Hacker News new | past | comments | ask | show | jobs | submit login

I feel like in the future, there needs to be a move away from teacher based education to some sort autodidactism. The current models waste so much time.



I'm working on a project just like that. My MVP is going to be an entire course in an undergraduate college course. It will be the full course, nothing skipped or watered down like you see with most MOOCs these days. I am writing the textbook, study notes, practice problems (fully adaptive with analytics to help the student) and quizzes/exams. Everything is focused on the concept of "have you mastered this topic/skill/subject?". There is also going to be an interactive lecture system which basically works by the student watching some high-quality videos on the subject followed by doing practice problems or simulation exercises.

There is also more advanced stuff you can do with the concept like an intelligent tutoring system or adding in features to the lectures where students can ask the "teacher" questions in real time. I have already built a very basic prototype of it and it definitely works. Just imagine that. You're watching a lecture, learning some topic and then it just isn't clicking for you so you can ask a question immediately or ask for clarification on some subject. I am not talking about typing your question into some text box. I am talking about having a head set on where you can ask your question with your voice and then being able to get a response back that really solidifies the concept for you or makes it crystal clear.

This is all possible with existing technology. Nothing new needs to be invented, it simply needs to be built. Most MOOCs/online courses are heavily watered down because the workload required to build these types of systems is immense but I think the potential is huge in this area. From what I've seen, the young people definitely want to learn online, at their own pace and have access to high-quality learning platforms. I am also planning on offering the entire courses for free. Not really sure how I'll monetize it. Maybe donations?


I’m intrigued. Email me. My email is in my profile.


Even trying to learn autodidactally, there's still need for a teacher. Especially if you can work in smaller class sizes where you can ask for personal help and clarification. It's nice to have a teacher work with you in person over problems and other things you're unsure about. Stuff that a one-size-fits-all computer program (or educational system, like we have) can't do.

I've noticed that when trying to teach myself advanced math. I wish I had a teacher for those topics, where I could ask questions and just make sure I'm understanding things correctly.

Also, you'd have to address on whether they need to do this at home, or if there'd still be a school building they'll need to be at each day. If the former, what do you do for those whose parents aren't home during the day, or who sometimes only get to eat the two meals they're provided at school? How will you deal with kids interacting with each other, and making sure they're not isolated, etc. etc. There's a lot more to think about than just "Online == less time wasting" (and even then, I think it highly depends on the student)


The first two questions can be solved through software. I've noticed that most people heavily underestimate what an algorithm is capable of doing. I know most will not believe this but an algorithm can do the same things as a human teacher. In fact, I would even say that in the future the best teachers in the world will be algorithms, not teachers.

Also think about how the whole teacher education model currently works. Students go to a lecture with 30 other kids and then the teacher presents the material, explains it and so on. Students ask questions during the lecture but there is very little actual one on one tutoring going on. If the students want one on one time then they will need to go to tutoring center (inconvenient) or teacher office hours. Even then, during those hours you have to hope that the teacher isn't already occupied by another student. Then think about this issues on a global scale. Do you think Zimbabwe or Ethiopia have competent teachers? very few. So most students get nothing at all. They are completely left out. If a global education system is developed with algorithms being the teachers/tutors then that would be a huge revolution. Every student in the world would get an equal shot at having a high-quality education.

For the last question. That is a complete separate topic. Kids can get social interaction by a variety of other areas, such as, going to the park, playing with other kids in the neighborhood and so on.


I seriously doubt the first two questions can be solved via software, unless you want students to give up privacy in their own homes. There's nothing to prevent cheating and the student from using the internet or else.

As for meals, if you think software will cure all societal ills, then you might be the most naïve person alive. There are a lot of things software can do, but provide two free meals a day to students is not necessarily one of those things. Especially when getting meals to them outside the school year is more a political issue than a scholarship one.

As for your argument that an algorithm can do the same thing as a teacher, well, I'll believe it when I see it.

And I'm not arguing that the model we currently have is great. There's many ways it could be improved and fixed; but turning everything over to software is not the answer. Especially because kids need less interaction with software as it is, and more with each other.

Which brings me to the last topic... You're coming from an extremely privileged view if you think kids can get a lot of social interaction otherwise, especially when younger. Try living in a rural area without a car or access to someone to drive you around. Or in a neighborhood without many other kids. You're basically just stuck at home, with no interaction outside family. Which means even more screen-time, which means kids will be even more addicted and have even more problems with empathy and other skills that have to be learned from interaction.

Software can't solve every problem, and there's some that it shouldn't try to completely solve anyway. It can definitely be a great tool to help students learn, but it shouldn't replace classroom interaction, if only because of the social aspect of it and to decrease reliance on mobile devices and electronics.


You're right that software can't solve every problem but for education, it can solve a great deal. I know it is hard to believe what I am talking about but it is true. Software, when designed correctly, is an extremely powerful and amazing tool to use. Even if there was an all-knowing AI teacher there would still be schools and teachers. Those facilities are paid for by tax dollars so as long as the state keeps funding it, then it will continue to exist, even if students can get a superior education online.

I see these tools mainly being used by students that want to self-study a topic that is not available to them. For example, high school kids that want to maybe take college classes while they are in high school. It is also great for students in other countries that don't have access to high-quality educational resources (teachers, books, tutors etc.).


Sure. Self-taught systems put a lot on responsibility on the student. Apprenticeship systems are very mentor-time-expensive and/or put responsibility on the student.

If the students were already so great at this, they could make do with so-so teachers. We do not expect as much from K-6 at least, for good reason. And that will not change soon.


> If the students were already so great at this, they could make do with so-so teachers.

It wastes the student's time tho.


The hard truth is not everyone can do that. The current models are broken, but something like them is necessary. I think what you really reach for here is Teachers as mentors when the door is open. We used to allow for that. We don't very much today.


I honestly think the best way might be the method advocated by Sal Kahn. [1] Basically, instead of students going over a lesson in class then taking home work to do to master it, they reverse that. They watch the lesson at home the night before, answer a few questions as an "exit slip" and then sleep on it. The next day they go to class and work on problems/ask questions with the teacher present to help them.

While it does have its issues, such as making sure students actually watch the videos, I think this is the best approach for certain subjects, especially math. It allows the students to do the problems where they can get feedback and guidance, instead of struggling with something they're failing to comprehend, which just leads to demotivating them. It also allows them some time to come up with questions related to the topic that might further their learning. There will likely be abuse from students who don't pay attention, and there are plenty of other issues (internet access being a huge one), but I could see some hope in this method.

[1] https://blog.blackboard.com/why-salman-khan-thinks-students-... is a nice write-up about it.


That method is far superior to the current method but teachers don't like it because it takes "resources" away from the teacher. That's the argument I've heard on other forums when discussing the flipping the classroom method. The teachers say that it is better if they explain the material in the classroom and answer questions immediately. So expect to get quite a bit of push back on the flipping the classroom method. That's the reality. I've noticed that there are quite a few teachers that are basically fearful of some new technology or method popping up that replaces them. Education has become far too political from what I've seen.


They have to know the material with that method.

I have a great career record in training. This is the method I use, but compressed a bit, due to shorter learning times.

The teachers employing this method will need to have very fully explored it. Then, when students see them, they are ready for all those interactions.

The difference is experience in guiding people through material, vs assisting them in applying and exploring it.

Frankly, guidance can be replaced.

Teachers who embrace the method, amd who make the investments will be invaluable. Each class they do improves them as much as it does students. Over time, there will be very little they cannot help students with.

Everyone gets real mastery.


Yeah, I've heard that argument as well, but I think it fails. As a math teacher myself, I would love it if kids would go read about the lecture at home and then we could work problems the next day. I try to do that anyway, but it's difficult when I've got a lot of material to cover and can't give them a day to practice after every lecture. Not to mention that a lot of kids don't know what they don't understand until they attempt to do it


Those attempts do three very important things:

Clarify poor understanding.

Result in awesome (for them) questions. They will be the ones they need, and often the more aggressive people will ask want type questions, seeking. All high value.

Educator will be challenged more deeply and broadly over time as they see better questions over time. New ones require effort, and that brings both greater mastery of the topic and how best to frame it up for people about to jump in and start understanding.

Should educators analyze those questions, they can augment all this by a quick "chalk talk" prep session to help all this along, stimulate more and better thinking.

This has been standard for me in various education roles over the years. It is rough the first few iterations, then very rewarding for all involved after that.

Turns out people, and their diverse backgounds and experiences will both approach things in surprising ways, as well as think in diverse, at times, equally surprising ways.

Understanding that better, as well as having some broad experience seen over time, very seriously improved my ability to identify with a given student and have a more accurate mental map of where they are. I can then seek to fill in gaps, or entertain extensions (their seeking) as time and their desire permit.

Worth it. Not the easiest way, but is my goto every time.


I completely agree. I try to do it as well, but my biggest issue with it ends up being time constraints and the teacher I mentioned in another post who wants us to all be doing the same thing on the same day (said teacher also prides herself on how fast she covers the material... Kids don't like it as she never reviews or spends more than a day on any topic)


Just revisited this. There should be topics and time.

How that time gets used is really driven by students. They get there, or they do not.

And, doing all this is expensive human time. They really want, need and should get there.

One thing I get to do, as my role is rarely formal education, is prepare 200 percent topic agendas. Prep as if I have Feynmans in the room. Look out! Here they come. :D

(And I am always hoping, because when that happens I myself am definitely going to learn something. Often pain in the ass learning, but it keeps my blood pumping, leaving nothing but getting after my own understanding.)

Then, assign priorities.

Basics get done no matter what. I do not want fails. Just quietly drop extras.

From there, I am highly likely to be golden. Usually, someone asks something new. Good. I got the prep done solid, so I research that one, add to the library of stuff I have ready.

Then, I coach them, they tackle stuff, wash, rinse repeat. I will, depending on material, also show a solution or few, again depending. Discuss, on to the next.

Over time, I get super good at whatever it is, and students can generally max out, getting all they can.

When I can, I set a chunk of the time aside as open time. People can fill a gap with that. Or explore, seek. Whatever. I am there to assist and challenge as they need.

This is a challenging way to do the work, but it works well. Often, I am doing this for people who just took jobs, or who have some requirement. Fails are painful. That is why I do it this way.

Just some context for my comments above.

There is a bonus. Once I have prepped and delivered a few times, I can often do the course by rote. My own mastery ends up solid, and I've got a ton of context, accumulated one student at a time, to work from when identifying where people new to me are, and what they are likely to need. Those go into the chalk talk, which almost always has custom bits tuned for the group at hand.

And that is tools, not solutions. They do that, picking up the reference material, exercizes, and the bits from the coaching talk.

Competency tends to come reasonably quickly. Then we do fun stuff, entertain any seeking, special needs, wants as time allows.


Too bad the students are not at the same frame of mind each day.


Yeah, for sure. I think it is a great method as well. When I was in college, I had a calculus teacher teach using the flipping the classroom method and it worked amazing. Probably one of the best math classes I have ever taken. At the time I had a serious "math-phobia" and calculus was something I had previously struggled in.

The teacher basically just made youtube videos and explained the concepts. If you didn't understand something you could just re-watch the video. After the video another set of videos explained problems and how to solve them.


I am a fan of that method.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: