[saivan]

I get the point you're making - but honestly I have to raise the counter argument, which I think is equally valid. Take for example, the average lecture video. The information density is so low, that I'd imagine that a 60 minute video could be compressed into 10-12 minutes without any loss of information. It really depends on what you're listening to/watching.

[someRandoJunk]

I think some of the replies and likes to your reply are kinda hilarious.

You went through the entire article, misunderstood the point (aka he's talking about people who are cramming information, not people who are using it to skip filler content and contemplate over the actual information like you do), and this misinterpretation is fair, it happens to all of us. Few people corrected you in the reply.

But a lot of people instead of reading the article, took the title of the article and your comment as what the article meant, thus fulfilling the entire thing his article mentioned. Speeding through information. Kinda hilarious.

[jasode]

>, misunderstood the point (aka he's talking about people who are cramming information,

The author is making multiple points and it's fair to consider each claim in isolation.

One of the points is that that active learning is better passive learning. And another point is that reviewing the information multiple times is better than reading it fast once. No disagreement about those. However, the other claim that the speed of 3x is always less retention than 1x isn't true for every listener, every speaker, and every topic.

- 3x can be better for focus because some speakers talk so slowly than listeners tune out at 1x

- 3x lets you listen to 3 different presentations of a topic for reinforced learning rather than only getting 1 perspective in 1x time.

- 3x lets you get past "easy sentences" and selectively slow down to 1x for the "hard dense sentences".

- 3x increases the wpm (words-per-minute) into the normal/natural speed of the reader's "imaginary voice in their head" when reading written text

The author should have titled his essay "Against Passive Learning" because that's the stronger point rather than highlight "3x".

[float4]

> some of the replies and likes to your reply are kinda hilarious

> You [...] misunderstood the point

> Few people corrected you

> Kinda hilarious

If you genuinely want people to understand why they're wrong, then know that this is not the way to do it.

[will4274]

I don't agree at all. The author doesn't mention filler content. He seems to implicitly assume that Mike's podcasts have no filler content. And he assumes Mike's motive to be trying to learn faster - rather than skipping filler - without presenting any evidence that Mike believes this.

I'd say the author constructs what is probably a strawman, that Mike is consuming so fast because he desires to learn as fast as possible, rather than the other obvious hypothesis - Mike is probably consuming so fast because he finds the content a little boring.

I don't think saivan is misunderstanding the author by pointing out the authors (mis)assumption - I think the author is misunderstanding Mike and you are axiomizing the author's misunderstanding to criticize other commenters.

[phgn]

To be fair, the article is quite long so I for example gave up reading after a while. Which from my brief skimming seems like what the author advocates for -- reading less things but more deeply.

[mcguire]

And then there's that...

[daenz]

I use the same youtube plugin mentioned in the article, and watch many videos at 2.5x-3.0x speed, for the exact reason that you state. There is a lot of "filler" content that I either a) already know or b) is not relevant to what I'm trying to learn. I'm really just trying to get through that content quickly. When I get to some really dense portion though, I will turn the speed down to 1x to learn it.

[gumby]

> When I get to some really dense portion though, I will turn the speed down to 1x to learn it.

An alternative that often works is to open the transcript and simply read it. If you find something unclear you can click to jump to that point. Coursera classes have this feature too.

Obviously doesn't work for everything, but it's especially useful when you want to know more about a subject you already know about (say a programming language you've used but never formally learnt).

If I can't do this I usually just close the tab -- the information rate of a video is typically quite low.

[saghm]

I've noticed that I'm fairly unusual in my generation (late millennial) in that I strongly prefer to consume information through reading compared to listening or watching. My brothers and friends and girlfriend all love stuff like podcasts or casual YouTube watching, but I find that the increased effort needed to arbitrarily change speed or skip around always makes me end up not retaining or enjoying the content I consume as much.

[tempestn]

It depends on the type of information, for me. I love podcasts (history, policy, some news) but when I'm trying to research a topic or find instructions on something, or that kind of thing, I also vastly prefer text. Kind of drives me crazy when the most relevant source I can find is a 15 minute youtube video explaining something that could be distilled into a paragraph of text.

[usrusr]

I particularly miss the feeling of being in control: with text I can skip scan, reread and so on with just an eye motion.

The first time I encountered the concept of a 3xer was in the context of political radicalization, people infusing their mind with YouTube self-radicalization content on 3x (or higher) every day. My mind conjured up images of Malcolm McDowell in that A Clockwork Orange scene, only that it's self-inflicted and with content aiming at the exact opposite.

[Thlom]

Same. Especially Youtube videos explaining and showing something really simple that takes like 5 seconds but they go on for 10-15 minutes. I suspect it has something to do with Youtubes algorithms that encourages creators to make long videos.

[pbhjpbhj]

In part. There was definitely a 10 min target time for a long while.

However, a lot of TV shows - particularly USA ones seem to needlessly repeat everything like there running a lecture for amnesiacs. Here's what we're going to say in the first part, here's the first part, we say what we said we would, now a recap, then a break so we review the whole first part ... now we're 10 minutes into the show and we've seen about two minutes of unique footage. It's harrowing -- I'll take overdrawn explanations in preference to that.

[yodelshady]

I would add for practical skills (including some research!) videos and podcasts seem to offer more feedback. Nobody in a book ever tells me what a flange or spline or baulk ring actually is, nobody in a video does either, but in the latter I get to see it and make my own, usually fit-for-current-purpose, inferences.

Closer personal example: I spent weeks trying to bully a supervised machine learning approach into a reinforcement learning one, because the 800-page reference book I used (that claims to cover all machine learning, and is well regarded!) in no way acknowledges the existence of this sub field. For whatever reason, and across multiple fields, I've never found static text to be good at "here's what you should be looking for", and I don't think it's reasonable to discount that knowledge as being valuable.

[b22d432e]

I've long since come the the conclusion that most 800-page texts are terrible.

Nearly all very-large-texts I've read on technical subjects are poorly written. The early and later sections seem to have little relation to each other; Some parts will be too general and other parts too vague. It's like the author totally loses perspective.

There's a sweet spot of around 250 A5 pages where a subject can maintain consistent scope and have meaningful relationships between chapters.

There are exceptions, but they are few.

[jules]

I've noticed that reading is faster than listening at 3x speed. A quick way to test this is to enable subtitles.

[BlueTemplar]

Are you ? After all, YouTube is a recent phenomenon for us, and we've even known a time without widespread Internet when knowledge was still overwhelmingly in books...

[bajsejohannes]

In addition to what you're saying, it can also be _harder_ to watch something at 1x speed. I've found that at 1.25x-1.5x my mind is more engaged. If it's too slow, I start thinking about other things and end up getting less from the video.

[khariel]

To me this is fairly dependent on how fast the speaker speaks. Most of the times 1.25x works well. Sometimes 1.5x is too fast.

[michaelcampbell]

I find this true on podcasts; I'm normally a 1.3-1.5x person listening to podcasts, except "No Such Thing as a Fish". That one I go slower on, just because they seem to talk quite quickly, comparatively.

[nextlevelwizard]

I think the slowness of the videos is for non-english speakers. Meaning people who can understand english to a point, but aren't using english every day.

Some years ago when I used to play games there was these awesome guides to some hard challenges a guy made where he was speaking pretty fast because there was a lot to cover and it was narrated over live footage. It was perfectly understandable to me, but the comment section was full of complains about the speed and how it was too hard to follow. This to me suggests that most of people would prefer if you paused the action to make your point slowly and after that continued with the footage.

[BlueTemplar]

Wait, you can also slow down YouTube videos, can't you ?

[freemint]

Yeah but it sounds like shut then since there are fewer samples per second of Playback.

[nextlevelwizard]

yes, but that's an option you need to select. Normal people aren't going to even try searching for an option like that. Also it was 8 years ago, so I don't know if that was an option back then and in any case that would slowdown the footage as well.

[Zarel]

I also have this problem (it might be related to my ADHD). 2x speed (with occasional pauses/rewinds) works much better for my retention; I've often explicitly noticed myself not paying attention and learning nothing at 1x.

[elil17]

https://addons.mozilla.org/en-US/firefox/addon/sponsorblock/

This lets you skip filler content automatically for many popular videos.

[nextlevelwizard]

It is more that these days everything is a video. Even things that should be two to three paragraphs of text or maybe a picture or two.

[aspyct]

I agree with this, though it depends on what you're learning and what you already know.

For example, I do not care for videos about code. I'm experienced in that domain and I want to get right into the meat of things: scroll to the appropriate paragraph, see the example I'm looking for, and move one.

I could see why a beginner would need a slower pace with more "filler" explanation and background information. Videos are a nice format for this, because they allow one to just sit back like we did at school and take in the information.

But that's only considering programming. Other domains are better suited to videos. For example, visual arts in general: painting, photography, filmmaking... I couldn't imagine explaining a picture with words only, or a human interaction with pictures only. Perhaps when I have more experience, but for now, I like videos.

[fouric]

> I could see why a beginner would need a slower pace with more "filler" explanation and background information.

But that's what links are for - allowing you to deduplicate information by merely providing a link to some other content instead of replicating it entirely.

Moreover, you don't know each beginner's background or desired pace. "Fixing" a certain set of information into the video is worse than providing the appropriate links that allow the beginner to read exactly what they're unfamiliar with, and videos hard-code the pace in a way that written material is not - they're the opposite of what a beginner needs.

> Videos are a nice format for this, because they allow one to just sit back like we did at school and take in the information.

The article specifically addresses this - passive consumption (which better describes videos than reading) is scientifically shown to be less effective for learning than active consumption:

> One study[1] found that active learning makes students think they’re learning less even when they’re actually learning more. That’s one reason why, even though they’re less effective, lectures have persisted for so long.

> Other domains are better suited to videos.

The parent comment ("It is more that these days everything is a video.") wasn't taking any issue with the fact that some things are represented as videos, but that everything is. Of course most filmmaking education (modulo some stuff like maybe an introduction to optics) is best done by video - but nobody is complaining about that.

Also, in terms of education, these subjects, while they exist, are a minority. The majority of stuff you learn in school is better done in a non-video format. Not a text format - diagrams and interactive simulation are incredibly valuable for understanding. But, specifically, video is almost exactly the opposite of a good format for learning most things.

[1] https://www.pnas.org/content/116/39/19251

[nextlevelwizard]

It is just sad when I you search for "how to do X in Linux" you get a video in search results first and only second some article where you can actually copy-paste the commands

[thepasswordis]

Which addon are you using for youtube?

[jholman]

I wrote a bookmarklet. Works on all non-iframe audio and video.

javascript:void%20function(){document.querySelector(%22video,audio%22).playbackRate=parseFloat(prompt(%22Set%20the%20playback rate%22))}();

Here's a non-interactive version

javascript:void%20function(){document.querySelector(%22audio,video%22).playbackRate=2.7}();

[vbezhenar]

"<"/">" buttons decrease/increase playback speed (up to 2.0).

[razh]

Ilya Grigorik's Video Speed Controller?

https://github.com/igrigorik/videospeed

[zoul]

For Safari there’s Dynamo, too, that lets you skip ads apart from changing the speed:

https://dynamoformac.com/

(I am the author.)

[daenz]

https://chrome.google.com/webstore/detail/youtube-playback-s...

But I suspect they're all mostly similar

[Topgamer7]

Enhancer for Youtuber

[michaelcampbell]

This is a game changer.

[gonehome]

I just modify it directly in the console (I have to search to look up the command every time).

[CompuHacker]

Hit the Up arrow key in the console; in Firefox at least, command history persists across sessions.

For reference; $('video').playbackRate=3.33

The playback engine mutes audio below 0.25x and above 4x, not configurable.

[Nursie]

Honestly, maybe I'm just some old, out-of-touch luddite, but I think that using videos to pass information is sub-optimal all around.

Unless we're actually showing audio-visual phenomena, a page of text is almost always more useful to me. I can absorb it at my speed. I can go back and forth within it easily. I can search it. I can copy bits out if I need to. It's just better.

[chronogram]

Nothing to do with your age, it's always been either a difference between users of the media, convenience for the producer, that it's easier to monetize video, or the ability for instant feedback when it's in-person or live.

Socrates to Plato perhaps: maybe I'm just old but using text to pass information is just making your memory weak.

[albert_e]

you are not wrong or outdated

different people prefer different methods of learning

they may or may not be more effective -- just that they are more preferred -- even if for no other reason than ease

sitting down with a page of text and focusing on it to learn new information is becoming harder and harder for me personally ...with the bad habits of constant smartphone and social media use

I fall back to have someone do the reading and explain it to me

videos let us pause / rewind / skip / slowdown as needed ..so I am noticing that I am depending on his control also and sometimes zoning out of videos too ...

...which sometimes bites me when I am watching a live stream that has no rewind or worse ..attending a real meeting and hear someone explain something at length

[BlueTemplar]

But video controls are tremendously worse than just being able to focus your eyes at any specific portion of the page in an instant ?

[gibolt]

Subtitles matter! While a video is not equivalent to just text, reading at 3x with audio is similar to why people enjoy audio books.

Visual content is a bonus, to remain more engaged and maybe impart information via a third medium

[sb057]

They key advantage to video (and the reason why YouTube seems to keep expanding to encompass more and more subcultures) is that video formatting is also inclusive to text, audio, and still image formatting. You can only upload text to a blog, and you can only upload audio to SoundCloud, but you can upload everything to YouTube.

[cthor]

Blogs are perfectly capable of embedding media as appropriate, including videos.

On the contrary to your suggestion, blogs (or more specifically, webpages) are what can do more than anything else, since they can also feature interactive media. See <https://ncase.me/trust/> for an example.

Whether or not this is expected, standard, or the author thinks it's worth the effort is another thing.

[Nursie]

> video formatting is also inclusive to text

Are we talking slide shows on a youtube video here? Because to me those are probably the worst of all worlds. Low information density and not searchable.

[memco]

Long before audio books were a public thing I received a special tape player for listening to audio recordings of my school textbooks because I have low vision. I was ecstatic to learn that I could adjust the speed and could still understand the book at something around 2-2.5x speed. I don’t know to what extent other bling or impaired persons use the speed controls but I’m guessing it’s designed because some people can process the information and others more slowly and this isn’t a bad thing per se. I also agree that the information glut is not a good habit but listening to something at a faster speed is not in abs of itself information glut. Sometimes the bottleneck in presentation speed is the speaker not the listener.

[kixiQu]

> I’m guessing it’s designed because some people can process the information and others more slowly

Familiarity with the speaker's voice and presentation can influence this hugely. There's a news podcast I like listening to much faster than most other stuff I listen to just because I know the newscaster's voice so well.

[luguenth]

I think it's not really about the speed of the information, as long as you are able to process it, but the sheer amount of information. If you use time savings of 50 minutes to just consume more information, not much will settle in your long term memory. But if you process and repeat the information in those "saved" minutes, you'll get away with much more in the end. I think this is more the point of the author's view.

[rsanheim]

This ignores the point (that Perell makes in the essay) that learning via lecture is a horrible way to learn much of anything to begin with.

Now I could see a counter point where you speed up a lecture, find the new information that interests you, and then use that as a jumping off point for repetition. For example, dive deeper into other sources, take notes, use flash cards, try applying what you've learned, and so on.

But just consuming 10 mins of new info from a full lecture at 2.5x speed and then moving on probably isn't doing much long-term learning.

[Joker_vD]

Off-tangent: modern lectures are still a better way to learn something than original lectiones were ― the lecturer would read the book by some prominent author, and students would listen to it and take notes... and that's it. That's what lectio literally means: "[an act of] reading". And before the invention and spread of the printing press, it absolutely made sense ― books were rare and expensive.

Today, of course, lectures during which the lecturer simply reads the textbook and does nothing more, are rightfully considered to be the worst: a student too can read the textbook himself just fine!

[TrackerFF]

I think that tutors like Khan of Khan Academy kind of revolutionized the shortening down of lectures. The videos are split up into smaller chunks, they're much more direct, and go straight to the point.

Andrew Ng is a bit similar. Much shorter, more dense videos.

[faeyanpiraat]

I have to use x1.7-x2.0 speed on pluralsight courses, even though English is not my native language, and I’m learning completely new stuff.

Otherwise I’d almost fall asleep.

It’s better to occasionally rewind a couple of unclear sentences than wasting 2x the time.

[dredmorbius]

I'll often seek out both the written and lectures on material I'm particularly interested in.

I can read far, far faster than I can listen (and at what seems to be 4-5x the projected reading time in most article guides ... I'd really prefer a simple word count).

But ...

... listening to the spoken lecture can result in a very different understanding of material, hearing the lecturer's intonation, emphasis, humour, and more. This applies both where the reader is the original author and, at least in cases, where not, if the reader knows the material and its author well.

This of course depends on the material and ones level of interest in it. Multiple exposures for high-quality and complex material, or particularly compelling dramatisations, are worth this in my view.

[JackPoach]

It's not just density, many books are pure nonsense. So 200 books of pure nonsense a year won't teach you much. They'll just introduce you to a ton of terrible ideas.

[visarga]

Hey, that's how AI's learn as well. Reading up all the nonsense indiscriminately and making no effort to make the ideas consistent. But it's better than not reading because you get exposed to a larger variety of text so you can draw upon them when it's time to get creative.

[RandomLensman]

You describe a symptom of a high data/low effect situation that is pretty common these days.

The real question is what to do with the 50 minutes left after distilling the information down to 10 minutes. Just absorb more data or do something with it?

[lrem]

From learning point of view? Unanimously do something with it. Or at least see some examples of someone doing something with it to make you care about what you just learned. It's very high to remember a fact, as in commit to long-term memory, if there was no feeling attached.

[dheera]

I'm very slow with books. I'll read a couple paragraphs and then ponder for a minute. And then maybe look up some tangential, maybe even radial stuff. Even audio books at 1X move too fast for me.

I read maybe 4 or 5 books a year. I don't understand how people do 100.

[gonehome]

For an example at the extremes, you can run congressional testimony at 4x, but Andrej Karpathy or John Carmack at only 1x.

It really depends on the quality of the speaker and the content.

[quickthrower2]

Yes anything about category theory is 0.1x speed. Or I need to watch 10 times at 1x speed to understand anything :-)

[sschueller]

I would prefer a plugin that could remove "ehm"s and pauses from speakers that are new to the game.

[mb_72]

If only this wasn't necessary - a lot of online stuff has had a single 'take' done and minimal editing out of such time-wasting utterances or silences. And once you start to notice such characteristics in some poor speakers it can be a complete deal-breaker in terms of actually learning something.

[dredmorbius]

I have quite literally gone through long lectures and edited out such filler words (Audacity is good for this), where the material is sufficiently compelling (an extreme rarity).

It's really a telling level of contempt for an audience to allow unedited material containing excessive fillers to be released. I'm not at all a fan of the "one take, FI/SI" school of podcasts, and will bail out of virtually anything that features this.

For vapid voiceovers, I'll often just watch the video with sound off. My response is similar to how Douglas Adams described Marvin the Android hearing people count.

[tsmeets]

Not sure how well it removes "ehm"s, but I use unsilence[1] a lot for lectures. It removes the silent bits from a video file. It isn't a browser plugin however. You have to download the lecture before converting.

It works quite well in my experience.

[1] https://github.com/lagmoellertim/unsilence

[dgellow]

I believe that’s what Overcast “Smart speed” does for podcasts. Would be nice to have for videos.

[Thlom]

Pocketcast can remove silence at least. Not sure about filler sounds like ehm.

According to Pocketcast I've "saved" over 3 days by trimming silence and over a month by speeding up. I've listened to a lot of podcasts it seems ...

[skrebbel]

The article uses 3x speed as a metaphor for a broader mindset. It's not actually about 3x speed.

[18al]

Using rate of speed up is probably a bad metric due to varying densities, but even if one were to account for that and use some kind of smart speed up app that maintains constant information throughput, the issue is with not taking pauses to ruminate.

It's more of an information retention problem rather than an information loss one. IE not committing to long term memory as the author states.

Not very unlike consuming food without chewing.

[BlueTemplar]

Yeah, I've noticed that with text I'm going to make more pauses thinking about what I just read (especially printed text for some reason). Video is the worst, while audio only in the middle. Maybe because of clunky controls ?

[thrower123]

I really hate how everyone has shifted to video and podcasts over the last ten years or so.

Personally, I'm an "in one ear, out the other" type as far as auditory memory goes. So I can read something written out, or even just transcribed, in a fraction of the time and actually remember it.

[tombert]

I suspect someone with better understanding of psychology can tell me if I'm way off on this or not.

-----

Sometimes I'll put a lecture at 2x speed if the professor is talking way too slow. Every ten minutes, I will pause the lecture and try and "teach myself" what the professor just said, giving a quick summary of all the information I remember. If I feel like I got a reasonable understanding of the gist of it, then I keep going at 2x, and if I had a lot of trouble with the summarization process, I drop it back to 1x.

More often than not, I end up dropping back to 1x.

[dqpb]

My experience is that speeding up low-quality content makes it more tolerable, whereas speeding up high quality content makes it less enjoyable.

So now I just don’t speed up. If the quality is low, I don’t listen to it. If the quality is high, I enjoy having the time to think about what the speaker is saying, while they’re saying it.

[SergeAx]

And that's why reading a book or an article is 10x better than listening or watching a lecture.

There's another perk of reading: one may easily jump back couple of words or even sentences in case they need some clarity. You can't do that with A/V-recording.

[onion2k]

Listening to a lecture at a faster speed doesn't change the information density. You're just compressing everything; you're not editing out the useless bits.

[saagarjha]

Yes, it does: the useless bits remain as time to help you digest the information. It's just that you don't need as much time to do so, so making everything faster is just fine.

[Aeolun]

I could watch my work meetings at infinite speed, and the information density would still be zero.

[cush]

Seriously. It depends entirely on the content