-based layouts or float hacks. The old adage of “know your audience” will eventually dictate the tech.
It's a valuable skill to have that will save you thousands of dollars over a lifetime. Give it a try next time you need work done.
It is a fun and interesting puzzle project putting it back together blind. It's not impossible. But it is tedious.
I would add an additional note that the choice of topics to dive into is now both more vast and accessible than it has ever been.
Not only can you learn what you want to, but you can stumble across that which you'd never thought of or scarcely knew existed and discover more about yourself along the way.
When you search for things to learn you're merely bolstering your predisposition, but when other things emerge before you and stick, something altogether more interesting is taking place.
Even if I'd think fablab/shop per city or neighborhood might be more efficient, it's still great.
[0] to me the only good thing there was khan academy which was basically video tutorial you can rewind.
It feels like a category error because it is a category error. Competitive games have a finite set of explicitly defined rules and win conditions; jobs do not.
> Even small changes along these lines would be a big deal. Imagine if each of us got as good at our jobs as the average teenager is at Fortnite.
I think the average person is already probably better at their job than the average teenager is at Fortnite; it's just that the "win conditions" for an average person at a job are significantly more complex and can vary quite a bit between different people and situations.
For example, besides whatever people themselves consider to be a "win" at their jobs, there's also things like this: https://dilbert.com/strip/2012-05-29
We are joking that people no longer are flying to the Moon because NASA have lost all of the videos of how to assemble a rocket. ;-)
Mexico Population: ~128 million
Fortnite Peak monthly players: 78.3 million
LoL peak montly players: 111 million
Fortnite total accounts: 250 million
you can say whatever you want if it's a blog article
I just want to de-normalize that a bit: blogs should strive for high standards, even though some don’t. They are a major form of information these days!
This is also certainly a contributing factor to the flatline of median incomes for 40 years in the US - among the dozen other influencers including globalism, automation, deregulation, deunionization, etc.
But we, as the labor class, are in a constant ongoing rat race to the bottom. The endgame is near total automation of production, near total availability of expertise on demand, and total capture of all resources by those own the robots. And this isn't like any other time - an able bodied man in any age before this one had intrinsic value to their community. They could be taken advantage of and thus provided for. We are going to see billions who cannot contribute meaningfully to economics. And I do not think those leading the charge towards automation are all that interested in seeing quarterlies dip to reduce suffering in the world.
But the internet completely destroyed information asymmetry, many people have access to lots of information now. I don’t need to take my car to a mechanic or pay someone to teach me, I can watch a YouTube video on how to change my oil.
The side effect of the destruction of information asymmetry is that a LOT of people depended on it, and they are all going to have to figure out where they stand in the new marketplace of information. I think we’re also seeing huge social movements because people are reading stuff online and reflecting about their city/state/government and saying “hold on, what you’re doing isn’t right, I’ve read about why you’re wrong”. The Arab spring, occupy wall st, the european refugee crisis, the Hong Kong protests, the Chilean protests, basically any major social movement in the last 20 years could be argued to have been triggered by the internet.
This is historically why reading comprehension has been such a big deal. This information has always been widely available, in a maintenance manual you could borrow from your local library.
Just as an example, I've decided I want to grow microgreens and tomatoes indoors. To do that efficiently, one needs grow lights. Thanks to the CBD community, there is sufficient information online to spec and assemble yourself grow lights. Whether you want to solder individual LEDs to a backing or glue COBs onto a heatsink, it's all there. Sure the initial investment is more, but product you can produce is staggeringly more powerful and the lifetime cost dramatically lower than what comes prepackaged.
Of course, then again this might not all happen if children are forced to be in school most of their waking hours.
Sometimes directly or indirectly (public repos/forums) working with each other. Even amateur groups produce stuff much more quickly these days. You can see it play out in college dorms when an enthusiastic group gets together.
So, for example, learning a new programming language - I'd say commit to completing the same low level assignments that you needed to do for your first language. Hell, if you still have them, redo them in the new language. As you move up in your understanding, then start to see what you can build with it. Or even simply "try" to implement it in the new language.
Much of that first hurdle comes from devoting the same time amounts that you did in the past. It's easier to do at a younger age due to less responsibilities. As we get older, we still need to devote the same amount of time, but it's harder to do with competing activities vying for our attention. I've liked the term "attention economy" [1]; however, instead of advertising, I view a similar thing for education.
My main problem is what to focus on. Not how to learn it. I guess reading a blog post instead of paying Amazon would be good for me. :)
Then once the thing arrives my attention has been directed somewhere else and I'm much less motivated to actually read or do anything with the thing, and it just goes on the pile.
At least for doing practical projects, I've found that poverty is much more conducive to getting things done. Having to make do with what's avilable, free or cheap, and not getting distracted by shopping for the perfect solution. Likewise, I find that some of the most creative people on YouTube tend to be either pretty young or extremely frugal.
On a related note, I just read nassim taleb's anti fragile which has a section arguing that most progress comes from people tinkering with things to get something they can sell or to solve some problem they have rather than academics trying to understand things from basic principles. I don't agree with it entirely but that and my way of learning has led me to value experiential learning much higher than society does currently
To piggyback on that, I need either a real project or genuine curiosity. I'm terrible at making myself learn something that I'm not 1) immediately applying to real work—paid or personal—not a make-work learning project, or 2) just really dog-chasing-a-bone interested, and I cannot gin that up or fake myself into that mode, though sometimes I can start poking around a topic and find something that gets me going.
Fortunately for me I'm fuckin' good at working that way. Unfortunately it made school a real pain in my ass.
I watch Youtube mainly to relax. I watch more Youtube than Netflix at this point. A lot of the content I enjoy appears to be teaching about stuff.
Some of the stuff I watch sticks. E.g. I'm baking a sourdough bread right now. In recent months that has become a Sunday morning routine for me. I learned how to bake bread by stumbling on some people explaining how to do this on Youtube. Since I enjoy eating bread, my mind went from "hmm tasty" to "hmm, I could do this". And then I had a lot of fun doing it. In the same way my cooking has improved dramatically since I figured out that implementing a recipe is vastly simpler than implementing an algorithm and infinitely more tasty. I have dozens of youtube cooks in my subscription feed. Most of what I see I never cook but some of it I do try out.
I just gravitate to that kind of channels and I subscribe a lot of youtubers that inform me on a wide range of topics on a daily basis. E.g. I find it very relaxing to watch Adam Savage (of Myth Busters) go about doing some stuff in his workshop to build all sorts of wacky stuff. But I'm hardly a maker myself. I rarely use what little tools I have (i.e. non programmer tools). But when I do, I find I know that I seem to know what I'm doing.
But is it more for professional growth I want to learn, I want to be hands on ("forward leaning"). Typically, write a program or a tool using a programming language I want to learn more. This takes more effort, and gives better results too.
I've been checking out steno, and I'm really still not sold on it as far as coding is concerned. I've mostly seen a bunch of bad demos.. (https://news.ycombinator.com/item?id=21490348)
I still think the keyboard as well as the software side of it could be improved quite a bit. I'd like to start at the keyboard, but it is surprisingly hard to get started. There are lots of custom keyboard kits out there but they all seem to be using the same standard size switches & caps that aren't what I need. So what, do I have to start by designing and making my own microswitches?! I wish I had a youtube video or three for that ;-)
I've been using the Dvorak layout since 2009. First on a TypeMatrix 2030 USB for a long time and currently with an ErgoDox EZ Shine that I bought like a year ago or something (wow, time flies! I thought it was only a year ago that I bought it but I looked it up in my e-mail archive and it was in fact December 2017 that I bought it) and have been using since then. Right now I am typing on the QWERTY keyboard of my MacBook Air, just because of where and how I am sitting at the moment. But I love my ErgoDox and use it a lot.
Here is my custom Dvorak based layout for ErgoDox EZ that I created basically recreating the position of most of the keys that I was used to with the TypeMatrix 2030 USB that I had, and with a couple of minor but important modifications.
https://configure.ergodox-ez.com/ergodox-ez/layouts/ENb6/lat...
This layout has been working perfectly for me and I am very happy with it. It's mostly looking the same now as when I first created it (that link is for the most recent version, the one I am using now).
Some things to note about this configuration and about ErgoDox EZ in general (for the benefit of others ITT):
- In the typical position of caps lock I have Escape when I tap it, but Control when I hold it. Awesome!
- This whole layout resides on the keyboard itself. I have all my computers set to US standard layout and am able to have my layout work always on all of my computers across all operating systems with no configuration or anything. This alone is worth like a couple of hundred bucks at least, seriously!
- The previous keyboard that I had, you know, the TypeMatrix, had hardware Dvorak support too, but I always had to press a button on it each time I unplugged and replugged it, or woke my computer from sleep, or turned my computer back on after having shut it down. It's actually surprising how annoying pressing that button each time where I needed to, which wasn't always (because back them I was remapping the keys to Dvorak in software on most of my personal computers), but often enough that it got annoying. None of that B.S. with the ErgoDox! :D
Prior to purchasing the ErgoDox EZ those two years ago, I too had looked a bit at stenography tools and projects. My conclusion back then was that it seemed useful, and was interesting, but cost prohibitive and time consuming to learn, with the added factor of uncertainty that I have no idea if writing code with stenography methods would actually be faster or in other ways better than my current means of writing software.
Anyway, if you do end up learning stenography and use it to write code I would love to hear about it and I hope you post about it to HN then.
> This whole layout resides on the keyboard itself.
This was a game-changer for me. No toggling between DV and US, anymore.
Regarding stenography for coding, I will probably only use stenographic typing for software documentation (prose). The main use for stenographic typing will be blog posts, emails, HN comments, etc.
With the ability for the Ergodox EZ to emulate a serial port , I can use Plover+Ergodox EZ as layer 1, say, and Dvorak as layer 0, without needing to change anything at the OS or application level when I do this (Dvorak will be seen as a USB HID which Plover won't see and steno-mode will be seen as a serial connection, which gets processed by Plover and forwarded to the OS).
Thanks =) ... I'm always disparaged with HN when you say something that can be viewed as not supportive. (even -1 votes feels like "oh come on guys!")
I think the thing that scared me the most out of that video was using things that are approx. 90% globally supported for layout.
Totally agree with "know your audience". Specifically, I teach the habit of look it up on MDN, then caniuse, then have VMs/devices ready-to-go and test it even further.
Simple rendering/functionality issues have turned into legal action against my past employers. Ex: if you cause even a 5-10% drop in ecommerce conversions due to what shoulda been a simple CSS fix, and that goes unnoticed for a month, that's a potential lawsuit/termination with very "black and white provable damages". Unfortunately, I've lived this =(
It's worth noting that I work on conversion-based web solutions that are "mom and pops" small, all the way up to "if it's down for 1 minute we just lost thousands of dollars"... but really I've learned to be incredibly careful with both. Every solution I build online at some point means $ even if it's non-direct conversions (ie: purely informational, no sales online sales funnel, etc).
Glad to hear that :)
> Where's your shift key, though?!
The main shift key is on the left side, below the "Esc / Ctrl" button. Additionally I have a secondary shift key on the opposite side that is activated when holding it pressed, but I actually never use the secondary shift key, only the main shift key. Same with right hand side ctrl, I never use that one either, and layer 3 is empty so I don't use the activation key for layer 3 either. Other than that I use all of the buttons on layer 0.
The other layers have quite a few keys that I don't use however, so I should probably update my config to reflect that.
While I am at it I am going to rearrange the Fn keys a bit, because I always struggle to remember where I put F12 exactly so I end up counting from the known position of F1 up to key F12. And the situation where I need F12 of course is when I am going to do something in BIOS/UEFI settings on some computer and like you know you only have like 2 seconds to press the correct key :P So I should make it so that I have F12 at the natural position that is the very top right corner.
Actually, let me go ahead and do that now real quick.
Here: https://configure.ergodox-ez.com/ergodox-ez/layouts/PBKjJ/Zq...
This one now has the unused keys removed, and layers 1 and 2 merged, as well as the () and {} and [] quick-access keys repositioned because while I've been using the [] and the {} quick-access keys a lot I've been a little bugged by my choice of position for them. After making the modifications above I flashed that config to the ErgoDox, meaning that what you see there is matches exactly what I use which is now all that is on my ErgoDox and none of the keys that I thought might be nice to have but which I never use :)
> Regarding stenography for coding, I will probably only use stenographic typing for software documentation (prose). The main use for stenographic typing will be blog posts, emails, HN comments, etc.
I see, makes sense.
> With the ability for the Ergodox EZ to emulate a serial port , I can use Plover+Ergodox EZ as layer 1, say, and Dvorak as layer 0, without needing to change anything at the OS or application level when I do this (Dvorak will be seen as a USB HID which Plover won't see and steno-mode will be seen as a serial connection, which gets processed by Plover and forwarded to the OS).
You have me convinced. I do type quite a lot of prose text too, and I knew vaguely that someone was using the ErgoDox EZ for stenography, but I didn't know how and I didn't know that Ergodox EZ could emulate a serial port. Do you have any links about how to set this up for Plover with the ErgoDox EZ? Otherwise I am sure I will figure it out eventually :)
Edit: Actually, tab is missing from the layout in the web UI for some reason. Put it back in place and changed the link in this comment to point to the version that has tab put back in place.
Edit 2: Made some small adjustments to the positions of some of the keys and flashed the modified to my ErgoDox EZ. Updated link and some of the text in this comment surrounding that link.
> Where's your shift key, though?!
palmface I had meant <kbd>Tab</kbd>. But, you addressed this in spite of my mistake at the end of your response.
When I get Plover running with my keyboard and can toggle between modes successfully I'll send you a link to my explanation to get it started. From the looks of things I'll be cobbling together a few different sets of instructions.
This is the point where you start bleeding value your videos might have for the audience. When content becomes a vector for marketing messages, it almost always goes very hollow very quick.
I think 3Blue1Brown is a good example of things being trivialized to the point where you're tricked into thinking you learned something new, when you actually didn't learn much at all.
Different people learn in different ways.
https://www.psychologicalscience.org/news/releases/learning-...
Motivation and styles is probably a different matter. There is the problem that self motivated learners can only learn if they are motivated, but also that they can be led astray by materials that primarily motivate and entertain.
I’m not at all convinced that programming is primarily text oriented in any case. The underlying concepts are mathematical and logical, not textual. People don’t learn through text or video. Those are mediums. They learn through metaphor and relationships.
So the question is which medium gets those metaphors across most efficiently. Consider, for instance, a binary tree. It’s a metaphor. And I’d be willing to bet most people new to the topic would learn that metaphor better (or at least faster) with a person drawing out a tree and pointing to the path a program would take to add/remove/list/etc. the nodes on the tree than they would seeing a block of code and having someone explain it in text.
In practice, I wonder whether this non-text based learning is more useful in the earlier stages of learning to write code. This might because your learning much more than how to write a valid set of lines that compile to an expected result, there's a lot about style, motivation and norms that you have to soak up too. A lot of that is often implicit in text based tutorials, whereas the video based format makes many minor details explicit (e.g. seeing someone type a line of code wrong, and then explain, 'oh yeah this is a commmon mistake'... this actually encodes a lot of information a text based tutorial would typically miss)
I think this is same with today's maps: it's very easy to find a road with a navigator, but it doesn't leave any lasting impression. Fine if it's a one-time task, but not that good if you want to become an expert on local roads.
With a PDF about data structures you do have the executable environment: the compiler or the interpreter the PDF uses to give examples. The compiler has no manual on data structures, but you have the PDF instead. This is the real thing, by the way, that compiler. An medium that combines both instructions and some interactive thingie that shows the data structure will be much farther from the actual compiler than that poor PDF.
This seems too broad to make a general rule. Shall I not use calculators and memorize trig tables to better understand the tools that are these functions?
I would be interested in seeing that experiment, I cant come up with a concrete example in my head which would follow that pattern.
> maps
I agree, I think most people should not have to become experts on local roads, then again, I think you over simplify the issue. I would bet that most people are experts on their local roads. I am in my hometown, I am not when I travel.
I do not know how pdfs are made today, but keynote/power point are generally drawn with WYSIWYG tools rather than post script or latex. I am not aware that I have ever read a pdf or been to a presentation with slides not generated by one of these WYSIWYG tools in my entire life. I have never actually opened one of these files, but I also don't expect that I could. I would bet they are not stored in text files.
Overall I contest each of your points, though I appreciate the ideas. I understand the sentiment but don't agree on the reality. Sorry for the late reply, I'm happy to continue if you have the time.
Your pencils, paper, and brain are an executable environment.
I contest your second point, by definition my pensil and paper are not executable environments. My brain is, but this gets us into the game of "how much computer can you simulate with your brain."
Why do we play the game of simulate-the-computer? We've had computer environments for at least 40 years now that do that thinking for us, that answer our questions about control flow or definitions or alternate scenarios. We have them, they were thought up, written up and delivered. Yet as a field we value this stupid idea of working alone without help and don't teach these workflows using the computer as a partner. No, in 2019 if you can't do it on paper you don't know it. That's bullshit, yet here I am and I can do it on paper. I'm just sad for those who come after me and have to go through the same wasteful hazing.
Given that subjects have inherent styles of medium that become more essential as you approach expert and need to interact with experts in an efficient way, if you convince someone they have an innate problem you deprive them of the motivation to become an expert, and you see more and more people going to a junior level and dropping out.
Troubleshooting became much easier as I took that advise.
I fundamentally disagree with the two assertions "it's the act of using your brain to simulate the computer that teaches your brain how the computer works so you can reason about it later" and "In most domains, I find it best to start with doing things yourself, and only move on to the tool-assisted version once you thoroughly understand what the tool is doing for you".
I agree entirely with the notion that ability to reason about the underlying system is incredibly important, but I disagree about the methods to get there.
I disagree with those two ideas because (and maybe we have different perspectives here) but the choice of whatever level is the "base level" or "bottom of the stack" seems entirely arbitrary every time. Is assembly the bottom? Or C? No it's machine code. No it's the physical wires.
I think I should come back to my original comment here and reiterate. I'll clarify what I mean about those "computing environments that do that thinking for us" because I expressed myself poorly.
There are computing environments and workflows that people have built which expose to the end user (the programmer) deep information about the state in which they are working. As a field, as a culture we have not embraced this thinking and rather stick to the simplistic notion of working alone to build understanding for ourselves alone.
Sharing is discouraged (outright banned at school under penalty of expulsion) and difficult requirements are kept in place primarily for hazing purposes rather than pedagogical(I have this from a one-on-one discussion with the course designer at my school). There's this thinking that "I had to go through it, and the system produced me, so it must be good" and to think otherwise would be a recognition of being failed by the system, of missing something. A recognition that you could be smarter now than you currently are.
So to give a concrete example, as my sibling comment talks about "being the packet". Why is the 'network', and 'understanding the network' not realized as exactly the same idea? Why should I ever have the simulate the network in my head? The network is an man-created artifact, my understanding of the network should come from the network, the literal code defining it, not from text file RFCs however well they are written and whatever brilliant ascii art they have (because they ARE well written, and well explained by their diagrams). Programmers should have inspection tools that are borne out of the definition of the artifact they are inspecting.
This is of course being done in other disciplines first. An architect working in Revit is orders of magnitude more powerful than an architect working on pen in paper. In every case where an architect prefers pen on paper it is due to a failure of technology to realize practical or targeted workflow affordances, not due to the superiority of paper.
I used tool-assisted incorrectly as a proxy for this scale. There are tools to help provide a more detailed, concrete view in addition to tools that speed up repetitive tasks.
There is some evidence that assistive tools may impede learning, particularly regarding navigation with GPS vs a paper map. If you just want to get the job done, tools are great. If you’re going to be doing similar jobs frequently, it’s probably worth investing the effort to try the lower-level/older method a few times; it’ll make you better at using the more modern methods. It seems that going more than about 2 levels of abstraction away from my original problem gives too little return for the effort, though— Diving into the debouncing circuitry of my keyboard is unlikely to help me diagnose a syntax error.
For learning, I’ll stick to paper and pencil, personally. I could never get my head around double-entry bookkeeping until I started keeping my personal accounts in a paper ledger book. I wouldn’t hire a professional accountant doing things that way, though: I now understand what the tools are doing well enough to appreciate both the benefits and drawbacks.
Similarly, spaced repetition systems like Anki never really clicked for me. It only started to work when I made a physical Leitner box and wrote out cards by hand. I eventually moved to a hybrid system where I computer-generate most of the cards but still print them out and review with physical cards. The extra flexibility of the concrete system let me try thing and figure out what I needed to make a system that works for me; I doubt it’s right for anyone else, but that doesn’t matter.