-- Edited by Jason on Friday 3rd of April 2009 02:39:38 PM
now if we can teach people the difference between "data" and "datum", we'd be set.
that's an impressive program: it can derive a relevant equation for a given system. however, it can't make its own measurements, it can't interpret its results, and it has to hope unrelated data with similar underlying mathematical models aren't fed to it. for example, newton's law of gravitation can be derived from gauss' law of electromagnetism. if the program saw data described by both of these laws, it might freak out about the differences in sizes of proportionality constants (newton's big G vs. coulomb's constant) and the fact that mass is unsigned (gravity never repels mass). it might think that mass is a charge, and a very feeble one at that. it might think that negatively charged particles have negative mass.