Other Can AI Assistants Improve Online Learning?

[jedishrfu]

This wired article talks about what you can and can’t learn online comparing it to a chocolate chip cookie:

You Can Learn Everything Online Except for the Things You Can't

https://www.wired.com/story/you-can-learn-everything-online-except-for-the-things-you-cant/

 

[phinds]

Good article. I agree w/ him.

 

[symbolipoint]

This wired article talks about what you can and can’t learn online comparing it to a chocolate chip cookie:

You Can Learn Everything Online Except for the Things You Can't

https://www.wired.com/story/you-can-learn-everything-online-except-for-the-things-you-cant/

Good Article!

My usual comment about any online learning is that the online situation does not give you a true laboratory section. Some types of courses are missing too much if no laboratory section is included.

 

[jedishrfu]

It’s somewhat like the problem with SAT style tests where they can only test for certain types of math problems, the ones that make good multiple choice questions.

 

[Qurks]

The only downside to online learning is the people may be less invested, either economically or socially so they stop doing it. If you need equipment then obviously that's an issue.

 

[StoneTemplePython]

qtd from the article:

College is about taking classes that might not have anything to do with work. Art history is a great class—even if you aren't going to work in a museum. Algebra should be taken by all students—even though you probably won't need it (most humans get by just fine without a solid math background).

This seems like a false equivalence / sloppy thinking to me. I got almost nothing out of an art history class I took when in college and I don't know of a single deep insight someone has had from art history. (And I've looked...)

On the other hand, I don't feel that way about mathematics, especially high school level algebra which is the kind of 'algebra' that I think is being mentioned here. As I've said here in the past, and setting aside the workplace, I think basic numeracy is civically important whether in the voting booth or while serving on a jury. I don't feel that way about Art history. My sense is the author of that article would be rather uninspiring to have as a fellow juror since clear thinking and basic numeracy quite important.

 

[jedishrfu]

qtd from the article:
This seems like a false equivalence / sloppy thinking to me. I got almost nothing out of an art history class I took when in college and I don't know of a single deep insight someone has had from art history. (And I've looked...)
.

Its good you liked math, but you haven't lived long enough or spent the time to look at art to realize that much of art is an advanced form of mathematics working with patterns of light, color and nuance. Its influence on your being may be very subtle. The problem for schools is they need to give you an education but they don't know what you'll become because you can't yet know unless you're exposed to these different areas or unless you live in a society where your parents have decided on your career.

I too had some bad experiences in English class in school where we were forced to read various authors like Shakespeare, Jack London, Mark Twain, and Edgar Allan Poe. The funny thing is though as much as I hated reading them and being tested on them, they still stick in my mind as yardsticks and I can begin to see just how terrible some writers are at story telling.

I remember liking Ernest Hemingway's Old Man and the Sea. It was only recently that I learned that he was strongly influenced by Cezanne and his cubist art and wanted to write like Cezanne painted. I also remember liking the movie Forbidden Planet and finding out later that it was based on a Shakespearean play The Tempest. There are many sci-fi stories whose authors were influenced by earlier authors and by art but you never hear about it but can see it in their writing if you know what to look for or have read the earlier works.

I think the one thing that made it hard to appreciate the arts was I could never figure out how to get a good grade. In math, you learn the rules of the game, play the game and get rewarded with good grades. When a teacher said you did something wrong you could look back on your work and realize where and why. In english, my writing would get constantly red pencilled and I never knew why or how to fix it. Now of course, I write better (no more redpencil right? darn the redpencil got underlined in red as I write it).

In high school, I had a really tough art teacher who told us outright if you want an A then your art will need to match mine. I got B's so for me that was close enough. In college I took an art course where we were to make a painting by dabbing a sponge of paint on paper. I made a starship and the teacher frowned and I decided then and there that art wasn't for me. Now I realize I should have been more open, more free and not so literal in my thinking.

That's what art can teach you: to feel and not be so literal in your thinking and perhaps that why its hard to accept what the article author has said.

 

[StoneTemplePython]

I think the one thing that made it hard to appreciate the arts was I could never figure out how to get a good grade. In math, you learn the rules of the game, play the game and get rewarded with good grades. When a teacher said you did something wrong you could look back on your work and realize where and why. In english, my writing would get constantly red pencilled and I never knew why or how to fix it. Now of course, I write better (no more redpencil right? darn the redpencil got underlined in red as I write it).

I can relate to this -- Art history seemed to consist of memorizing a bunch of facts about famous works and not much else. I thought this is very strange, so I asked people majoring in it and they told me that's what you do. I have a good memory so I memorized a bunch of stuff and got an A but not much else out of it.

- - - -

Its good you liked math, but you haven't lived long enough or spent the time to look at art to realize that much of art is an advanced form of mathematics working with patterns of light, color and nuance. Its influence on your being may be very subtle. The problem for schools is they need to give you an education but they don't know what you'll become because you can't yet know unless you're exposed to these different areas or unless you live in a society where your parents have decided on your career...

That's what art can teach you: to feel and not be so literal in your thinking and perhaps that why its hard to accept what the article author has said.

I think what you've written here is quite thoughtful, though missing the mark. But that's the nature of this -- we're very much in opinion not fact territory here.

As a second point I'd mention that my role model here has long been Ben Franklin, not a mathematician or scientist per se. He's perhaps the most high dimensionally talented and accomplished person I can think of. My sense is he'd approve of communication courses -- even with red ink -- but again not art history. It seems that you and the wired author have one view, whereas (Ben and) I think differently. (I can drop in some Mungerisms here too -- he has much more radical views though the gist is similar.)

A third and final point -- I'm not sure how one estimates the general (read: not narrowly domain specific) insights from studying art -- or more importantly, falsifies 'bad' lessons learned. It's not that everything that's 'true' has evidence it's just that everything immune to evidence is on very precarious ground. Very few people learn to appreciate the difference.

(A special exemption is "evidence" and math though for purposes here it should be interpreted as a logical construct and that's it.)

- - - -
n.b. I have had several variations of this conversation with friend who has formally studied art and to varying degrees practiced in many countries. On multiple occasions the conversation has ended with him telling me that I should at least learn to play an instrument so that I can appreciate art, and then me reminding him that (a) he has a bad memory and we've had this conversation before and (b) I played a couple different instruments over 8+ years.

 

[ZapperZ]

I have a different take on this. First of all, I follow Rhett Allain's article frequently and I find his point of view is often interesting. Secondly, while I have not been involved in a purely online class, I have taught blended classes where a substantial portion of the lesson is done online. So I have a flavor of the issues surrounding online lessons and classes.

While I agree with Allain's take that an "education" (not simply a class lesson) involves more than just the material, I also think that the issue of online education and online lessons have problems directly, not just esoterically. The past many years, there have been numerous research on physics education that showed that passive learning that many of us are used to is not the most efficient and most effective means of learning, especially for non-science majors and non-physical science majors. So there are efforts to change how we present the material to the students which involve a significant portion of active learning, and there are research to back a number of different pedagogical and methodological approach to this.

I'll give you one example via Eric Mazur's "https://www.uq.edu.au/teach/flipped-classroom/docs/FAB/FABPeerInstructionTipsheet.pdf" technique. Here, the instructor becomes almost as a coach, directing students into the correct understanding of the concepts. But it is the students who are actively participating in the class where they are not simply sitting and listening passively. A lot of studies have been done to show that this method is more effective than passive learning, and a lot of schools are adopting this technique at the General Physics course level.

This active learning is what can't be done (at least, not easily) with online lessons or classes. Inevitably, online lessons involve someone sitting in front of a PC or electronic device, often alone, and either reading or watching a video. Even in an online class where there is live interaction with the instructor and/or students, active learning that is required in Mazur's Peer Instruction is simply tedious and slow to do.

So, while online learning is new and makes use of the latest technology in teaching, its philosophy and methodology are actually quite old, because it is still essentially a passive learning. It means that the students are not learning the material in the most efficient and effective manner when compared to those who are engaged in active learning. The students are not just handicapped for not having the "college experience" and a broad education. They are also hampered with an old and inefficient way of learning.

Zz.

 

[Tosh5457]

I have a different take on this. First of all, I follow Rhett Allain's article frequently and I find his point of view is often interesting. Secondly, while I have not been involved in a purely online class, I have taught blended classes where a substantial portion of the lesson is done online. So I have a flavor of the issues surrounding online lessons and classes.

While I agree with Allain's take that an "education" (not simply a class lesson) involves more than just the material, I also think that the issue of online education and online lessons have problems directly, not just esoterically. The past many years, there have been numerous research on physics education that showed that passive learning that many of us are used to is not the most efficient and most effective means of learning, especially for non-science majors and non-physical science majors. So there are efforts to change how we present the material to the students which involve a significant portion of active learning, and there are research to back a number of different pedagogical and methodological approach to this.

I'll give you one example via Eric Mazur's "https://www.uq.edu.au/teach/flipped-classroom/docs/FAB/FABPeerInstructionTipsheet.pdf" technique. Here, the instructor becomes almost as a coach, directing students into the correct understanding of the concepts. But it is the students who are actively participating in the class where they are not simply sitting and listening passively. A lot of studies have been done to show that this method is more effective than passive learning, and a lot of schools are adopting this technique at the General Physics course level.

This active learning is what can't be done (at least, not easily) with online lessons or classes. Inevitably, online lessons involve someone sitting in front of a PC or electronic device, often alone, and either reading or watching a video. Even in an online class where there is live interaction with the instructor and/or students, active learning that is required in Mazur's Peer Instruction is simply tedious and slow to do.

So, while online learning is new and makes use of the latest technology in teaching, its philosophy and methodology are actually quite old, because it is still essentially a passive learning. It means that the students are not learning the material in the most efficient and effective manner when compared to those who are engaged in active learning. The students are not just handicapped for not having the "college experience" and a broad education. They are also hampered with an old and inefficient way of learning.

Zz.

I find that online classes, at least in programming, are much more interactive than it'd be possible in a physical university, and I'm sure that the same could be done for other subjects. After a theoretical presentation, the course can make you complete some exercises to make sure that you've understood the material, automatically check if it's correct or not, and give you suggestions on how to make it right. This is the sort of thing that just won't happen in a physical class because the teacher won't have much time to spend with any specific student.

 

[ZapperZ]

I find that online classes, at least in programming, are much more interactive than it'd be possible in a physical university, and I'm sure that the same could be done for other subjects. After a theoretical presentation, the course can make you complete some exercises to make sure that you've understood the material, automatically check if it's correct or not, and give you suggestions on how to make it right. This is the sort of thing that just won't happen in a physical class because the teacher won't have much time to spend with any specific student.

Interactive learning is not the same as active learning. Read the example of peer instruction method that I linked to. This can’t be done easily with online classes.

What you described is already done in some form for many classes as part of their lessons,especially with their homework assignments.

Zz.

 

[gleem]

About 40% of my courses in college were from the liberal arts. This was the standard curriculum for science majors. I have been a slow reader all my life perhaps some undiagnosed functional issue. This made it a struggle in the liberal art classes where reading assignments usually where extensive. However at the college level the courses seem to become more interesting perhaps because of increasing maturity. As an adult I have developed an interest in a very wide range of topics which in HS I would never have dreamed. My last course was World literature and by far the best. One reason was that the professor had the knack of conveying his insights of the works of the likes of Milton to Balzac to Dostoevsky.

I get the impression from the inquiries in this forum, from budding physicists seeking advice on curricula that the number of such courses seem to be dwindling in many universities. Are the curricula becoming too vocational. This will be unfortunate for many decisions must be made by citizens which are not of a technical nature. The indigenous city dweller may not have a sense of the value and worth of our natural wonders. He therefor may not appreciate/care about the problems our planet's exploding population will create. So to the overly focus person on science may not be able to participate effectively in solving societies problems and may depend on bias and hearsay to make decisions.

When I graduated my feeling was that all should go to college. Not only for the subjects of personal interest but for a taste of the endeavors of mankind in general particularly and why do what they do and think it is worth doing. Equally important is learning to live and function in a world of new ideas and more diversity which is far removed from the somewhat sheltered environment of our childhood.

I really liked the chocolate chip cookie analogy.

 

[Dr. Courtney]

So, while online learning is new and makes use of the latest technology in teaching, its philosophy and methodology are actually quite old, because it is still essentially a passive learning. It means that the students are not learning the material in the most efficient and effective manner when compared to those who are engaged in active learning. The students are not just handicapped for not having the "college experience" and a broad education. They are also hampered with an old and inefficient way of learning.

Zz.

Not necessarily. My sons have taken a couple of distance learning courses (chemistry and physics) that included laboratories. They bought the kit, they did the labs. It was genuine active learning. I'm putting together resources for a distance learning physical science lab course right now. Students who do the labs (not just watch) will get more benefit than most brick and mortar school students, because the labs are better.

 

[ZapperZ]

Not necessarily. My sons have taken a couple of distance learning courses (chemistry and physics) that included laboratories. They bought the kit, they did the labs. It was genuine active learning. I'm putting together resources for a distance learning physical science lab course right now. Students who do the labs (not just watch) will get more benefit than most brick and mortar school students, because the labs are better.

Yes, I know of lab kits since those were used in my blended courses. Still, that is not the "active learning" that I was referring to.

Zz.

 

[Stephen Tashi]

It's interesting to browse a list of autodidacts https://en.wikipedia.org/wiki/List_of_autodidacts and speculate whether the internet will lead to great expansion in prominent autodidacts or whether they will be produced at about the same rate.

The qualities of the student are far more important that the qualities of the course or the instructor. I think the pros and cons of techniques in education are important for teaching "the average student" or, in the case of technical material, "the average intelligent student". The brilliant students and the lost souls seem to reach outcomes independently of how they are instructed.

 

[Vanadium 50]

speculate whether the internet will lead to great expansion in prominent autodidacts or whether they will be produced at about the same rate.

It doesn't seem to have many modern examples in science, and some of those have PhDs. So I'm not so certain we have a good handle on the baseline.

 

[Dr. Courtney]

Yes, I know of lab kits since those were used in my blended courses. Still, that is not the "active learning" that I was referring to.
Zz.

So, while online learning is new and makes use of the latest technology in teaching, its philosophy and methodology are actually quite old, because it is still essentially a passive learning. [emphasis added] It means that the students are not learning the material in the most efficient and effective manner when compared to those who are engaged in active learning. The students are not just handicapped for not having the "college experience" and a broad education. They are also hampered with an old and inefficient way of learning.

Zz.

I see. You have a false dichotomy between your narrow definition of "active learning" and less effective "passing learning" which is an "old and inefficient" way of learning. I strongly disagree. Active means engaged and participatory. That active participation can take many forms. Passive means being more of a viewer (like watching lectures or videos without active participation.)

My experience is that many students learn very well through distance learning - and it is often more efficient and effective for several reasons:
1. Costs and time and inefficiencies associated with commuting to class are often eliminated or greatly reduced. Many times, the "in person" experience simply is not enough better to justify the costs and time of commuting.
2. Accountability tends to be better in online courses - both the professor accountability toward the students and the university/department/system's accountability toward the professor ensuring course credit is not being gifted.
3. Online courses leverage technology more to provide assessments and feedback.

I don't consider online learning to be a panacea, but for some students and some courses it can be just as good or better than the average "in person" classroom experience. The students I mentor tend to average 1-2 online courses each semester for various reasons - usually related to difficulties getting all the courses they want to take at a time and location convenient to their schedule. These courses have included Composition, Calculus, Chemistry, Physics, World History, and Philosophy. The students have done very well both in their distance learning courses and in downstream courses where mastering the material in the online courses was critical. Online learning can be active enough for learning to be complete, efficient, and effective.

Of course, students need to be self-disciplined and committed to excellence and timeliness in completing all the assignments. Students who struggle in distance learning courses tend not to be disciplined in completing assignments.

My own experience and that of most students I mentor is that real learning hardly ever happens in class anyway, especially in physics, math, and chemistry courses. Most of the important real learning happens when doing the homework.

The qualities of the student are far more important that the qualities of the course or the instructor. I think the pros and cons of techniques in education are important for teaching "the average student" or, in the case of technical material, "the average intelligent student". The brilliant students and the lost souls seem to reach outcomes independently of how they are instructed.

Yep.

 

[ZapperZ]

I see. You have a false dichotomy between your narrow definition of "active learning" and less effective "passing learning" which is an "old and inefficient" way of learning. I strongly disagree. Active means engaged and participatory. That active participation can take many forms. Passive means being more of a viewer (like watching lectures or videos without active participation.)

My experience is that many students learn very well through distance learning - and it is often more efficient and effective for several reasons:
1. Costs and time and inefficiencies associated with commuting to class are often eliminated or greatly reduced. Many times, the "in person" experience simply is not enough better to justify the costs and time of commuting.
2. Accountability tends to be better in online courses - both the professor accountability toward the students and the university/department/system's accountability toward the professor ensuring course credit is not being gifted.
3. Online courses leverage technology more to provide assessments and feedback.

I don't consider online learning to be a panacea, but for some students and some courses it can be just as good or better than the average "in person" classroom experience. The students I mentor tend to average 1-2 online courses each semester for various reasons - usually related to difficulties getting all the courses they want to take at a time and location convenient to their schedule. These courses have included Composition, Calculus, Chemistry, Physics, World History, and Philosophy. The students have done very well both in their distance learning courses and in downstream courses where mastering the material in the online courses was critical. Online learning can be active enough for learning to be complete, efficient, and effective.

Of course, students need to be self-disciplined and committed to excellence and timeliness in completing all the assignments. Students who struggle in distance learning courses tend not to be disciplined in completing assignments.

My own experience and that of most students I mentor is that real learning hardly ever happens in class anyway, especially in physics, math, and chemistry courses. Most of the important real learning happens when doing the homework.

You are missing the point. I never claim that students can't learn anything via online classes. I also do not claim that for many students, this is not the most viable means to get credit for certain courses. Those are not the issues!

And yes, I have a clear and narrow definition of "active" and "passive" learning, and those are the ones I was comparing with. Can active learning be done online? Sure! I've done it, but it was with a very small online class, not a class of 20-40 online simultaneously. This is what can be done easier in-person.

While what Allain's stated can't be remedied with more technology, I am not that fatalistic, and I am sure that there will eventually be technology that will enable many of the active learning techniques that we have employed in class to be available to online classes. I just do not think we currently have that.

Zz.

 

[e-pie]

Nothing can fill the places of the good old book. But I do encourage to learn new topics online through authentic sources.

And to watch videos of topics like Orbital theory, s p d f orbital shapes etc. Solid state. The best method to learn these topics is to visualise the 3 d structure. The books lack this feature. Not that nobody knew these theories before the internet, but modern generation students needs the visual and olfactory approach.

 

[gleem]

While what Allain's stated can't be remedied with more technology, I am not that fatalistic, and I am sure that there will eventually be technology that will enable many of the active learning techniques that we have employed in class to be available to online classes. I just do not think we currently have that.

This might be a good application for AI.

 

[Dr. Courtney]

You are missing the point. I never claim that students can't learn anything via online classes. I also do not claim that for many students, this is not the most viable means to get credit for certain courses. Those are not the issues!

And yes, I have a clear and narrow definition of "active" and "passive" learning, and those are the ones I was comparing with. Can active learning be done online? Sure! I've done it, but it was with a very small online class, not a class of 20-40 online simultaneously. This is what can be done easier in-person.

While what Allain's stated can't be remedied with more technology, I am not that fatalistic, and I am sure that there will eventually be technology that will enable many of the active learning techniques that we have employed in class to be available to online classes. I just do not think we currently have that.

Zz.

Perhaps our difference of opinion stems from differing experiences. My experience and that of students I mentor is that "peer instruction" is a noble goal, but more likely a relatively inefficient use of time as "the blind leading the blind." Students I mentor report that the most likely number of students as interested and engaged as they are in math and physics classes is 0 or 1. My own experience was similar in college. Most of the time peer instruction is the lazy students trying to ride the coattails of the diligent students (to course points rather than real learning). Most aha moments and flashes of insight occur either when working alone on assignments or when asking an instructor specific questions to get over a hump that was encountered working alone on assignments.

There may well be in person core math and physics and chemistry classes out there (perhaps you are blessed with them) where most of the students in the class are willing to apply themselves to actually learn consistently over a 15 week semester - coming to class having read the assignments and made honest and timely attempts at the homework problems. But I only experienced this as a graduate student at MIT. (In both undergraduate courses and graduate courses.) In most places the majority of students put off taking these courses as long as possibly, lack enthusiasm, and don't do much more than they figure is required to reach their grade goal in the course.

 

[StoneTemplePython]

It's an interesting point... the linked article from Crouch and Mazur said:

Here we present the results of ten years of teaching the two introductory physics courses for nonmajors at Harvard University with one such method

This should set off alarm bells of overfitting. The method may work ok in Boston/Cambridge area and perhaps at some other ultra-elite schools. But the success likely tells quite a bit about the idiosyncracies of Harvard Students (probably smarts and motivation in this case).

- - - -
Is there much any data testing it elsewhere?

 

[gleem]

While what Allain's stated can't be remedied with more technology, I am not that fatalistic, and I am sure that there will eventually be technology that will enable many of the active learning techniques that we have employed in class to be available to online classes. I just do not think we currently have that.

One problem with online information sources is if you have a question or need a clarification of the material. University courses may have someone available to help you out. But for free courses it is probably not available especially if the course is available to anyone at anytime with perhaps thousands of participants is multiple time zones. AI to the rescue. At Georgia Tech Prof. Ashok Goel has developed and is improving an AI teaching assistant.(alias Jill Watson) to answer questions in his knowledge based AI class. Only one student out of 300 was suspicious that it was not a human.

https://pe.gatech.edu/blog/meet-jill-watson-georgia-techs-first-ai-teaching-assistant