What is the Capital of Brazil?
And solving problems as athletes and spectators
“What is the Capital of Brazil?”, a friend once asked me.
My immediate reaction was to say “Rio de Janeiro”, and the cities enormous Christ statue atop a mountain visualized in my mind. But then, almost immediately, my brain said, “No. That’s wrong.”
Why did I think of Rio? Probably because a few hours before my friend asked the question, I’d been listening to some music, and happened to look-up the composer’s name; he was born in Rio.
My next reaction was to say, “Sao Paolo”. But again, my brain said, “No! Brazil is one of those countries, like Australia, where the capital is not the biggest city.”
Finally, my brain said “Brasilia”; and, of course, that was the right answer.
Utterly unrelated to my friend’s question, I felt a very subtle sense of disgust; and I soon realized why. I had been reading about how Brasilia, designed by Oscar Niemeyer, was a planned city, and hence had an “artificial”-ness which was, to many, ugly. For some reason, I had internalized this ugliness.
…
I have work to do this morning.
To be more exact, I have to design the architecture for a Machine Learning system.
“So, why are you wasting time thinking about Brazil?”, you might ask.
Well, because “How do I design an ML architecture?” and “What is the Capital of Brazil?” have a couple of things in common. For one thing, they are both “problems”, for which I have to (or had to) find “solutions”.
As I had thought of Rio de Janeiro, Sao Paolo, and finally the artificial Brasilia, I don’t remember spending mental energy at the time. My brain seemed to “flow” from one to the other, eventually reaching the solution. It was almost as if I watched my brain do the work; as if I was a spectator of a race, and my brain was the actual athlete.
In contrast, as I tried to solve the ML Architecture problem, I was trying to be the athlete; and was not making much progress, with no finish line in sight.
…
But what if I tried something different? What if I tried to solve the ML Architecture problem by being a spectator, and not trying to be the athlete?
And so I did.
Without trying to solve the problem, I let my brain wander. I thought of this and thought of that; concepts in ML, the business problems that the architecture was trying to solve, completely disconnected things like what I was going to have for breakfast.
And then, almost like a thunderbolt, my brain said something.
“Why am I trying to solve the ML architecture problem?” my brain asked me. “Is this the most important problem to solve? Is it not better to spend your time working on a different problem? Like what your ML strategy should be for 2021.”
Not only did my brain ask me this question, but it also suggested some interesting answers. And so, I’ve decided not to worry about architecture, and think about the “bigger picture” for 2021.
…
In retrospect, I wonder if we ever “willingly” solve problems, or whether our brains or minds or other subconscious processes solve the problem, and we later take credit for them; rather like spectators claim (at least part of) the credit, when their athlete or team wins
I wonder if there is any significant problem that I or anyone has solved as an athlete, as opposed to as a spectator…
