thought experiment

Transcript of a chat between Alan and me yesterday. Posted with his permission and edited to remove irrelevant elements.


Alan: I had an idea for a thought experiment I wanted to run by you.


me: Oo. Hit me.


Alan: You could get a bunch of rocks and a bucket,
and if all you were interested in doing was keep track of how many things you had at a particular moment,
it serves as a "computer."
similarly, if you want to do some arithmetic you can use an abacus, or for more complex math, a slide rule.


me: Not unlike counting on your fingers, but with the added benefit that it serves as a memory.


Alan: If you keep a notebook, you could even have a rudimentary storage medium, writing down equations and results.
What makes this situation unsatisfactory such that you want to build a computer?
Is there some minimum bound that those tools don't meet but your computer does?



me: OooOooOoo.
I like this!
I will contemplate my answer for a few and get back to you...


me: 

• minimal real time required to reconfigure, reorder, recalculate information
• information density - to - weight ratio
• information density - to - volume ratio
• entertainment value inherent in programmability

Alan: That is an excellent list.
I've experienced the recalculation one personally.
I probably wouldn't have come up with the other ones.


me: The middle two are actually probably the most significant to me,
and it surprises me how mechanical they are.
If there were such a thing as re-writable microfilm, a portable reader/editor could satisfy those.
(Though I suspect that rendering such a device portable would be prohibitively expensive.)
On the other hand,
one of the thoughts on my mind lately has been something along the lines of "what I really want is a portable word processor, but since that's inherently a computing device, I see no reason not to build it in such a way that maximum access to its capabilities is exposed to its owner."


Later...
me: Look at this: http://en.wikipedia.org/wiki/Aperture_card

Alan: How fascinating.


me: I think so.
But, yes. I look at devices like the old Tandy portable word processors, my Kindle, etc. They're amazing little devices, and they'd be so much more exciting if they were intended for hacking.
As it is, there's a whole culture around jailbreaking closed devices, but I think the process of doing that alone is too satisfying. People get to that point and don't use the freedom they acquire to do awesome things, most of the time.
Two reasons for that, I think.

1. The original designers didn't ever have to say "we don't really know all of the ways our customers would like to use this"
   So it's not actually very comfortable to do what you want with most of these objects.
2. Like I said, there's an inherent rebellious satisfaction that comes from opening a closed platform, and a lot of energy gets expended there, goes no further.

Alan: well said.