Here is one of my longer emails referencing the RALPH technology platform.
This also tells you the state of affairs as regards.... well, certain things.
Please note that the nickname "Digital Blue" came after 2009.
~
Dear Professor Dixit,
I hope this finds you well.
I wanted to share a small but personally significant milestone: I’ve just finished writing my first book in the Esperanto language. What makes this even more remarkable is that I’m still only at the A1 level — I’m literally sitting with beginners in my Esperanto classes. And yet, despite that, the book is done.
The responses from the Esperanto community have been encouraging. One comment in an Esperantist Telegram group simply read “Mirinda” — “amazing.” (Please find attached to this email two screenshots from my cellphone, where you can see the "Mirinda" comment.) What no one said was “this is impossible.” That’s important, because it reflects an underlying truth that I’ve been exploring for a while: constructed languages (ConLangs) can be picked up far more quickly than natural languages, and here is the best part: they can be specially designed so that technology systems today can easily impart them to children and adults.
This aligns with the central thesis of my Digital Blue paper, which I wrote in part while on a break from HBS. In that paper and subsequent work, I have argued that technologies built on ConLangs — and platforms that support the learning of ConLangs — offer a dramatically more efficient route to literacy. If the core goal of literacy is to help someone go from zero to one — from not being able to read, write, or speak any language to mastering one — then ConLangs make sense as a solution. In a nutshell: Build technologies on ConLangs, and get people literate fast.
Natural languages are full of irregularities and exceptions that make them slow to learn. English, for example, is riddled with quirks that frustrate learners and educators alike. Most teachers of French or English say that it takes at least two years to learn French or English. So why not take a more engineered approach? Why not design languages explicitly for the purposes of rapid uptake, aligned with current technologies, and teach them — even if not as a first language, then as a gateway second language?
I have already done a Reddit AskMeAnything for one of my poetry books, because it was written so quickly (using NLP and A.I.) and with certain ideas in mind. Now, everyone knows about ChatGPT and knows that this can be easily done. I’m planning to try and do another Reddit AMA soon to discuss these ideas. Even if formal academic interest is limited, perhaps someone in the philanthropic world — a millionaire or billionaire aligned with education or digital equity — might see the value in exploring this approach. But this seems like such a roundabout approach. I would like to be able to explain the whole thing to someone such as yourself. Almost everyone who has spent the time to understand this idea sees the merit in this approach. One does need to support this idea further with empirical research, but interventions involving development economics often cost a lot of money. A quarter million dollars would not be considered outrageous, and, of course, we would need to find sponsors. I would appreciate it if perhaps, I could meet someone in your department, or indeed, yourself, if you have the time some time later in the year.
In the meantime, if this line of thought sparks any ideas or connections on your end, I’d be very grateful for your thoughts.
Warm regards,
Anand