A tweet (or any piece of social media passing in front of our eyes) is not really information until we think about it.  It's just STUFF.  And we don't think about stuff until it catches our attention.  What we really do with social media is SCAN it.

Alvin Toffler coined the term "information overload" to describe a state of too much, too fast ... a worsening condition that would seem to afflict all of us ...

Americans Consume 34 GB of Data Daily

The concern that we are somehow generating "too much information" goes back many years:

"As long as the centuries continue to unfold, the number of books will grow continually, and one can predict that a time will come when it will be almost as difficult to learn anything from books as from the direct study of the whole universe. It will be almost as convenient to search for some bit of truth concealed in nature as it will be to find it hidden away in an immense multitude of bound volumes." –Denis Diderot, "Encyclopédie" (1755)

And yet, far from scaling back because of "overload", many people continue to opt for tools and platforms that feed them more content faster than ever before.  When given features that enable them to limit and filter that flow, they often choose to broaden their consumption instead.  

(Twitter is a great example - the attraction of the tweet flow (for those who are attracted) is that it provides a single stream to which many people contribute - one place to go to see what's happening.  More sources, smaller chunks, greater velocity, less filtering.)

Here's an hypothesis and a metaphor.  What we do as humans, and descendants of Ardipithecus who stalked the physical jungle, is to scan a complex environment continuously and with all our senses.  Arguably what we are scanning today is no more complex than what our remote ancestors had to deal with: a dense living ecosystem full of danger and opportunity. 360 degrees of sound, sight, smell, touch, kinetics.  All of it is data (STUFF) and always has been.  What digital networks and community interfaces have done is not so much increase the amount of data that we have to scan, as to abstract and thereby extend our senses around the world and around the clock.  

So maybe the concept of information overload is really beside the point.  We can scan a practically unlimited amount of input, which is not the same thing as processing it.  We process only what catches our attention for some reason.  We don't necessarily need filters (it's a matter of taste and lifestyle, really) and any fixed filter will often be abandoned because it biases our view of the environment. Ardipithecus would have found filters dangerous.  

Paleontologists have just about decided that walking upright, not intelligence, was the turning point into humanity.  Walking upright freed up hands to carry things, which made it easier  for Ardipithecus to collect stuff from the environment and carry it back to the group to consume.  (I wonder if this implies that the first useful human constructs were not sharp stone tools but bags and baskets?)

Today what we might find useful is better bags and baskets to carry digital STUFF ... more and better ways to capture and manage the content and stories that do catch our attention ...  flexible, convenient and intuitive "shared personal knowledge management" tools.  

Digital networks and social media have stretched our senses; now we need them to grow our memories and provide a collective workplace for collaboration.