Saul Griffith
U PRSOEDAUHCEEAINDS-MTORUUNCT TIEODNSVIDFOEO R C MAAMKE INRGATTHOINGS.
One can now picture a future investigator in those aren’t the ones that really count. They’re
his laboratory. His hands are free, and he is not high production value (and high cost) and have
anchored. As he moves about and observes, he been labored over by teams of designers. In a long-
photographs and comments. Time is automati- tail world, the instructions you’d like to be obvious
cally recorded to tie the two records together. If are your auntie’s Roomba hacks, or the Tron guy’s
he goes into the field, he may be connected by instructions on el-wire Halloween costumes. The
radio to his recorder. As he ponders over his notes problem is that documentation is generally so
in the evening, he again talks his comments laborious that it is rarely complete or well done.
into the record. His typed record, as well as his I listened keenly to the observation by Neil
photographs, may both be in miniature, so that he Gershenfeld (of MIT’s Media Lab) that the best
projects them for examination. moment to capture documentation is immediately
—From “As We May Think” by Vannevar Bush, after someone has just successfully done some-
Atlantic Monthly, 1945 thing for the first time. It’s at that moment that
M people are high on success and wanting to dance
Y CURRENT OBSESSION IS around and show the world how they did it. If you
meta-making, or the documentation can capture that energy toward writing instructions,
side of how-tos. A source of inspiration for you’ll get the best how-to. The other element that
my thinking is Vannevar Bush’s seminal essay “As makes instructions top-quality is images, and plenty
We May Think,” which has been an influential piece of them. When inserting a 4/40 left-hand threaded
in computer science since it was written in 1945. hex bolt into the diaphragm pump housing of your
Most people read it for its incredible foresight as to washing machine, no number of images is too many.
what the internet would become. I’ve been thinking a lot about what can be done
The part I love to latch on to is the “memex,” a now to make how-tos a no-brainer. Here’s the prob-
device Bush describes that links researchers every- lem: you want to capture the details as you do it,
where and the contents of their notebooks so that and probably the first time you do it. Second time
documentation is seamless, and negative results are around, your hands move fast, and you’ve figured
registered as often as positive results. It sounds like out the hard bits, so they don’t seem so important.
my fondest dream: it’s all wireless, hands-free, and The other big problem for me is that I generally use
auto-documenting. But that’s still the future. both hands for building things. I often use my mouth
Elegant instructions are quite rare. There are a lot too, as two hands aren’t enough, and the mouth is
of how-tos out there, in books, on websites, in the such a versatile tool for holding screws and gripping
instructions for toys and washing machines, even in things. Cheek pouches full of pop-rivets — that’s my
the seat backs of airplanes, but few are truly great. type of squirrel. So with mouth and hands occupied,
Lego’s original building instructions are good, but stopping to take the photograph is damnedly hard.
