Comment Exceptionally weak typing of data ??? (Score 0) 75
IBM technologists approached cancer medical centers, but "were frustrated by the complexity, messiness and gaps in the genetic data at the cancer center...
This sounds like a problem with data typing and attempting to translate between differing and likely mutually-contradicting [if not outright self-contradicting] weakly-typed data sets. Actually, if they wanted to do something interesting, they'd build an AI for classifying poorly typed data and then see whether the AI could make a best guess at matching up different pieces of data without necessarily having to shove square data into round bucket holes for data. But do not kid yourself, weakly-typed data spread out over databases from every manner of competing vendor is a phreaking disaster to try to work with, and will tax the patience of even the smartest & most experienced of human programmers. And then there's the problem of multi-dimensional data - no one can ever possibly hope to collect enough actual real world biological data to fill an entire multi-dimensional hyper-rectangular grid of data, so inevitably you're left with just a two-dimensional rectangle here and there, and no sense of how to fill the catastrophically overwhelming number of points on the grid with no known [or even knowable] data to assign to those points. [And that's about all I can squeeze into my allotment of two "P for Paragraph" tags.]