Feeds:
Posts
Comments

Posts Tagged ‘performing’

As far as I understand it when a computer generates a random number it uses the current time value from its built-in clock as the starting point (seed) in a mathematical formula which generates reasonably randomised values.  They’re not truly random because if the formula is started again at the exact same moment, resulting in the same seed being used, the same result will be generated but the formula. 

Whilst this problem is avoided by the handiness of the linear nature of time and us therefore never repeating the same moment, it does raise interesting issues.  Will computer generated improvised music (generated solely by the computer, not just preprogrammed synthesis) ever be so convincing that we wouldn’t be able to tell the difference?  Will nature of programming and randomness in computers allow them to?

Algorithms such as A.L.I.C.E. simulate responses to enquiries in a somewhat realistic way, but fall far short of earning a place at the table of your ideal dozen-strong dinner party.  Will similar algorithms in the future be able to improvise music in the same way they might improvise conversation?  And if they can, how capable will they be of taking a leading role in the music by with a sudden impulse decision (or in their terms random decision)?

Relevant Wikipedia articles:

http://en.wikipedia.org/wiki/Random_number_generation

http://en.wikipedia.org/wiki/Random_seed

http://en.wikipedia.org/wiki/Pseudorandom_number_generator

A.L.I.C.E.bot

http://www.pandorabots.com/pandora/talk?botid=f5d922d97e345aa1

Any thoughts? Preferably in an over-the-top grandiose Tomorrow’s World style?

Read Full Post »