Mobile SMS-based basic e.g. math & language education (AKA mSara prototype v0.1)

Recently we were travelling on a vacation through India, and while driving for hours between cities, I did some math with my children. We occasionally use Khan Academy at home, so I gave them similar exercises on paper. Later on the ride while looking out of the window and letting my mind flow freely, I started thinking: Why do my children have the opportunity to learn like sponges as much as they like, while others that we were driving past lack access?

So I had an idea, and used a few nights at the hotels while travelling to change the world one line of a code at a time by hacking a v0.1 prototype of a mobile SMS-based basic math education platform which I'll tentatively call "mSara" - as in "Mobile Saraswathi" - think mPesa for Education. It's already running online up in the cloud/s (thank you Heroku) , and you can try it out yourself by sending an SMS text message to +1 646 791 3552 (thank you Twilio), and it will respond something like this:

The idea is that system will track per user progress, which is easy as the sender number can be used as ID, gently ramp up difficulty, help with hinted solution suggestions if it detects one does not know the correct answer yet, and keep track of achievements, so one can learn more as you keep using it. Extreme simplicity in usage is paramount - no "menus" to "choose lessons" or anything like that.

For math learning, full literacy isn't needed, as basic number recognition should be sufficient (and reading a few keywords like "Hi", "Yes", "No" in English or a local language may be generally available; if not the exercises are probably self explanatory enough). If that is not a given, perhaps it could even teach basic numbers, using a dialog like this:

Similarly one could imagine some basic language vocabulary learning, initially certainly using multiple choice answers and not actually T9-style typing words, sending questions like "(local unicode script lang's word for 'house') = 1. mother,  2. house,  3. tree" - to which "2" has to replied as correct answer.

Assuming some of above is completed, the next and actually more interesting part is how to get millions to use such a platform... you would certainly have to go through local field partner organizations of various kinds which can reach into the target audiences. But even with that, what's the incentive for someone to use this? I reminded myself what got my young son interested in programming, and I'm sad ;) to report that it wasn't passion for OOP, "class" keywords and such; instead this young child's initial driving force was refreshingly openly monetary: "But papa, could I SELL the game we're making, you know like Minecraft?" ;-) So that then gave me a wild idea: What if this platform, which can easily track each participant's progress actually rewarded learning achieved through persistent trying not only with regular "WELL DONE" responses (which ARE important and can mean more than you may think), but actual... money? What if some donor funds, instead of leaking through gov programs or well meaning but perhaps inefficient NGOs, were to directly reach the top performing most able children per region on this platform? Mobile Airtel Money or Vodafone mPesa virtual currency bitcoins in exchange for making an effort to learn with an old phone? Tag line: "Earn a modern Android smart phone / tablet which we'll ship to get to you (with which you can access the next level of knowledge à la Khan Academy..) by successfully completing the SMS-based math course on your parent's old phone!" Perhaps this is completely crazy.

The "target audience" for such a platform of course isn't necessarily constrained to "children" of course, there is really no reason to artificially age limit, or say "time bomb" any of the exercises. If I can help someone learn, at whatever pace that they can afford (perhaps in parallel to working), that's great.

So what next? I'd really be very curious to try out how well this crazy idea actually works in practice in real world field test(s). It certainly needs a little bit more development; notably incl. more country local SMS gateway bindings. For example from India, using my first POC's US number isn't really viable, because its much more expensive to SMS an international number, and because incoming SMS From: phone numbers get screwed up with a prefix to track "telemarketers" due to TRAI government regulations - well meant spam protection, but in practice making a two way service such as this one not viable easily. So I'd need an Indian local short code for sending “non-commercial” messages without that prefix, but in the limited time I had to investigate this found out that set-up is somewhat bureaucratic (as in it's certainly feasible, but a little more time consuming for me than just "one click away" as offered internationally by e.g. Twilio). I'm also unclear if and how a SMS short number that is free to text to could be set up e.g. India - as traffic will likely be high, cost of even just Rs 1.00 for each answer may be an issue for the target audience? Incoming SMS are free in India, BTW.

I'll see how much of next steps I can do when myself; but alternatively, someone's philanthropic funding money to pay for other people's time would certainly help me.. ;) In parallel, reliable local field partner(s) to reach out to possible target audiences to see if this really creates value would be needed. If you have (access to), are or can support me in any of this with your time, money or connections, please do contact me!

PS: See the Ignite talk presentation about this I gave about this at Devoxx Belgium 2014, and see the open source code of this project on GitHub.