Drowning in Data?

That probably translates to 12 exabytes of valid content....

Okay, 12 hexabytes then....

I have often maintained that the "value" of information decreases as the volume increases. One can only assign a "value" after the fact- after someone has actually paid for the information, or after someone has used the information to make a decision, the results of which can be evaluated. Couple this with the fact that critical information becomes more and more difficult to access as the volume of information increases, and one can see that the "cost" associated with extracting pertinent data increases. Timely access is pertinent as well, and as the volume of data increases, it takes longer for pertinent information to be accessed, and, after some point in time, the data may no longer be appropriate, adequate or useful for anyone. As an example, let us say one is monitoring a series of temperature or vibration sensors in some sort of plant, with the intent of determining imminent bearing failure, so that bearings can be replaced prior to catastrophic failure. If one has only a few bearings being monitored, it is relatively easy to glance at the raw measurement data once a day or so and identify trends indicating one should consider replacing a certain bearing. On the other hand, if one is monitoring thousands of instances, such "quick glances" are unlikely to result in the appropriate insight (unless, of course, circumstances have progressed to the point where other indicators are also screaming for attention). Thus, one has the added cost of introducing data analysis systems that can automatically extract trends from multitudinous data records, and flag those of particular interest. Now, if our monitor neglects to review the chart recorders, or the data in our automatic recording and extraction system is not providing the critical data in a timely fashion, it is very possible that critical trends have been overlooked, and we get a catastrophic bearing failure. At this point, the historical data has absolutely no value whatsoever.

Wait a minute- let us throw a lawyer or two into the mix, and maybe start a law suit or two to assign blame for some catastrophy, and suddenly this historical data, that has no further predictive use, rises in value considerably...unless the lack of records can be construed as an inattention to detail.

That would be a good place to hide something, and you don't even have to be discreet about it