scripsit

Elias presents ... a worm!    Thoughts on family, philosophy,
and technology

Profile

Friday, February 16, 2007

The Reader




Painting of Aristotle by Chris Howard.

Labels:

Thursday, February 08, 2007

Scratch

Perhaps the best first programming language: Scratch.

Sample code:





When I was in college, a professor was hired whose research area was graphical programming languages, and I remember her presentation looked like this.

Labels: ,

Monday, February 05, 2007

The death of privacy

It surprises me how many smart, technical people will line up to use a full suite of online applications, such as Google's. When is it going to become well-appreciated that the data which we are now offering to Internet hubs, such as Google, is being integrated and stored in an emerging supercomputer -- and someday AI's which we cannot now imagine will "understand" it all, and people will do things with the results which we cannot now predict?

Here's a small example: Ever notice how people will try to encode their email address when posting to a newsgroup, such as "joe@dont.spam.me.hotmail.com"? This is fine for fooling today's email scrapers (let's presume), but how does Joe know that in a few years the email scrapers won't be clever enough to parse his feeble attempt at data-hiding?

The same principle -- that tommorow's AI will crack today's data -- applies to every email or blog post you now write, to any documents or spreadsheets you upload to Google Office, and to your browse and search histories. All your data goes onto the hard-drives of network computers -- which are relatively relatively dumb and inert, now. But in 30 years those computers will not be dumb. They will, somehow, be part of a supercomputer which will parse language and make (artificial) conceptual connections among all of your data; an integrated, subtle, accurate picture of you will emerge. This cannot be in doubt. We have every reason to believe that Google's fundamental strategy is to build that supercomputer. And why wouldn't they (or someone else) do it? There are six billion market forces behind such a strategy. All that is in question is the length of time until that supercomputer exists, i.e., the length of time before privacy as we now know it is dead.

We cannot predict what will be done with the data we are thoughtlessly dumping onto today's servers, but we can predict that someone -- maybe you, maybe me -- is not going to like it.

Labels: