Youtube "how to" videos are mostly useless

[tggzzz]

If the how to video is how to fix something, they are usually quite valuable. They may not give you a complete way to do a good fix, but boy can they save you time figuring out where all the obscure fixings are, without poking around on the real product so much you break something.

That can be true, especially with modern electronics. I used one when replacing a screen on my Hudl tablet.

I still had to actively avoid several pieces of crap.

[IanB]

For example, I still have not found any description for choosing a solution method for partial differential equations other than "try each until one works".  I know such a method works, because ask any mathematician, and they'll show you the working method in real time; I just haven't found one who can explain what they based their choice on (except "I've seen this form before").

Interesting. I would say the answer is one of pattern matching. You could probably train a machine to recognize which solution method to apply given a sufficiently large training set of PDEs. With experience, a human brain could apply the same learning. We tend to call this "intuition", but in reality there is no such thing. Intuition is just a fancy word for "I can guess what to do because I see a pattern in the data."

[CatalinaWOW]

I want to double down on one of my prior points.  A good textual document is more than enough for almost all of the topics mentioned in this thread.  Such a text is difficult to generate for most people, and the number who can do it well is in my opinion declining in this video focussed era.  Meanwhile the number who can do an adequate or better job of video content creation is increasing.

This partially changes the discussion from suitability to availability.  Trash text content has always existed.  Time winnows out the lesser documents.  Presumably the same will happen for video.

[coppice]

I want to double down on one of my prior points.  A good textual document is more than enough for almost all of the topics mentioned in this thread.

For repair work the visual aspect of a video is its greatest strength. Most of what I want from those videos is to see how things fit together. For most things its complex. If learning from books was so great, why would people spend a fortune to learn from lectures in colleges? Really smart people learn quickly from book, while less able people gain massively from lectures. However, even the smartest people pick up new things faster with a combination of books and lectures.

[tggzzz]

Such a text is difficult to generate for most people, and the number who can do it well is in my opinion declining in this video focussed era.  Meanwhile the number who can do an adequate or better job of video content creation is increasing.

I don't think the number of people who can do a good job is diminishing. Instead, the number of people that can do a bad job is increasing.

I think the problem is that the low cost of publishing is the root of the problem. That removes the need for editorial decisions.

This partially changes the discussion from suitability to availability.  Trash text content has always existed.  Time winnows out the lesser documents.  Presumably the same will happen for video.

The problem is that the trash text/video is increasing faster than the good content, so the good content is drowned out.

That's exacerbated by the video medium: it takes me 15s to scan text before ignoring it, but 15 minutes with video. 60:1 is a killer.

When it cost money to publish, a suggestion for an "I opened the box and it looked like the advert" article would have caused howls of laughter.

(Oh hell. That reminds me of the later Jerry Pournelle articles in Byte.     Damn; now I need to go and practice some relaxation therapy )

[coppice]

That's exacerbated by the video medium: it takes me 15s to scan text before ignoring it, but 15 minutes with video. 60:1 is a killer.

That's not really an issue of the medium. Its mostly the choice of the author, looking to keep you viewing. Few videos make a clear statement up front about what's to come. while most lengthy text has a contents page up front.

[tggzzz]

I want to double down on one of my prior points.  A good textual document is more than enough for almost all of the topics mentioned in this thread.

For repair work the visual aspect of a video is its greatest strength. Most of what I want from those videos is to see how things fit together. For most things its complex. If learning from books was so great, why would people spend a fortune to learn from lectures in colleges? Really smart people learn quickly from book, while less able people gain massively from lectures. However, even the smartest people pick up new things faster with a combination of books and lectures.

Remember the old Haynes manuals showing how to repair your specific type of car?

Once upon a time they were full of clear drawings showing the essential points of the next step. Then photo reproduction became cheap and easy, so unclear visually confusing photos were used instead.

[tggzzz]

That's exacerbated by the video medium: it takes me 15s to scan text before ignoring it, but 15 minutes with video. 60:1 is a killer.

That's not really an issue of the medium. Its mostly the choice of the author, looking to keep you viewing. Few videos make a clear statement up front about what's to come. while most lengthy text has a contents page up front.

Most articles and videos contain a couple of thousand words. That's not lengthy when speed reading (30s) but is when watching (30mins).

[Ranayna]

I think YouTube is so big it takes time to sort the gold from the dross.

That is what this problem essentially boils down to, not only regarding Youtube, but in general.

You as an individual cannot know everything. There is just too much of everything out there. Everywhere.
If you know, you know. If you do not know, you either sometimes painstakingly search for it yourself, with a dubious chance of success.
Or you ask someone whose opinion you trust and ask them. And even then what they tell you might be not applicable for you.

This is already difficult enough for technical topics. Add softer topics and politics (shudder) into the mix and well.... That is going to hell quickly at the moment.

[coppice]

I want to double down on one of my prior points.  A good textual document is more than enough for almost all of the topics mentioned in this thread.

For repair work the visual aspect of a video is its greatest strength. Most of what I want from those videos is to see how things fit together. For most things its complex. If learning from books was so great, why would people spend a fortune to learn from lectures in colleges? Really smart people learn quickly from book, while less able people gain massively from lectures. However, even the smartest people pick up new things faster with a combination of books and lectures.

Remember the old Haynes manuals showing how to repair your specific type of car?

Once upon a time they were full of clear drawings showing the essential points of the next step. Then photo reproduction became cheap and easy, so unclear visually confusing photos were used instead.

Why are so many of those photos so poorly lit, and poorly focussed? For understanding ideas, graphics are generally the best choice, but for seeing around the details of a specific product, and how awkwardly shaped things fit together, video can be a real boon. A key issue with the proliferation of how to videos for niche things, like specific product repair, is they just wouldn't exist if they were not as quick and easy to produce as a crude YouTube video. So, your choice is a less than ideal format, or nothing.

[coppice]

That's exacerbated by the video medium: it takes me 15s to scan text before ignoring it, but 15 minutes with video. 60:1 is a killer.

That's not really an issue of the medium. Its mostly the choice of the author, looking to keep you viewing. Few videos make a clear statement up front about what's to come. while most lengthy text has a contents page up front.

Most articles and videos contain a couple of thousand words. That's not lengthy when speed reading (30s) but is when watching (30mins).

I think it really comes back to what I said. A 2000 word article probably doesn't have a contents page, but it probably does have some bold section headings that are quick to scan. Effective videos give you a good outline in the first 30s, but that massively cuts total viewing time, and potential revenue. There are perverse incentives at work.

[IanB]

If learning from books was so great, why would people spend a fortune to learn from lectures in colleges?

Well, probably they shouldn't spend a fortune. Especially in the USA, college fees are way too high.

But putting that aside, I think you are paying to know what to learn and to have a structured learning experience, and have the incentive to complete the program, as well as access to tutors and teachers to ask questions. How often have you read a book, and thought "Hey, I don't quite get this, I would like to ask the author a question about it."

[coppice]

But putting that aside, I think you are paying to know what to learn and to have a structured learning experience, and have the incentive to complete the program, as well as access to tutors and teachers to ask questions. How often have you read a book, and thought "Hey, I don't quite get this, I would like to ask the author a question about it."

Sounds like you want a private tutor, not a college course.

[IanB]

Sounds like you want a private tutor, not a college course.

I can tell you with absolute certainty that my undergraduate engineering degree was essential. There is no possibility that it could have been replaced with book learning or watching YouTube videos (even if YouTube had existed back then).

[tggzzz]

That's exacerbated by the video medium: it takes me 15s to scan text before ignoring it, but 15 minutes with video. 60:1 is a killer.

That's not really an issue of the medium. Its mostly the choice of the author, looking to keep you viewing. Few videos make a clear statement up front about what's to come. while most lengthy text has a contents page up front.

Most articles and videos contain a couple of thousand words. That's not lengthy when speed reading (30s) but is when watching (30mins).

I think it really comes back to what I said. A 2000 word article probably doesn't have a contents page, but it probably does have some bold section headings that are quick to scan. Effective videos give you a good outline in the first 30s, but that massively cuts total viewing time, and potential revenue. There are perverse incentives at work.

A clear statement would be welcome. I'm sure you are familiar with "Tell them what you are going to say. Say it. Tell them what you have told them"

But that requires the look-at-me-merchants to have planned, in which case they would also avoid ums and ahs, and realise they haven't got anything worth saying.

Summary: not going to happen.

[Nominal Animal]

I would say the answer is one of pattern matching.

Quite likely.  The tomes of formulae tggzzz mentioned and their popularity indicates the same.

As to "intuition", mine is all wonky.  Sometimes it makes these leaps that end up being correct, without any pattern similarity.  Sometimes it jumps off the bridge to the la-la land.  I like my intuition, but I'm pretty suspicious of its correctness, too.

I call the understanding of the properties and behaviour without understanding the internal details "intuitive understanding", mostly for a lack of a better term, and because to me it is similar to being able to catch a thrown ball without being able to describe the trajectory of the ball (except in the crudest of approximations).

[tggzzz]

Sounds like you want a private tutor, not a college course.

I can tell you with absolute certainty that my undergraduate engineering degree was essential. There is no possibility that it could have been replaced with book learning or watching YouTube videos (even if YouTube had existed back then).

Precisely.

 "Push this button to floggle the spatchcock and make it yellow" should not form a central part of university education. It should be left to on-the-job training.

[nctnico]

I want to double down on one of my prior points.  A good textual document is more than enough for almost all of the topics mentioned in this thread.

For repair work the visual aspect of a video is its greatest strength. Most of what I want from those videos is to see how things fit together. For most things its complex. If learning from books was so great, why would people spend a fortune to learn from lectures in colleges? Really smart people learn quickly from book, while less able people gain massively from lectures. However, even the smartest people pick up new things faster with a combination of books and lectures.

Remember the old Haynes manuals showing how to repair your specific type of car?

Once upon a time they were full of clear drawings showing the essential points of the next step. Then photo reproduction became cheap and easy, so unclear visually confusing photos were used instead.

I agree. The Haynes manuals have gotten worse. Often I try to find repair videos for the car but I find skipping through the videos in order to find the relevant 5 seconds out of 30 minutes a tedious job to do.

[Nominal Animal]

A clear statement would be welcome. I'm sure you are familiar with "Tell them what you are going to say. Say it. Tell them what you have told them"

But that requires the look-at-me-merchants to have planned, in which case they would also avoid ums and ahs, and realise they haven't got anything worth saying.

Summary: not going to happen.

Considering how quick and easy it is to just record a video on your phone, then upload to Youtube, compared to actually writing the same information, I'd say you have hit on one of the core reasons for the proliferation of bad "how to" videos.

[IanB]

I agree. The Haynes manuals have gotten worse. Often I try to find repair videos for the car but I find skipping through the videos in order to find the relevant 5 seconds out of 30 minutes a tedious job to do.

Where it gets amusing is when the first 30 minutes is spent dismantling and removing practically everything in the engine compartment in order to gain access to the part that needs replacing, with the mechanic calling down all sorts of curses on the engineers who designed it.

[IanB]

Just for fun, here is an example of a short "how to" video with 100% information content, and no fluff:

[joeqsmith]

Youtube "how to" videos are mostly useless

[tggzzz]

Just for fun, here is an example of a short "how to" video with 100% information content, and no fluff:

Oh, good videos do exist, just as nice opera arias punk songs exist. The problem is finding them amongst the deluge of awful stuff.

BTW: that's not a yooootooob vid, it is a TokTuck vid 

[mwb1100]

Just going to throw my 2 cents in:

There is certainly a lot of junk on youtube.  But I have found it to be an amazing resource even more.  When I need to repair something in my home there is often an easily found video explaining and showing exactly how to do it.  And often I'm looking for a repair for a specific appliance or auto part and I can find a video showing how to do the repair not just on a generic dishwasher or car, but my exact model.

Most of the time it's not too difficult to avoid the junk or quickly ignore it since it's usually clear within a few moments that it won't be much use.

On the negative side, there are many tutorials that would be much better in written form than in a video.  But who am I to tell someone how they should offer/present something I'm not paying them for?  I'm not their boss.

It still amazes me how I can go to my computer or phone and ask nearly any question and much more often than not get useful answers immediately. 

It's like I'm living in Star Trek!  Captain Kirk:

Subject: former governor Kodos of Tarsus IV, also known as Kodos the Executioner.  After that, background on actor Anton Karidian.

[tggzzz]

It's like I'm living in Star Trek!  Captain Kirk:

Subject: former governor Kodos of Tarsus IV, also known as Kodos the Executioner.  After that, background on actor Anton Karidian.

That is a decade too late...

Have a look at Arthur C Clarke's "The City and the Stars" (1956), and the earlier variant "Against The Fall of Night"(1948).

Or, if you prefer a version from 1975, "Imperial Earth", which does feature technology like the web - a decade before the internet and two decades before the web.

Clarke was well aware of Vannevar Bush's memex from "As We May Think" (1945).