There is no reason why time of the day, temperature outside, whether Obama made a joke or not, etc can't all be features for a classifier to learn.
I love the idea of an algorithm anticipating my next move and I'd argue that for the vast majority of people you can find repeating patterns that are almost always relevant. These patterns will not be simple or obvious and as you currently pointed out, the lack of data is what is holding most innovation behind. I feel that companies that own the data like Amazon will have a lot to gain if they do a Netflix-style competition to improve their recommendations engine.
I find it surprising that they did not and do not plan to employ any sort of interlocks/padded walls. What I mean is, if the system is exhibiting some very abnormal state (e.g #remirror_event above a fixed threshold or more than x standard deviations above average) then automated repair actions should probably stop and the issue should be escalated to a human.
Problem with RAID is not just drives failing, but also the UBE (Unrecoverable Bit Error). Lets say you have a RAID5 configuration. In addition to the probability of disk failure, you need to account for the probability of some failure plus the controller's inability to repair the problem because a sector on the good disks experienced an UBE. The probability of UBE on enterprise level disks is rather low (10^-16 I think) but that quickly shoots up if you start considering 1TB and 2TB drives.
Problem with EBS benchmarks is that they largely depend on who else is sharing the spindles with you at the time the benchmark was run. Given the large variance in performance that is being reported, the sample size for some reliable statistics would need to be quite large
That is not what the researchers suspect. The article misquotes the study. You only have to go as far as the abstract to find: "Regardless of the amount of sedentary activity reported by these men, being older, having normal weight, being normotensive, and being physically active were associated with a reduced risk of CVD death."
I don't think he is saying that. He is saying that in a non-dedicated environment, you share the same spindles with other tenants who may have different I/O access patterns than your application. Careful choice of indexes, good data locality for fast reads, making sure writes are sequential - all that goes out the window if some other application is causing the disk to seek all over the place.
I love the idea of an algorithm anticipating my next move and I'd argue that for the vast majority of people you can find repeating patterns that are almost always relevant. These patterns will not be simple or obvious and as you currently pointed out, the lack of data is what is holding most innovation behind. I feel that companies that own the data like Amazon will have a lot to gain if they do a Netflix-style competition to improve their recommendations engine.