I think we'll find that 'big data' is simply too encompassing. There will be the desire to again break into the independent practices of storage and analytics.
I believe we'll come up with a term to scale beyond 'data warehousing' for storage and access. And "data mining" along with "predictive modeling / analytics" will sustain as very accurate and fitting terms for that practice.
My next prediction is the very near-term death of "data scientist", for several reasons -- - at least in commercial circles:
1. The vast majority of analysts are not truly "scientists." And I'm not sure that they even aspire to be.
2. "Scientist" carries too much academic stigma to be commercially attractive. People are not failing at analytics because they're not building technically adequate models. They're failing because they're not being effective pragmatic business practitioners: evaluating the environment; communicating effectively with leadership; preparing a solid project definition before throwing data and software together; not translating results in terms that are actionable or understood by leadership, etc. etc.
The experience required to make a "Data Scientist" commercially viable do not align with the standard definition of a 'scientist.' It dilutes the formal discipline and focused practice of a true scientist, and is off target strategically for the broader skills needed to make the practice succeed commercially.
"Data Science" and "Data Scientist" will evaporate faster than the 'millennium bug'!