I honestly feel like I’m developing attachment issues with language apps. Duolingo started pushing plus and ads way too hard and stripped off most of the functionalities I used so I haven’t used it in a long while. And now this is happening.
Guess this is the push I needed to get a better grip on Java so I can make my own app and actually have some stability lol.
2 Likes
My guess is that they want to get rid of the Community courses eventually and keep only their courses. This was here already couple of years back and by removing mems, they take away one huge incentive to use Community-created courses.
4 Likes
- No more mems: I use them all the time and have created hundreds myself
- No typing: Fine for word learning but what about longer sentences? I don’t want to have to speed read when I can type them myself (and learn the structures better that way)
- I turned off the beta program ages ago because I found it unbearable, and I still can’t bear to turn it back on. The thought that this will soon be the only way is depressing.
Way to throw cold water in someone’s face. I wish you would reconsider.
13 Likes
Amen to that…
Bye bye Memrise on january 1st 2022 
10 Likes
I have been studying Korean for 8 years - 5 years using Memrise. The memes have been invaluable for me and the few words I struggle to remember are the ones I couldn’t find a good meme for. What am I paying a subscription for if not for the memes? Seriously!
5 Likes
Hi everyone, I’m a co-founder here at Memrise and as Chief Strategy Officer, this decision around mems comes back to me. I’ll let others on the team talk to other aspects of the beta rollout, but I think that mems deserve a special discussion.
I first want to acknowledge that we haven’t explained this as well as we could have. I also acknowledge that this has been a long standing pattern. Alanh and Hydroptere have had this discussion with me many times over the last decade. I am old enough to suspect I personally am unlikely to change my ability to get these comms right up front, but I will keep trying, and I hope that I am able to correct that failing now and through open discussion in the coming days.
First I want to talk about the intention we have.
For that, a little bit of background on Memrise and what our company aims are. I apologise that this is a bit long - but I think that any answer I give you that doesn’t start at the beginning is likely to be incomplete and cause more confusion.
So, here goes:
I founded Memrise because the way that I’d been taught languages at school hadn’t worked for me. Then through my studies of psychology and reading the likes of Steven Krashen and Michael Lewis, I found a way that did work for me, and I wanted to share that with as many people as possible.
The method involves combining memory techniques to learn words and phrases fast with spending time in the right kind of “immersive” experiences so that you convert “knowing” the language into the ability to actually “use” it to understand and communicate.
This method was how I learned to speak Chinese, while working in a motorcycle factory in Beijing. That was where I learned how powerful this method can be.
The first content for Memrise came directly out of that experience: our first course was made using the mems I created for myself to learn Chinese.
Mems are a way of “richly encoding” memories. That is important. (if you are wondering how I can say this and still have made the decision that we should remove them, please read on below).
The other part of this approach to learning languages is to have “immersive” experiences that use the language. The only way that you gain the ability to hear language spoken and for meaning to appear in your mind is through listening to and trying to understand lots of content in the language. If you know a lot of words and rules, but you haven’t spent time processing target language input, you treat language like a puzzle: you can work out what it means, but you don’t speak or understand with fluency.
This second part if what we need to add in to the Memrise product. We are doing this in a couple of ways:
-
We are building a separate site - MemWorld - that is a collection of immersive experiences which learners are led through by native speaker “coaches”. This includes live video calls as well as video playlists and podcast selections. This is all free to use.
-
We are adjusting the structure of the Memrise product to allow it to effectively integrate with this new site.
With these two pieces together, we will create a product that covers both aspects of the language learning methodology that we founded the company to bring to the world.
I want to give a little example of what this looks like:
-
In MemWorld you might have a video call with a language coach and some learners. During the call some words and phrases come up that you want to learn.
-
At the moment you could make a UGC course on Memrise to learn those. However the course that you create would have no video and would have limited test types - no comprehension tests, no pronunciation tests, none of the other tests that are being developed.
-
In the future structure, a wordlist could be automatically generated for you on Memrise that has all the rich content of the official courses. You also would not have to make it yourself.
-
The same thing could happen when you watch a video or listen to a podcast: any immersive experiences you have can be dynamically converted into a wordlist on Memrise that has rich media test types integrated into it.
This structure will also allow users to make courses that are integrated into the core Memrise experience: in the mobile apps, and complete with rich media.
This is the future product that we are committed to building over the next few quarters, and why: it is a product that delivers on the learning method that we believe in and which is evidenced by the work of people like Steven Krashen. THe current Memrise structure odes not allow us to do this effectively. We need to make some core changes.
Memrise has been around for 11 years. Some of the code is very old. Making the changes that we are making is complex. In order to make the change that we want in the timeline that we need to, we need to make some touch calls. This is one of them.
Mems are part of the old tech stack. Keeping them alive in their current form as we move to the new tech stack will add a great deal of time to the process. We still believe in the power of mems. We have every intention of building them back into the experience in future. But we have to take one step back so that we can take two steps forward.
I appreciate that this won’t make it any less annoying for all of you in the meantime: I’m asking you to trust that we are going to take two steps forward after taking this step back. I know that is a big ask. If I could see a way to get to the product that we want to create that didn’t involve taking this step back - believe me I would grab it with both hands. But after a lot of work and research by the team, this is the route that gives us most certainty of getting where we want to get as fast as possible.
I appreciate that is a lot of information. And there is still a lot more to discuss around the subject. I’d be more than happy to get on a zoom call later today or tomorrow if there is interest in it - I’ve set up a call here at 2.30pm UK time please do join me if you can - but I can set up another tomorrow or later in the week if this is too soon.
To set expectations, we are unlikely to change the high level decision that we need to retire mems in their current form in order to move forward and then re-introduce them in future - but I would certainly like to hear your input and see if there are any mitigations that we can make on the way to do that.
I’ll reply as much as I can this evening and then each evening this week.
Ben
Please, just leave “only typing mode” in “Classic review”! I beg!
15 Likes
Apparently I was right. See the creator’s answer about their plans. Effectively removing everything Community-created and focus on their product. Although maybe with some customization.
3 Likes
Guys, I have to apologize to all of you. This is obviously all about me and Memrise discouraging me from purchasing the subscription. Couple of years back, when i considered it for the first time, they came with the Decks and removing user-content. Well, why would I pay them in that case, right? Even though they reverted their decision, I was still sceptical, assuming similar decision may come any time. Anyway, now I started considering the subscription again, because I’m using Memrise heavily lately.
Well, with such change, I’ll better quit altogether. The beta looks awful. It doesn’t work as it should. The Community courses will be gone soon. at least I’m glad I know about it before I started making new courses, which I just planned.
7 Likes
Sorry for the miscommunication here - the intention is 100% to put user created content into the main Memrise product rather than having it separated out in a separate experience as it now is. This is what the new structure is aimed to do. It is not about removing everything user made - quite the opposite. But to do that we need to take a step back - that is what this is: a step back so we can take two steps forward.
1 Like
Hi Ben @BenWhately,
so nice to see you around. I have decided some years ago not to take part in the fora any more - aggressive co-memrisers, every now and then not very nice Memrise staff – but they came and went, Alanh and Hydroptere and others are still here) … but I still feel indebted ( 
) to you as part of that wonderful site that made me learn some Mandarin and Japanese, I cannot hold myself back.
I still do not get it: do you indeed plan to delete soon the user content, the user-made courses ???? i am afraid that even Alanh and Hydroptere will leave then…
I’d like to push forward some tiny requests for customisation. Would it be possible to have some options such as: 1. all typing; 2 no typing at all (for situations when the number of items in need for watering is overwhelming – which happens often to Alanh and Hydroptere and others 
); this because speed review is not a happy solution at all 3. timer ?
The new UI is really disconcerting, big letters, empty spaces splashed on the page – especially in review, and writing new user scripts means so much work….
These being said, I’ll miss my mems….
thanks for listening
7 Likes
BenWhately: Given the sorry history of app updates, missteps, user-hated changes, back-stepping, etc., I simply don’t believe you have the technical or creative competence to achieve the very ambitions user-immersion goals you are dreaming about. It’s going to take a pretty grand success to gloss over betraying your loyal users.
6 Likes
I understand that, though obviously its hard to hear. Nonetheless, I’ve still got to try to deliver the best thing I can. After 11 years building this vision I’m not about to give up. I also genuinely believe we now have the team in place that we’ve ever had, and we now do have what we need in order to deliver on the vision. Part of that ability is the ability to make hard decisions like this one.
That isn’t going to make you trust me. Only time and results can do that.
1 Like
Hi Ben,
Thank you for coming on and engaging with us in person. I, for one, appreciate it.
I also welcome the more detailed explanation of the rationale for the changes being made. On mems, for example, there is a big difference between “we beleive that [mems] no longer fit the Memrise experience that we’re creating, so we’ve made the difficult decision to retire them” [see OP in this thread] and “we still believe in the power of mems and we have every intention of building them back into the experience in future”.
On user-created content, could you clarify your follow-up comments above for me, please? Are you giving us your assurance that our existing user-created courses will survive in your new world - or not?
Whilst dipping in and out of Memrise Beta over recent months, I have preferred to turn it off and use the ‘all-typing’ script on a selective basis because there are some courses where I don’t find multiple-choice or tapping tests helpful. I tried Beta again today on a range of course types and subjects and did not enjoy the experience at all. An option to be able to select the type of testing and whether tests are timed or not would be a step in the right direction.
Thanks for listening.
11 Likes
Hi Alan, it really is wonderful to see you and Hydroptere still such great contributors here.
In the initial thread I think the team were reluctant to give the message that “we are taking a step back and you’ll have to trust that then we’ll take steps forward.” They feared a response such as @ drifterprof quite reasonably gave.
I understand why that was the message, but I think it is more helpful to give the full picture for why we’re doing what we’re doing.
On user created content, I can share the approach we are taking and what we are aiming for. I can’t guarantee that we won’t have to take some back steps on the way to get where we are aiming, but this is what informs our thinking and planning:
- We believe user made content is incredible important. The ability to list out words that you want to learn is a ubiquitous language learning need, and one that Memrise was designed to serve.
- We believe that serving it really well means making it super simple - ideally automatic - for lists to be made from the films you are watching, podcasts you are listening to and conversations you are having.
- We believe that the best possible way to learn words and phrases involves using test types that go beyond word-definition pair drills, and move into tests from context and other tests from rich media involving a range of voices using words in a range of contexts. These tests are only possible in a world where all course can draw from a central database of “target language items”.
Our strategy for building an experience that serves all these beliefs is this:
- We will create the ability to make lists of words within the main Memrise experience. These will look more like current “levels” than like “courses”, in that they will not have subdivisions. But you will be able to make many of them.
- These lists will draw from a central database of words that means that they will all come with video content and multiple rich test types that go beyond word-definition pairs (yes, and in future the plan is for these also have mems). If you want to learn a word that isn’t in the database, then it will be possible to add it, though the exact mechanism is yet to be determined.
- We will then start recreating all the current UGC courses in the new within-Memrise system. We will keep the current UGC in parallel. We will recreate as many courses as are needed (there are literally millions of courses many of which are not used at all. We will not recreate all courses)
- We believe that the new course creation tools within the Memrise experience will be much better both in terms of learning and in terms of ease of creation (eg being able to generate them from text articles, and from videos). Therefore it is our hope that over time usage will move into the main experience and that usage of the legacy will fall. At some point we would stop supporting the legacy UGC courses, once enough people are satisfied with the new experience.
So the short answer is that existing UGC courses will survive for the time being. The ones that are in active usage will then be migrated to the new world. Then once the new world is working well enough for enough people, we plan to retire the old courses. That is not an irrevocable decision - but it is our intention.
1 Like
In short, this feels like the crux of the problem.
I get that maintaining legacy code is hard and that what you’re talking about almost certainly isn’t a trivial effort. It probably comes with considerable expense and what you see as opportunity cost. I’m genuinely sympathetic to those being hard choices.
However, I’d ask you to again look at the feedback here and ask yourself if you’re making the right trade-off here. You might be doing what feels right for you and for your team, and that’s a powerful incentive, but if you look at the level of anger and disappointment in this thread, I think it’s pretty hard to argue that you’re doing right by your users and your paying customers.
I’ll put it this way – the consensus in this thread seems to be that if you asked your users whether they’d rather wait an extra year for a set of features that they’ll really like or they’d rather see their favorite features immediately deprecated in order to maximize the speed of that feature delivery, it seems pretty clear that the current proposal isn’t in alignment.
10 Likes
NOW you are getting me worried.
5 Likes
Hi Ben, I’m sorry, but I don’t get what you are trying to build. Some parts of what you’ve said sounds like utopia.
Free cheese is only in a mousetrap, and you are promising free video lessons and other free learning tools. You have to pay your teachers and developers, have to pay for servers, and so on. Are you rich enough to sponsor millions of people in their language journeys? What is the purpose of building a product without perspective to gain?
Folks around here have their own goals in learning, why do you think they need something universal and that this product you are offering will cover their needs?
The most difficult thing to understand is that instead of tuning up, dedicating your resources to old good Memrise you are building absolutely new product, which is totally ephemeral at the moment. In the last 3-4 years you’ve hardly made any improvements here, were methodically destroying Memrise, and now you will build even more powerful learning website? Sounds doubtful.
This promise of bright new future provided that you’ll suffer a bit I’ve heard somewhere (not from good people). By the way, why did your team lied to us about mems statistics if it’s not the main reason to take them down?
This is so suspicious and dubious, I wouldn’t bet anything on it.
11 Likes
The business model is that the sessions are available for free (you can sign up now if you are learning Spanish to video sessions 3 times a week), but there will be other tiers and features that are paid for. But the core feature set will be free. The gain is in the overall ecosystem that we are building: we believe that this combination of effective learning and immersion is what is actually needed to successfully acquire a new language. That ecosystem will present lots of ways to monetise, but the heart of it should remain free.
I’m afraid I don’t know the details of the incident when you felt lied to, but knowing the people involved I don’t believe it was intentional; but again I recognise as per the above comments that you are unlikely to take my word for that.
I really do appreciate your continued engagement