(This started as a comment to the previous post and grew into a post of its own. Thanks Phil, Jay and Jane.)
I think Jay is right to point out that there will be limits to what we can learn. Historians of the future, however, may gain access to the "Rank Page" algorithms of our time. These will be telling, i.e., they are in many ways decisive.
For example, it will help us to understand the sense in which George W. Bush epitomizes the concept of "miserable failure" (I'm sure you've tried this) and, of course, the limits of that application.
Jane brings up a vital issue. I like to think of myself as a technology conditionalist, not determinist. That is, technology conditions both aesthetic and historical progress and this, in a sense, is what makes aesthetics less conditional on history (i.e., the historical epoch of a particular aesthetic) than we think.
No understanding of X without an understanding of the technology of X.
But, no, I am not saying that technology, in the narrow sense of "how the equipment works" (which is the sense I've been trumpeting here), will answer all the relevant questions.
Even in this narrow sense, however, (cue trumpets!) technology has a way of summarizing and stabilizing social conditions. What I am suggesting is that the physical properties of texts (how they work, again in the narrow sense: their readability, durability, searchability, editability) are currently rendering substantial habits of cultural sensibility (the sort of thing we learn in our humanities courses; the sort of thing we teach there because we don't know enough to do otherwise) are now, or will soon be, peripheral to the real enterprise: "the process now going on," as Pound might say.
That is, aesthetics, along with history (but not "because of" this association), is becoming increasingly conditional on brute technical processes.
The phenomenal distance or ontological difference between our humanity and our technology (what we normally experience as "consciousness") is shrinking. The interface is being perfected.
It will never be perfect.
As the semiotic properties of texts are linked in ever greater detail to their material properties, their social properties will follow along. Indeed, semiosis just is the interlacing of social and material properties. What we are witnessing with search engines, blogs, wikis and Flarf, is a remarkable development in the mediation of social and material processes, cultural and natural events.
Picking up on Phil's enthusiasm here, I find myself wanting to announce (as if I'd be the first) a new linguistic era: from orality, to literacy, to "hyperliteracy". That is a word worth Googling for origins, but for me it simply means the ability to read and write hypertext--to see, e.g., a blog post as an occasion for further Googling (on the reading side) and to write a post knowing it will be read in this way and, of course, using links, comment boxes, etc.
I would caution against using words like "ubiquitous" here; as with literacy, progress belongs always first to the few. Which is also basically a reprimand to myself for all this (bordering on) techno-ecstatic hype I seem to be caught up in right now.
Coming back to Jay's point (and Jane's), yes, there will be old fashioned forms of power and ignorance to subtend these developments--just as literacy did not mean the end of oral communication.