AI concept of the '80ies ... why?

[PhilD]

chacal_ said:
Radioactive source seem to be a good source of true random
==> http://www.random.org/

Yeah, and electronic noise should be another good one, potentially with higher rates. If only we could get some electronics into computers...

I did some quick searches on the subject (as a side line to some "randomized algorithms" lectures I give), and I seem to remember reading that CPU makers (Intel et al.) typically have the means of including pretty good, unpredictable, irreproduceable random number generators in their chips, only it doesn't look like the demand is really there (since space on chips has to come at the expense of some other functions, the integrated RNG has to compete with other, chromier stuff that will make them sell a few more units.