Why is it that, sixty years after PLATO, we don't have automated teaching programs that really work? By now, the hardware is cheap, everything up to and including VR can be easily delivered, and machines can have semi-intelligent conversations.
Duolingo, probably the most widely used teaching program, is designed around motivating the user to come back every day to the point where learning takes a bit of a back seat, probably because a too-demanding curriculum would decrease engagement.
On the other end of the motivation spectrum, a textbook with exercises and solutions is a teaching technology that really works, with hardly more interactivity required than flipping to the back to check your solution. Of course nowadays you can also ask an LLM to critique your understanding and try to pinpoint exactly where you got something wrong.
Decreasing the effort required to learn something is probably hard (though quality-of-life improvements like making information more readily available are certainly possible) and so is motivating people to exert more effort, but there might be some value in establishing exactly how much effort a specific student is willing to expend, and then providing them with the most useful exercises that don't exceed that level.
That’s wrong.
Duolingo sucks. So does the conventional textbook approach. People can go through years of the conventional textbook approach and learn nothing. I certainly did.
The litmus test for learning isn’t how much effort it requires and how much it demotivates you.
The school of thought where language input with sufficient comprehension is emphasized has already worked for all those people who learned English as an nth language by playing video games, watching movies and Netflix, and without grinding through books.
And compared to that? Duolingo looks much more like the conventional approach if you look past the insipid gamification and toddler-level "the boy drinks water and sits on an apple".
Actually, Duolingo as a technology isn't necessarily that bad either, it's just Duolingo-the-company with its business incentives that limit its potential.
And of course not everything that requires effort is necessarily a good way to learn, but I don't think there's any way to learn that's also effortless.
E.g. the least-effort way to play a video game in a foreign language is to figure out how to operate the interface through trial-and-error (or watching someone else play) and then rely on spatial memory to press the right buttons without having to read the labels. And people who watch foreign-language movies or shows typically do so with subtitles that free them from having to learn anything, so they might just pick up the names of characters and maybe a few common expressions.
The comprehensible input approach requires a lot of effort upfront to make the input comprehensible in the first place (even if it's toddler-level content, that still means it targets people with a vocabulary of several thousand words) and then you still need to push yourself a bit to actually learn something, rather than mentally skipping over the parts you didn't understand.
Clearly that’s not the case when multi-year streakers don’t have a clue what to do.
They followed the instructions of streaking and didn’t get anywhere. The ones I have seen anyway (YouTube).
> And of course not everything that requires effort is necessarily a good way to learn, but I don't think there's any way to learn that's also effortless.
No one said effortless. You can’t bolster your side by making up that someone marked a place on the complete opposite extreme of the spectrum.
> E.g. the least-effort way to play a video game in a foreign language is to figure out how to operate the interface through trial-and-error (or watching someone else play) and then rely on spatial memory to press the right buttons without having to read the labels. And people who watch foreign-language movies or shows typically do so with subtitles that free them from having to learn anything, so they might just pick up the names of characters and maybe a few common expressions.
I don’t know why you are setting up this armchair theory. I said what has worked for people. Not in theory. IN REALITY. People who have immersed enough time into English indirectly by playing video games, watching Netflix and whatnot. And also seeing no real correlation with their practical, day-to-day abilities and their formal studies of English that they have to go through in school.
You can see this every time someone goes from relatively passive learning through single-player video games to participating in voice chat in multi-player video games and discussing things in English online. Do you think they stop themselves half-way through this process and grind grammar/whatever other English books in order to work up to talking to people in real life?
> The comprehensible input approach requires a lot of effort upfront to make the input comprehensible in the first place (even if it's toddler-level content, that still means it targets people with a vocabulary of several thousand words) and then you still need to push yourself a bit to actually learn something, rather than mentally skipping over the parts you didn't understand.
What it requires is a lot of time and exposure. Where your “actually learn” weasel phrase fits in is unclear since all that seems required is sufficient interest in the associated topics (not the language per se).
There's a long history of technology which promised to revolutionize education - records, radio, film reels, TV broadcasts, VCR, microcomputer software, the constructionist faith in "One Laptop Per Child", MOOCs, and more.
Add this to the pile.
There are always some students who can learn on their own, and who will, for example, learn Latin from a book at the age of 6.
Most kids are not thing way.
I think most kids need someone to be there for the long-term emotional connection. Ideally (in my view), someone who respects them, and encourages them to learn, which the student can reciprocate by success in class and personal growth.
Technology cannot make that connection.
I don't think revolutionize is particularly on the horizon, but "doesn't really work" really depends on what you're defining your success condition to be.
Including a copy of a book like "Introduction to Latin".
I have to be hand-wavy because the full analysis has to include factors like the cost of licensing, the cost of hardware and maintenance, the time to integrate into the curriculum, how often the UI and content changes, and the loss of space for and access to other resources.
Like, our local elementary school got rid of its staff librarian to have money for more student electronics, and now they are talking about bringing a librarian back.
I've only ever seen PLATO-based systems used at the college level, which is where we also expect students to be more self-motivated. That would be another factor.
I don't think I had any teacher who only talked in front of us. Not only did that do more than that in class, and via marked papers, we also had teachers who helped out in school clubs, and attended bake sales, and set up the room for parent visits, and lead field trips, and organized sports days and fire drills. One teacher went the same church as my family, and I bumped into teachers a few times around town.
1 to 30 is higher than average for most of the US. https://nces.ed.gov/surveys/ntps/tables/ntps1718_fltable06_t... and https://nces.ed.gov/surveys/ntps/estable/table/ntps/ntps2021... tell me it's more like 1:25.
I've read many many anecdotes over the years of Europeans learning or perfecting their English language skills playing classic point and click adventure games that were either unavailable localised or the localisation was bad. It feels like that's an interesting language learning entry point.
I also like the idea of an AI plugged into a glasses video camera just narrating what you're doing in your target language.
Also I wouldn’t trust the current crop of “AI” software to generate accurate VR images, say of the workings of a steam engine or mechanical clock.