Was reading the Alinea cookbook again, and Grant Achatz mentions that he was inspired for a dish by some tree roots he saw while walking with his kids through a park. He also talks about how he’s always pulling inspiration for his food from the rest of his life, and how basically, everything he sees passes through this filter of “how could this influence food?”
Reminds me of two things – the first, Shigeru Miyamoto’s descriptions of his inspiration for Zelda (exploring caves as a youth) and Pikmin (his then-recent obsession with gardening). And obviously, they’re fundamentally the same thing – someone who is so saturated by something that he breaks apart the rest of his (or her) existence and views the world through a very specific lens.
The second thing it reminds me of is a thought I had a couple months ago, but couldn’t really explain. I’d been reading a bit about how the mind processes things, and I haven’t read enough to really talk about it with any certainty, but there’s this idea of “chunking” things – creating shortcuts in how you process information around you so you don’t have to process every detail every time you see something.
Ei-Nyung used to say that she never really noticed what model cars were – basically a car had chunked to “car” where all cars were fundamentally the same, where for me, I’d chunked Mazdas and Toyotas and Ferraris and the like, then later broken up those chunks into RX-7s and F430’s. I find that the idea of chunking seems to explain the holes in my memory – how I process past events is really reduced down to a few (to me) critical details, and almost everything else appears to be disposed of. I dispose of who was at an event, for instance, while this is clearly one of the critical things that many (most?) people seem to place as a high priority.
So, the thought was this – basically, Achatz uses food as a hammer to break through the chunks. Most people look at a tree and see a tree. He understands food as flavor, form, presentation, etc., and then breaks the tree down into those things – the tree then becomes the inspiration for some aspect of the food. Miyamoto understands gaming as interaction, character, game systems, etc., and then un-chunks his world based around those things.
It seems to me that a lot of these sorts of creative breakthroughs seem to happen when you take something very ordinary, that you look at every day for weeks and weeks and weeks, and then use some “hammer” to break apart the chunks, and find some detail that’s relevant, or interesting, or that other people are overlooking. In which case, a person’s specific expertise is basically their mining hammer, and the way their brain has stored information is the quarry. You can enrich the quarry by learning new things, but until you take your hammer and break the stuff apart, you’re not really mining.
Hrm.
We definitely do chunking in both our inputs and our memory – there’s just too much freaking sensory data to process otherwise. We see a chair rather than a random collection of visual impressions. And at higher levels, we process input in light of the “cognitive subroutines” that we have formed.
So I’d put it slightly differently – rather than using a viewpoint to break apart chunks, a viewpoint determines the touchpoints which we use to categorize what we see/experience. There’s the study that showed that chess masters can memorize chess boards faster and more accurately than novices… unless the boards show an invalid position – the theory being that masters chunk the board into more than individual pieces (“Oh, that’s the XXX defense with a mid-game YYY gambit”).
Absolutely agree that innovation and creativity comes from applying new viewpoints to the world. Expanding one’s number of viewpoints is the way to increase the possibility of innovation. Alternatively, one can innovate by delving deep into a single viewpoint (like cooking or gaming) and applying that lens to everything else.
Anyway. I’ll shut up here.
Why shut up?
I wonder if you can accurately describe how a chess master stores a move as simply that they know how to chunk at different resolutions?
For me, for instance, I’d remember, “Hey there’s a chess game in progress!” and not be able to retain any detail about it. Or I might say, “Here are the exact positions of each piece,” and have a terrible time trying to recall them all.
A master, however, might remember familiar groupings – a knight is here, and it had this meaning because it was under threat from X other sources. From that, you can extrapolate three or four pieces from the location of that one. But that only comes from the knowledge of what the relative positions of the other pieces actually *means*, rather than simply knowing the literal fact of their position.
So in some sense, they have this midrange “chunk resolution,” where I only have the large and small (useless) chunk sizes because I’m only looking at the physical reality of the board.
This shows up even more vividly in Go, IMO – I’m terrible at it, but because there’re only two pieces, it’s much easier to see how the relative positions of those pieces has meaning – hell, it seems like a huge portion of the game is getting small groups of pieces to resolve to a known, larger quantity. A single piece, on its own, is largely useless until it’s in the context of a larger group of pieces.
So… yeah, I dunno where I was going with this. I guess it’s that you both have to have the hammer – you have to have some specific knowledge/experience to put something into context, and you have to know enough about the rock to see where the cleavage planes are, so you know how to break it apart. If you’re good, you end up with a cut diamond. If you’re bad, you end up with dust?
But that leads to maybe a different point. A while ago, I had this notion that what you’d have to do in the future would be to learn to think in a way that Google can act essentially as a massive brain you can seamlessly interact with. But Google stores, for the most part, literal facts.
It would be interesting to see if somehow, they can organize that data into “chunks” of various sizes, so that people of varying levels of expertise can access chunk sizes that are appropriate to how they think about a particular subject.
So, a search for Van Gogh, for instance, might look at other searches you’ve performed, realize that you have no idea what Van Gogh did, who he was, or what his impact was, and say, “he’s a painter.”
Or if you’ve been searching for Renoir, Matisse, etc., it could put him into the context of those other artists, and discuss his work in more detail because the system knows you have the knowledge (or at least can guess) to actually parse the information it gives you.
Whee!