02 CM-1 Design: Why all the blinking lights? – Tamiko Thiel and Brewster Kahle

–So how did you see it?
–I had the task of making the machine live up to the sentence, “You’ve never seen a
machine like this before in your life!” So that was … you can say it was a
marketing ploy, right? So the idea was that people come in and say, “Okay, what is this thing and why should I spend so much money on it? You know, what are these 25 year olds doing that is better than you know, Cray and the big boys haven’t
been doing for decades?” And the line was was going to be, “You’ve never seen a
machine like this before in your life!” And then it was clear! If you … if you
bring them into a machine room and the customer says, “Oh, you know, it’s
exactly like the refrigerator I’ve got at home,” then you’ve lost already. You
know, you’ve got to convince this person that it’s not – that it’s smarter than your refrigerator
(and refrigerators in those days were pretty dumb). So it had to be something that was drop-dead, that really looked like something completely new. And then I, coming from a design
background, growing up with sort of a Bauhaus father architect urban planner,
I had drilled into me from day one this idea that form has to follow
function. No decoration. Decoration is evil, and it has to… design has to really
explicate, and explain the function of the machine. And that was a huge
conundrum, because if I give you a chip, if I give you a board– I mean
probably even you who’ve designed chips, designed boards– if I give one to you, you
can probably say, “Well, here’s the processors, here’s the memory.” But there’s
no way you’re going to be able to tell me, “Well, this is exactly how this chip
works! And I can tell just by looking at it!” You can’t. Because the
functionality is buried in microscopic levels that no human can see, and it’s
probably also complicated and tangled enough so you can’t sit there and say,
well, you know the trace from here to here, and then it goes to
there, and therefore I think it’s doing that. I mean, can you look at a VLSI
drawing or a chip or a board and say, “Okay, I can tell how it works”? –No, not at all.
–Can anyone do that? I don’t think so. –But you can look at the Connection Machine, and see how it views the world. It’s cubes of cubes; it’s got lights that blink that
show things are being active and inactive at different times; and it looks
not like a direct analogy to anything that existed before, but it communicated something.
–And the idea was that it should communicate, “This is an
electronic brain, and this could potentially develop into a living,
thinking machine somehow.” So there were ideas from you, from Carl [Feynman], from
Danny. I know you were the three people that I explicitly can attribute certain ideas to.
And I remember Carl, for instance, talking about he had this fantasy of the
machine as this big cloud of lights that are blinking as they send their
messages to someone else. –In terms of the design of the Connection Machine, since I was on it early days
with Danny at MIT, all sorts of things came and went, whether there was memory
whether there was routers or there were CPUs– The only thing that was ever
constant were the lights. We needed blinking lights. And that was Carl
Feynman. And I think that’s the distinctive design element of this, it was
just a sea of lights that was indicating what it was doing. But it turns out,
that’s not true. The lights would mostly be in the idle loop. And Cliff Lasser,
always sort of the wry one, he called that loop ‘random and pleasing.’ And that’s
become the distinctive look of the Connection Machine. The other
person that did things with the lights– we only had four, the
first four [cubes] running in the lab. So everything was wired up with oscilloscopes and all these people and there was this lab, and Karl Sims came and he
wrote the first computer program that actually ran on the computers itself
when we had four. And it he made it a little ticker tape to bring over words.
And he called us down, and we saw and it says, “I blink, therefore I am.” So that was
the self-recognition of the lights, and the design aspect being so important to
how we thought of the thing that we were building. –So can you be more
explicit about why the lights caught people’s imagination, and why it was so
important that Carl [Feynman] said, no matter what we change, we have to retain the lights? I think it was a… Why the lights? I think it’s the dreaminess of it. I mean, we were doing something much more than just building a
functional thing that would sell well. The idea was to build something that
would be seen as kind of an interesting thing to be near. Something that you would
want to go and discourse with. And computers in those days looked like, well,
refrigerators. Maybe they’d have an on/off switch and a little light that
would say something. But it was long before– Or in science fiction you had the
lights on Star Trek, and lots of switches, but all of those actually on real
computers went away. And so the idea of having the lights come back… and I’m sure
that was just hell to try to design around. I remember the discussions about
trying to make it so it wasn’t just a radio beacon that would just go and flood all of the radio and TV stations in a neighborhood. How do you go and have exposed electronics?
And that was some of the design challenges that Tamiko and all sorts of others
had to go and wrestle with.

Leave a Reply

Your email address will not be published. Required fields are marked *