hedwards: You're missing the point here. Better technology has little to do with the problem. Current day educational titles aren't any better than the ones were decades ago.
[..]
It's a pipe dream where people keep thinking that it's going to work and yet, there's been very little progress in terms of something that actually works.
clarry: My point here is that I'm not convinced anyone's seriously trying, or at least I haven't seen or heard of the results if someone is. If you're saying current educational games are limited in scope and just reiteration of 90s crummy garbage with better graphics, you're sort of proving my point. There's so much more that could be done and nobody is doing it!
There are many fields where progress comes to a grinding halt (even if a few academics with limited resources are toying with their doodles that will never make any impact) until it resurges again decades later with fresh ideas and a different outlook, along with crucial changes in supporting technology. Of course, financial incentives need to align too. There might be "tons of money" in educational software (if done right) but getting started and getting started big enough to get a piece of that cake might just take a truckload of money. I can't imagine that many institutions today would pay for 90s tier stuff, and you can't just throw a quick prototype / proof of concept together in three months and show it off on kickstarter.
And that's kinda where I imagine the whole thing is right now, basically everybody gave up in the 90s when the "golden era" ended and nobody's doing anything serious right now. And fwiw the bar for serious work today should be much higher just as it is for games. And getting the funding to start anything serious is not there, ergo we
can't have seen anything serious.
This is one of the reasons why we require teachers to have a license and complete relevant coursework. I'm sure that it seems like the bar should be higher than it was back then. But, the games back then were quite a bit better than you're giving them credit for and the level of technical sophistication needed to get good enough is a lot higher than it might seem.
It's only been within the last 10 years or so that we got competent chatbots that could really interact with people like a real person in limited areas. Teaching and education will be one of the last areas where AI programs are able to replace actual humans due to the sheer complexity of the task.
We know from the last century of research as well as personal experience, that learning is social, it's messy and the student is ultimately responsible for figuring out how to own what's being taught. We also know that how precisely lessons go is largely unpredictable. You can have the same lesson and activity go right in one classroom and then fail horribly in the next one with no particular warning signs ahead of time.
Computers do not handle ambiguity very well and that isn't likely to change any time soon. Sure, AI is getting better, but we're nowhere near the point where it's good enough to provide the appropriate learning environment for anything other than basic rehearsal or discovery of very basic concepts.
Fundamentally, games are just not conducive to any sort of deep learning.
clarry: Where does this fundamental truth stem from? I'm not convinced.
On the flipside, I think that the vast majority of primary (and probably most of secondary) education isn't particularly deep anyway.
This is a massive topic. I don't think that I can provide enough citations to make this understandable to lay people. I've spent the last year studying on top of another year in grad school on a different certificate and the better part of 10 years of professional experience.
But, if you start with Piaget and Vygotsky and work your way forward, it should be relatively clear why it is that computers, at least the kind we'll have in the next 10-20 years, are unlikely to be capable of doing much more than what was possible in the past.
But, in short, computers are linear and they need to be programmed ahead of time to deal with the complications that arise during the process. Computers for flashcards or simple tasks tend to work great, when there's one, or a small number of, predictable answers. However, when the answers are subjective or possible results aren't predicted ahead of time, they fail miserably.
I've worked with students having to enter their math into computer systems for years and it's ugly. Taking an answer that you've worked out on paper should be a nobrainer in terms of computers, but even those programs frequently get it wrong. If they can't score math and science correctly, then why on earth would they do better at the humanities where the answers are frequently subjective?
