Hey HN - a little background, which might be interesting to some. I went through YC twice, the second time with a company called EveryArt. After working on that for two years as a solo founder, I decided to close it down. I moved from San Francisco to Istanbul and started a niche business (actually, just renamed EveryArt, Inc. in fairness to investors) that has done well enough to support me. I started teaching myself objective C in January to eventually build CleverDeck, which was motivated out of dissatisfaction with the available spaced repetition apps available (Anki, Memrise, etc.) I'll be hanging around if you have any questions. Thanks!
1. I just wanted something I could pull out of my pocket and tell me what to learn. Of course I want to be able to customize it (and you can in CleverDeck), but I found myself spending so much time in the others creating and managing different lists. With CleverDeck, the app is the list. If I don't have time to add words that I'm encountering in the wild, then the app will pull from a default ordering that's sorted by "usefulness."
2. I didn't really like the user experience of others. Too many confidence levels led to cognitive fatigue; I had to tap through menus to find which lists had which number of cards I needed to review. With this, just open it and swipe. It just tells me when cards come up for me to review, and automatically presents 10 new cards the next day (or whatever I have the setting adjusted to).
I use Anki for everything except language learning. It helps me remember books I've read, people, certain numbers, facts, etc. I am really grateful for a lot of the complex features Anki has.
My main beef with Anki is the developer's unwillingness to open up the API. Card creation is currently a pretty unpleasant experience, but I can't currently build a browser extension without doing a lot of hacky bullshit because the the developer won't expose any public endpoints.
AnkiMobile developer here. Are you referring to the mobile app or the desktop app? Because the desktop has API endpoints you can hook into with extensions [0]. If you're talking about the mobile app, I believe we originally decided not to provide an extension API because Apple forbids it.
I am talking about a REST API that hooks into your AnkiWeb infrastructure. I believe I asked about this in one of the google groups and was told you aren't interested in doing that because of the additional load it could put on the service. I would love to put together something like the Evernote webclipper for creating cards right here in the browser without doing this awkward alt+tab+click+click+click+etc dance to create cards on the desktop client. I admire the work you and others have done on Anki and its extensions, but if I'm going to use Anki as my primary learning tool it needs to be delightful and effortless to create/delete cards as needed.
Ah OK, in that case I can see both sides. From Damien's perspective, AnkiWeb is a free service, so development/support time and server load have to be considered for any new feature. From the user perspective, web clipping tools would be really nice.
Is there anything to prevent interacting between the desktop app API and a web clipping browser plugin, in the same way that the Zotero connector plugin does with the Zotero app API for example?
CleverDeck looks great and I've been looking for something to use alongside Duolingo. Do you have plans to expand from vocab to grammar? Duolingo can be too flexible and I'd love some more rigorous lessons on conjugation, etc.
Thanks! Right now, we just want to make the best piece of technology possible for learning thousands of words. I don't think grammar is learned very well through flashcards (though, every word does have an example sentence, so you could pick up on some through pattern recognition).
But, for sure, CleverDeck is just one tool in a language learner's toolbox. I have DuoLingo on my phone. It's great. But not ideal for learning tons of vocab.
I'm a Turkish teacher! I make podcasts to help people learn Turkish. Supplemental material costs a monthly subscription. It's here: http://turkishteatime.com
Istanbul doesn't generally feel much cheaper than San Francisco, except for rent. I pay $500 to live in a gorgeous apartment with a sea view in downtown Istanbul. That has really added up over time.
As another story to back this up, I moved to India when I was bootstrapping my first business. I got down to $2/day for food, and found free rent. This low personal burn rate massively extended the amount of runway available. Of course, Bangalore is no longer low cost, but there are plenty of places that still are.
Off topic: I recall talking about this last year and I believe we have met quite randomly over some beers when I was travelling in Istanbul in July 2013! Good luck for your business!
Isn't the cost of living there (eating) cheaper than in San Francisco ?
I am from Austria (Western Europe) and have been in Istanbul some weeks ago. If you are not going to eat at tourist traps I had the feeling that eating was cheaper than in Austria.
Buying clothes and shoes was cheaper than in Austria too - so I guess the cost of living there is less than in Austria.
Beer and alcohol is / can be pretty expensive - but other than that I thoughts it is pretty cheap there (as a tourist)
I find the Anki software to be perfect for my purposes of spaced repetition learning. I've invested a lot of time in creating my decks. Because I don't want to recreate those decks again, is there anyway to import those decks I've already created onto CleverDeck?
Not right now - I'm sorry. If Anki is perfect for you and you've made all your decks, that's great! Keep using it! If you ever decide to learn another language, we'll be here with a bunch of content ready for you :)
Yeah, definitely. Learning languages seems to be the largest use-case for spaced repetition apps, so building in lots of good content for language learning seemed like a smart way to differentiate and target ourselves at first. One of the biggest annoyances I had using the others was how much time I spent building my own decks - when, really, I just wanted to learn common, useful words. It's true that some apps include user-created content you can share, but the quality was pretty hit or miss - and usually didn't come with pre-associated images, sound, and example sentences.
I can't install it right now as I am upgrading to iOS 8 but why is the app free and the 3000+ cards mentioned in the description cost $15?
Is it because the App store doesn't support any form of trial? It may be unfair, but I always feel like I am being tricked when a purchase is set up this way?
Oh no, I really hope it doesn't give that impression. I'm open to suggestions on how to avoid that perception.
It's freemium. You can download the full-featured app with a deck size of 50 (also enough to learn just enough words for a short trip for free). If you like it, you can add different increments to the total deck size through in-app purchases: 99 cents for +100, up to 14.99 for 3000.
Didn't mean to be tricky. I thought this was a good way to price it so everyone could try it, tourists could get what they needed for free, and every gradient of language learner could only pay for the amount they needed.
Thank you for taking the time to reply. I apologize if my comment was rude or insensitive. It looks really nicely done and I can't wait to try it once the iOS 8 install has finished breaking my phone.
German and Mandarin are my two requests (even just 300 or so Mandarin cards would be great! I'll help!). Just started on French while I wait, looks great! :)
There is a fantastic dictionary app for Mandarin called Pleco that includes a SRS-based flashcards feature as an in-app purchase (around $10-15 if I recall). The base dictionary itself is free.
While of course you can just program your own cards into Anki or the like, being able to create new cards from dictionary definitions with one tap is very handy.
Ah man - one of the things I've learned from my launching my first iOS app is how painful it is to answer this question.
Painful because yes, I would LOVE to. Let's see how people like it on iOS, then I can decide the best way to go about getting it in the pipeline as quickly as possible.
If you sign up on the email form on the front page of cleverdeck.com, I will send something out when we launch an Android version. But that means you'll also be spammed when we launch new languages for iOS. Not ideal - I'll try to add a separate signup somewhere tomorrow.
Chrome 37.0.2062.120 m (latest version) for Win7 x64 has some sort of validation problem upon entering a perfectly valid gmail address in that signup field. Hitting subscribe just makes it highlight red. Works in IE though. PS: I'm looking for the android version too! Anki is great but clunky.
Bummer this wasn't planned on from the start, seems like this type of app is well suited for one of the many cross-platform tools.
No doubt that iOS is the bigger market, but Android is still plenty big and there is a dearth of well designed apps so I think it can still be rewarding.
Anyways, count me as another eagerly awaiting an Android port. Signed up on the website mailing list.
I almost hate to ask this question here because this seems like a really nice product, but does anyone know of any similar apps? I ask because I'm learning Polish, a language not yet available on CleverDeck.
I once wrote https://github.com/darius/spaced-out which is certainly not serious competition, but came from a similar idea: build the deck from the most common words in the language (automated using an aligned parallel corpus), and replace the Anki interface because I didn't like it.
It's probably a better starting-point than starting from scratch, I'll say that for it. (I wasn't interested enough in language-learning to take it further.)
Nice UI. I would love if you add a "Advanced English" deck. I have a 11 years old vocabluary (according to some online test I forgot the name).
There is a place for people learning english as a second language where you know enough to be able to communicate, but you don't have a good enough vocabulary to express more sophisticated ideas, or express simple ideas more elegantly. I am at this place and I am sure there are lots of others here.
Yes, when talking about the language learning market, English is the whale. We're going to get there as fast a possible, but it means we'll have to be prepared for everything that comes with having non-English speaking customers.
Though, having an Advanced English deck - so people have good enough English to tell us all the problems we have - sounds like a terrific way to start :)
As a non-native English speaker myself, I agree with GP. I have a bunch "English Vocabulary in Use" books of various levels which I almost never open because, you know, it's hard to carry it around with you.
you don't need an app for this, you need to read more books. Not about expanding your vocabulary, but just quality fiction novels. That's how native english speakers broaden their vocabulary and your ability is clearly far enough along that most books will be within your grasp. From what you wrote your vocabulary is fine, it's the nuaces of grammar where your English would be improved by reading copiously.
One thing that's unclear in the app description: does the free app come with 3000 words? And then is the $15 purchase is for an additional 3000? That would be amazing!
Are all the words nouns? Or are there verbs/adjectives/adverbs, too?
Is there a way to create a set of flashcards via a file upload? That'd be a wonderful feature!
Good questions! The free app (there's only one freemium app, actually), has all 3000 words built-in. We charge based on how many you swipe into your deck (i.e. you decide to actively learn). The app comes with a free deck size of 50, then you can increase that incrementally with in-app purchases.
There is a healthy mix of nouns, verbs, adjective, and adverbs.
There isn't currently anyway to add your own content outside of creating individual cards within the app. It's something that I'd like to get to (if people find the SRS system useful, I'd like people to do whatever they want with it).
I was genuinely confused about this too, and am still not 100% sure I get it after your comment. So the deck size means it's 50 concurrent cards, but I can add/remove card sets to my deck as much as I'd like using the free version?
If I remove cards from my deck to make room, does this mean any training on how well I've told it I know those words is lost?
Looking great, just a small pet peeve, in the coming soon section, above portuguese you have the brazil flag, feels kinda weird being portuguese and seeing that :)
I know, I know - I'm sorry. Despite some of the objections, I just think flags do make great language icons. I do understand that complaints, though.
In that particular case, we chose the Brazilian flag because the list and audio is target at Brazilian Portuguese. We should probably explicitly write "Brazilian."
Interestingly, we couldn't find a flag icon for Arabic that was neutral enough. So we just chose the green color from the Arab League flag (which I guess could itself be controversial) and wrote "Arabic" in white text.
I would write "Portuguese (Brazil)" rather than Brazilian, similar to how you often see "English (US)" or "English (UK)". If I saw "Brazilian" my first thought would be "whoever wrote this doesn't know that Brazil speaks Portuguese".
Looks and works really, really nicely. I am currently in the midst of cramming Spanish (trip to Barcelona in two weeks, trip to Cuba in four months, woo hoo!) and I've been using these two trips as a motivation to finally learn the Spanish I've always wanted to know.
I've been using Duolingo a lot, as well as Memrise. Memrise has been quite good so far, but it is certainly not very polished, whereas this seems extremely well-done.
Congratulations on the launch, this looks like a superb product and if it works for me I'd be happy to pay for it. That's a clever idea for in-app purchases too, by the way.
Hey, I lived in Barcelona throughout March this year and Catalan is the dominant language. I felt (and perhaps I was too self conscious) that when I spoke Spanish they worked out pretty quickly I was English and defaulted back to that.
Either way, learning Spanish is definitely an advantage as there's crossover to Catalan. Hope you have a wicked time :)
Well hey now. I can't believe I never came across that piece of advice before, but you're right! I knew that Catalan was spoken but did not realize it was the dominant language.
I guess I better learn some basics...as if I didn't have enough to learn already. ;)
Duolingo is nice, but it's not as good for cramming as for low-intensity, long-term learning. Anki is awesome for just "I'm going to memorize this speech like a poem"
I agree with you on DuoLingo - great app, not good for vocab learning.
Anki was my main SRS app for many, many years. I didn't like managing all my own lists (and was often buggy). I didn't like the UI. Adding pictures and audio was awkward. It didn't have some basic features like reminders or notifications built-in.
Also, I had to add all my own stuff to Anki. If you want to learn Spanish (for example), we've already done all the work of picking the 3000 words we think you should learn, added high quality images and audio to each one, and gave each an example sentence with translation. I think that's what's most exciting to me.
Thanks so much for building this! Just purchased the full 3k Spanish deck. You've saved me a ton of time vs. my plan of writing a script to build my own Anki decks.
One question: is there a recommendation on whether to go with English-first or target-language-first when studying? I prefer to go with Spanish-first, but I don't know if there's any data on which is more effective. Thanks again!
In building this app, we didn't survey the research addressing that question specifically (though googling "L1 vs L2 flashcards" pulls up some interesting sites).
For me personally, speaking from 15+ years learning French, Chinese, and Turkish, it's native --> target. Learning that way helps me recall faster when I need to construct my own sentences. If you're looking just to learn to read, I think target --> native is easier and might be a better approach.
I've also had success starting with native --> target to familiarize myself with the word, then quickly switching to target --> native to solidify my ability to produce it on my own.
138 comments
[ 2.6 ms ] story [ 195 ms ] threadBut to name the big two:
1. I just wanted something I could pull out of my pocket and tell me what to learn. Of course I want to be able to customize it (and you can in CleverDeck), but I found myself spending so much time in the others creating and managing different lists. With CleverDeck, the app is the list. If I don't have time to add words that I'm encountering in the wild, then the app will pull from a default ordering that's sorted by "usefulness."
2. I didn't really like the user experience of others. Too many confidence levels led to cognitive fatigue; I had to tap through menus to find which lists had which number of cards I needed to review. With this, just open it and swipe. It just tells me when cards come up for me to review, and automatically presents 10 new cards the next day (or whatever I have the setting adjusted to).
I doubt anyone can replicate everything that Anki can do, but there's probably some gains to be had by cutting out complexity.
[0]: http://ankisrs.net/docs/addons.html
Is there anything to prevent interacting between the desktop app API and a web clipping browser plugin, in the same way that the Zotero connector plugin does with the Zotero app API for example?
https://github.com/dsnopek/anki-sync-server
But, for sure, CleverDeck is just one tool in a language learner's toolbox. I have DuoLingo on my phone. It's great. But not ideal for learning tons of vocab.
Istanbul doesn't generally feel much cheaper than San Francisco, except for rent. I pay $500 to live in a gorgeous apartment with a sea view in downtown Istanbul. That has really added up over time.
As another story to back this up, I moved to India when I was bootstrapping my first business. I got down to $2/day for food, and found free rent. This low personal burn rate massively extended the amount of runway available. Of course, Bangalore is no longer low cost, but there are plenty of places that still are.
I am from Austria (Western Europe) and have been in Istanbul some weeks ago. If you are not going to eat at tourist traps I had the feeling that eating was cheaper than in Austria.
Buying clothes and shoes was cheaper than in Austria too - so I guess the cost of living there is less than in Austria.
Beer and alcohol is / can be pretty expensive - but other than that I thoughts it is pretty cheap there (as a tourist)
Looking at the network it seems the return value was success, but it says "We'll just need to confirm your email address".
Shouldn't there be some text or something besides a green border?
Either way this looks awesome...
Is it because the App store doesn't support any form of trial? It may be unfair, but I always feel like I am being tricked when a purchase is set up this way?
It's freemium. You can download the full-featured app with a deck size of 50 (also enough to learn just enough words for a short trip for free). If you like it, you can add different increments to the total deck size through in-app purchases: 99 cents for +100, up to 14.99 for 3000.
Didn't mean to be tricky. I thought this was a good way to price it so everyone could try it, tourists could get what they needed for free, and every gradient of language learner could only pay for the amount they needed.
While of course you can just program your own cards into Anki or the like, being able to create new cards from dictionary definitions with one tap is very handy.
Painful because yes, I would LOVE to. Let's see how people like it on iOS, then I can decide the best way to go about getting it in the pipeline as quickly as possible.
While I understand that porting and maintaining two applications is difficult, you're ignoring half your potential market.
Good luck, I hope to see the droid version soon.
No doubt that iOS is the bigger market, but Android is still plenty big and there is a dearth of well designed apps so I think it can still be rewarding.
Anyways, count me as another eagerly awaiting an Android port. Signed up on the website mailing list.
Cheers.
iOS users have historically spent significantly (4X? 10X?) more per user.
Check out Anki and Memrise. They, in my opinion, are the best SRS alternatives out there.
It's probably a better starting-point than starting from scratch, I'll say that for it. (I wasn't interested enough in language-learning to take it further.)
http://www.supermemo.net/
There is a place for people learning english as a second language where you know enough to be able to communicate, but you don't have a good enough vocabulary to express more sophisticated ideas, or express simple ideas more elegantly. I am at this place and I am sure there are lots of others here.
Though, having an Advanced English deck - so people have good enough English to tell us all the problems we have - sounds like a terrific way to start :)
$14.99 seems to be a right price, too.
For the original question, there's an Anki deck for English idioms with sentences.[0]
[0] https://ankiweb.net/shared/info/846593111
Edit: It seems Anki is mentioned quite a bit further in the thread, so I'll skip defining it here.
Are all the words nouns? Or are there verbs/adjectives/adverbs, too?
Is there a way to create a set of flashcards via a file upload? That'd be a wonderful feature!
There is a healthy mix of nouns, verbs, adjective, and adverbs.
There isn't currently anyway to add your own content outside of creating individual cards within the app. It's something that I'd like to get to (if people find the SRS system useful, I'd like people to do whatever they want with it).
If I remove cards from my deck to make room, does this mean any training on how well I've told it I know those words is lost?
In that particular case, we chose the Brazilian flag because the list and audio is target at Brazilian Portuguese. We should probably explicitly write "Brazilian."
Interestingly, we couldn't find a flag icon for Arabic that was neutral enough. So we just chose the green color from the Arab League flag (which I guess could itself be controversial) and wrote "Arabic" in white text.
Though, we will be launching Mexican and Canadian soon as well!
I've been using Duolingo a lot, as well as Memrise. Memrise has been quite good so far, but it is certainly not very polished, whereas this seems extremely well-done.
Congratulations on the launch, this looks like a superb product and if it works for me I'd be happy to pay for it. That's a clever idea for in-app purchases too, by the way.
Either way, learning Spanish is definitely an advantage as there's crossover to Catalan. Hope you have a wicked time :)
I guess I better learn some basics...as if I didn't have enough to learn already. ;)
Duolingo is nice, but it's not as good for cramming as for low-intensity, long-term learning. Anki is awesome for just "I'm going to memorize this speech like a poem"
Anki was my main SRS app for many, many years. I didn't like managing all my own lists (and was often buggy). I didn't like the UI. Adding pictures and audio was awkward. It didn't have some basic features like reminders or notifications built-in.
Also, I had to add all my own stuff to Anki. If you want to learn Spanish (for example), we've already done all the work of picking the 3000 words we think you should learn, added high quality images and audio to each one, and gave each an example sentence with translation. I think that's what's most exciting to me.
Most of the example sentences were hand-picked from tatoeba.org. That covered about 2700 of the words - our teachers wrote the remaining 300 or so.
One question: is there a recommendation on whether to go with English-first or target-language-first when studying? I prefer to go with Spanish-first, but I don't know if there's any data on which is more effective. Thanks again!
In building this app, we didn't survey the research addressing that question specifically (though googling "L1 vs L2 flashcards" pulls up some interesting sites).
For me personally, speaking from 15+ years learning French, Chinese, and Turkish, it's native --> target. Learning that way helps me recall faster when I need to construct my own sentences. If you're looking just to learn to read, I think target --> native is easier and might be a better approach.
I've also had success starting with native --> target to familiarize myself with the word, then quickly switching to target --> native to solidify my ability to produce it on my own.