It is a shame that for the longest time, MIT OCW's CS & Math lectures were the only high-quality course-videos on the internet. Leveraging the IIT system to get 100% course coverage over every possible engineering course is genius. It has democratized education better and earlier than any unicorn. (Coursera, Udacity, etc.). Everything from assignments to
It is an indictment of American universities that despite having the funding, resources and technology to do this a lot earlier, it was eventually a 3rd world country that pulled it off. All without any profit motive. "Non-profit universities"?, yeah right. Just take a look at the courses on offer - https://nptel.ac.in/courses . Every imaginable engineering specialization is on offer, down to the most obscure.
University classrooms do not seem to have caught up to just how easy it is to access this information. Most professors do not have the same caliber or flair to do better than the best NPTEL video. I would like to new coursework designed around integrating video lectures of this sort as integral part of engineering instruction, with the in-person professor taking on the role or assistance and disambiguation, rather than
being an inferior repeater of the same information.
I learned probably 90% of my math knowledge from YT videos and Kahn Academy. The fact is that the odds you get a decent teacher at a public school or even university are low, while the internet allows you to access the absolute best people in the world at explaining and conveying information for your selected topic.
it's like school choice at global scale, why learn from a person you happen to be geographically located close to at school when you can learn from the best? Combine that with some of the new AI personal tutor tools coming out and I don't see how traditional schooling competes beyond acting as day care while parents work. And even at the college level a huge number of professors are obviously not passionate or skilled at teaching their subject and class sizes are too large to give student specific feedback.
I'm really looking forward to when you can take a placement test and have AI look at your work and solutions and then tell you exactly where the gaps in your knowledge are and create a study plan. So many people struggle with math because they didn't properly learn lower level fundamentals and without a good tutor you can't figure out what you need to learn. I feel like AI is getting to a place where it can probably get there, something like the GPT and Wolfram Alpha integration
I absolutely can't wait for AI language tutors. Private tutoring is extremely expensive and if an AI could do it, even if it's not as good I imagine everyone could have their own private AI language tutor.
I don’t think you can make effective AI without them having boredom/frustration. Otherwise they might do very unintelligent things, like trying the same approach to a problem forever.
It's actually interesting. UC Berkeley had to take down their videos since they weren't subtitled (and therefore not ADA compliant).
There's an interesting industry around this. I wonder if they could have sold a license to them to me for $1 and then I would be free to put them on the Internet.
Alternatively, perhaps the YouTube automated captioning would have been sufficient.
I'm jaded because I know there is a cottage industry of attorneys who go around suing for ADA violations. I wouldn't be surprised if there is some special interest group that goes around claiming automated captions aren't good enough and using them violates their client's civil rights.
Isn’t that literally the enforcement model of the ADA? It isn’t a special interest group so much as attorneys engaging with the system as it was designed, right?
The challenge is that these attorneys aren't acting in good faith, they're doing it for profit. Although they purport to advocate for the rights of the disabled, their business model is to find violations, however small, file a suit, and then offer to settle for several thousand dollars. Most of these lawsuits happen without an actual disabled person (who isn't the attorney) raising a complaint in the first place.
There was a semi-famous example of an attorney suing Chipotle because the counter was too high and he claimed that he was being denied the full Chipotle experience of seeing the food be prepared in front of him. The workaround of literally bringing the food to him tableside and preparing it in front of him there was not deemed good enough, and Chipotle lost the suit. As you can imagine this attorney makes quite a bit of money going around suing people, for reasons the average person would not find genuine.
"Chipotle argued that its accommodation was adequate and told the Supreme Court that Antoninetti lacked standing to sue. The company cited a federal judge's findings in 2008 that Antoninetti had sued numerous businesses over disability access and had not shown that he was sincere about intending to return to the restaurants if they lowered their barriers."
https://www.sfgate.com/nation/article/Chipotle-loses-disable...
I've been putting my podcast voice recordings into Whisper and it produces 99.8% usable subtitles from the language recognition perspective, and maybe 95% usable from timings perspective... So yeah, it's been really good...
I was thinking of that scenario recently now that Youtube has Subtitles/CC with "English (auto-generated)" would that have satisfied the ADA compliance.
> It is an indictment of American universities that despite having the funding, resources and technology to do this a lot earlier, it was eventually a 3rd world country that pulled it off.
I assume knowledge discovery and sharing doesn't depend on your country's world status. A lot of fundamental knowledge was discovered and passed down to us from ancient times by people who would consider life in the so-called third world countries to be a significant upgrade.
It's a bad look because the idea and technology were there and the execution is low cost.
Presumably, American universities wanted to 'protect' their information from others, maintaining exclusivity and gating off knowledge. It is this 'profit seeking behavior' that's shameful.
One of my classes at IIT-M from 15 y ago or so is there holy shit. Haha! So what would happen is they'd put this single camera operator at the back of the class behind the last row. There was only one classroom where we had it.
The prof must have been mic'd up because the audio comes through quite clearly. We weren't told to act any specific way or anything and no students were told explicitly that it was for NPTEL.
But it wasn't recorded in a studio, just a normal classroom with students.
Found some really interesting courses like this one:
Introduction to Launch Vehicle Analysis and Design, IIT Bombay
https://nptel.ac.in/courses/101101086
> It is an indictment of American universities that despite having the funding, resources and technology to do this a lot earlier, it was eventually a 3rd world country that pulled it off.
I don't understand why this needs to so adversarial but for what it's worth, American universities do already provide videos. You mentioned OCW, but EdX does so too, Berkeley did what NPTEL is doing a decade ago. Many universities do so piecemeal.
Remember, that if an entity is a nonprofit, that entity still needs to raise money from somewhere. It needs to show statistics about how impactful its efforts are.
The main reason why there isn't more of this is because its educational value as far as we can tell is marginal, pretty close to zero, at best.
For example, if a course already exists on OCW, the value of having another course from another university seems very low. Why would someone pay for another archive of a course at another university?
Then there are the completion statistics. They show that virtually no one completes MOOCs. OCW and EdX numbers show the same. It's hard to find anyone to give money to these programs.
Now, we might be measuring the wrong thing. In the US if you want to study something you will definitely find a university to do it. It could be that there is great value when the baseline you're comparing to is Nigeria. That's just not what US or EU universities look at in terms of numbers. How could they and who would they pitch this to for funding?
It's extremely expensive to keep a video archive like this up to date. NPTEL videos contain what in the US would be considered copyright violations on their slides. You cannot imagine how much effort OCW has to go into to avoid this. You simply couldn't do what NPTEL is doing in the US. For what it's worth, I think we should relax IP laws, but it is what it is.
I'll say this for the NPTEL archive. The quality is extremely variable. From a few really good courses, to a lot of rather poor ones, to things like the "Modern Computer Vision" course recorded in 2022 that was "modern" in maybe 2013-2014.
Also for educational value, I agree with the US numbers. If you look at the most popular youtube videos called out in the article, they're not like the NPTEL archive. Dry courses presented in textbook order. They're much more interactive, much deeper dives, that slowly walk you over some piece of material from different points of view. More like Khan academy than OCW.
> For example, if a course already exists on OCW, the value of having another course from another university seems very low. Why would someone pay for another archive of a course at another university?
This is where you are sadly mistaken. There is no canonical course out there! Every course is different; every teacher is different. Just because a course on, say, "Data Structures" exists, does not mean that every person who's interested in Data Structures can just watch that and find it equally informative. Many people may not understand the teachers accent; many may not like their teaching style (speed, delivery, stress, etc.). There are so many variables!
NPTEL has been doing this for a good 20 years now.
> EdX does so too
In my experience, EdX videos feel like the pop-sci version of the real course. Some excellent courses are 1:1 copies of OCW courses (6.0.31), but most others have been disappointing.
I transitioned from MechE to CS/ML, so I have spent a lot of time wading through sub-par Edx courses (and on other platforms), before finding the good ones by pure brute force.
> needs to so adversarial
My experience at American universities is that the professors and teachers are wonderful. But, the administration operates on ruthless profit seeking incentives, which often leads to worse outcomes for the nation as a whole. I am not adversarial per se. I just expected better from universities I deeply respect.
> virtually no one completes MOOCs. OCW and EdX
Which is exactly why I am a big advocate of just releasing the videos and assignments. Let people do it at their own pace, and quit at their own pace.
> NPTEL videos contain what in the US would be considered copyright violations on their slides. You cannot imagine how much effort OCW has to go into to avoid this.
Haven't thought of this. I might be a good point. Could you give an example ? I am not fully aware of the kind of copy right violations that occur on slides. Either ways, glad India treats IP laws with the kind of reckless abandon it deserves.
It's reached meme status (as in people will make and share the funny pictures about it) in some international western circles, leading me to believe that everyone's favorite teachers are indian.
Personally I sometimes find the thick accents hard to deal with, but really, when all other materials let you down, you can count on the indian tutor to cover the topic.
Big thanks to all of the (largely indian) internet tutors doing my lecturer's job.
When I attempted to learn C back in 2016, one of my absolute favorite youtuber teachers was a native Indian (accent n all). His videos were much more detailed and better designed than what was out there at the time. After doing some research on his background, I found out he ended up getting a job at Google and lives in Cali now! Put a smile on my face.
When I was doing my CS degree I was watching youtube lectures in Hindi(I don't speak Hindi) and following along with what was on the board. I did my degree just before the wave of Indian teachers in English was a thing. Respect to these people.
I'm too old to have done this when I went through school, but I suspect a good practice is to go through your syllabus, and but the major unit covered into the youtube search bar.
I suspect you get all sorts of parallel presentations of the material in a way that is quite helpful for retention.
a lot of great tutorials and education is out there, especially on YouTube. glad others are making use of it today. I was watching the Korean drama 'Dr Cha', and the look on her face when she found out she could watch training videos on anything, and she said something like 'this is a brave new world'.
Nigerian reversal (of sorts): Indian vloggers are fullfilling the niche which Nigerians can't.
Of course it's always a question of how good the presenter and the course and how good are students actually learn the topic from these videos, but at least people themselves praise them so there are definitely a sense of gratitude.
What makes it so that Nigerian presenters _can’t_ while Indian presenters _can_?
I’ve certainly seen Nigerians who understand the material well enough to present… is it the means of production? The non-CS skills (editing, audio, etc)? Just straight access to resources?
It's not the quality, it's availability of means, for sure.
> Just straight access to resources?
IMO, as I understand it, it's just that. There are enough people in India who can do these videos, there are not enough people who can do that in Nigeria. And considering YouTube algos.. there are probably when you could had a more 'local' video but which was buried at th page.
That’s a shame. I mean I’d assume — admittedly, as an English-speaking American white dude who never had the slightest difficulty accessing “local” content — that having a presenter you can directly identify with using locally relevant examples (simple things like local currency, pricing, products and names) and adapted to local norms (maybe steering discussions of probability away from gambling or usury in Muslim communities, for example) would really help convert kids at the fringes of confidence / motivation to successful learners, if only by removing the need to translate already unfamiliar material into local / personal relevance.
If it were possible to get the small number of Nigerians who can do it to actually do it (maybe by providing a production suite in a box, if physical means of production is the issue), then gaming / overwhelming the algorithms probably isn’t that hard, just by actively encouraging local users to engage with the content actively (driving up the ads watched, and hopefully ignored).
With my background I could easily put together the suite in a box, I just haven’t the slightest clue who needs to get that box… that takes local expertise I just don’t have.
You get it. Evrrything can be done, but you need a talent 'on the ground' and someone who would guide them. There are tons of missions in Africa with the best intent in their hearts, but the reality is a harsh mistress. Just pouring money there doesn't do nothing, just like everywhere else.
You’d honestly think, in this day and age, when we’ve got all these matchmaking algorithms for helping rich folk hook up and get celebrity-endorsed food from ghost kitchen to door on an electric bike while still hot, that we’d maybe have some means of bringing the right people on the ground at the edges into contact with the means and the skills and the best intent nearer the root.
Clearly it can’t be something originated by a tech bro out of (until a few years ago) California - the literal last thing the world needs is another white guy with ideas about how to “fix” Africa - but there’s gotta be someone who can originate it demand-side, and articulate what they need rather than the other way around.
For one, I’m absolutely certain I could teach any motivated human on this planet the basics of how to make technically-adept content for YouTube inside six months, modulo someone to help with language barriers… but I haven’t the slightest clue how to identify someone in Nigeria (to keep running with this example) with the charisma and cadence and delivery chops and other local knowledge necessary to appeal to a Nigerian audience… hell, I barely know what language(s) that person should speak, never mind what music genre and tracks and cultural references they should use to get the attention of the age group they’d want to target. I mean I know how to choose a Nigerian presenter that would appeal to (and, perversely, not threaten) me and mine, but that’s not exactly the same thing, even when we’re talking about something as “universal” as CS instructional content.
But, hey, anyone in Africa (or anywhere with similar issues, really) with solid and serious ideas on how to fix that, hit me up.
Huh I'm always skeptical of anything that goes viral on twitter, usually theyre marketing campaigns looking at individual youtube videos most dont even have ~25k views. Personally, this looks like a lazy attempt at fake advertising when nigeria does not even have a large twitter presence. As most nigerian are not even online. Nor are they at the level of educational attainment to find anything of remote value from these lectures. The responses in this post as well point towards fraudulent and fake responses in the hopes of generating likely funding for a technology sector in a region that isn't ready for one.
I'll certainly check out some of the channels here but my experience with Indian-originating tutorials on YouTube is that they are almost uniformly terrible. There are a few good ones but so rare that if I see an Indian name I'll tend to skip over it.
My interests when I look tend to be about computing and mathematics. They may be of higher quality on other subjects.
The article discusses NPTEL. For most of its duration, NPTEL videos were straight recordings of classroom videos at India's top IITs.
I have found their rigor & instruction to be at part with MIT OCW (which was and still is the gold standard). The instruction style is more mono tone and less interactive. But, neither are negatives when watching a youtube video.
For me it's been the opposite. Though in my case it's been more Indian professors who often give extremely good (though sometimes perhaps a bit dry) lectures.
Depends on what you call quality. Production value or content.
In my experience it's true a lot of Indian content can improve their video / audio or script proofreading... But they compensate being the only ones sharing their knowledge in niche areas for free. And when you're lost, I take whatever I have at hand even if it's on 240p with a potato mic, I'm grateful it exists.
Sometimes the audio recording is so bad it hinders comprehension. Generally, this is the case where the camera mic is used, and it's set up on the other side of the room. I've encountered a number of tutorial-style videos where there's so much room noise, it's an effort to distinguish the syllables being spoken.
Perhaps it is. But if Indian YouTubers instead produced exceptional quality videos, I'd be as attracted to an Indian name popping up as I am discouraged by one now. I don't believe my discrimination is based on race, therefore isn't racism.
Names are often associated with various groups, trying to imply that this doesn't matter is disingenuous. I will continue to avoid low-quality tech content through various heuristics.
This particular heuristic will — self evidently, based on the article — result in you both missing a lot of very high-quality tech content while also revealing yourself as more than a bit bigoted… but hey, more power to you.
It will result in me missing mostly garbage content, occasionally quality content, in exchange for other content which is quality that isn't missed. It's not like I just decided one day to avoid content in this fashion, you waste enough time and energy on low quality content and you eventually learn ways to notice and avoid it.
> revealing yourself as more than a bit bigoted
A man was once asked by a journalist whether or not he was a "racist". His response was as follows, "I've thought about that a lot, and I've concluded that it doesn't matter."
I think it's time to start ignoring the accusation of "bigotry" and other similar freely-used now-meaningless terms. If recognizing a trend and acting on it in a way that improves my life makes me a "bigot" then the definition of "bigot" is to be reasonable.
Really impressed that you went to “not all psychopaths” there… the particular psychopath I’m recalling was, you know, an admitted multiple murderer, but sure, there’s psychopaths that aren’t. There’s also racists who aren’t assholes… I’m just not sure why they feel the need to cling to their right to be the former so hard they keep edging closer and closer to the latter.
It is a shame that for the longest time, MIT OCW's CS & Math lectures were the only high-quality course-videos on the internet. Leveraging the IIT system to get 100% course coverage over every possible engineering course is genius. It has democratized education better and earlier than any unicorn. (Coursera, Udacity, etc.). Everything from assignments to
It is an indictment of American universities that despite having the funding, resources and technology to do this a lot earlier, it was eventually a 3rd world country that pulled it off. All without any profit motive. "Non-profit universities"?, yeah right. Just take a look at the courses on offer - https://nptel.ac.in/courses . Every imaginable engineering specialization is on offer, down to the most obscure.
University classrooms do not seem to have caught up to just how easy it is to access this information. Most professors do not have the same caliber or flair to do better than the best NPTEL video. I would like to new coursework designed around integrating video lectures of this sort as integral part of engineering instruction, with the in-person professor taking on the role or assistance and disambiguation, rather than being an inferior repeater of the same information.