When they arrived in his office and Abir explained the concept for what is now called the decoder, Carbonell was floored by its elegance. “In the few weeks that followed, I kept wondering, ‘Why didn’t I think of that? Why didn’t the rest of the field think of that?’ Finally I said, Enough of this envy. If I can’t beat them, join them.”
I’m floored too. (And envious!) What Meaningful Machines lyrically calls «flooding» in a recent Wired article, Me Translate Pretty One Day, is a stunningly beautiful translation algorithm, baffling in its simplicity.
Though if it’s simple to state and understand, it’s only because it relies on operations on a terrifying (computational, mathematical) scale. (Like the first time one invokes inside a theorem, say, the set of all possible sets, there’s a mixture of fright and awe — we can barely believe our moxie to write such thoughts.) In a very real way, the algorithm is written in Moore’s law language and if it escaped us all it’s mostly because our words are so shy, so inadvertently constrained by past assumptions.
Ah! How exciting! Machine language translation is on the horizon.