Historians distinguish between remnants and traditions. The first designates sources not meant to be found by future generations, while the latter are (at least partly) created for that purpose. Caesars’ De Bello Gallico in which he described his wars in Western Europe were an incredibly effective propaganda campaign, not only written with the intention of making him popular, but also for immortal fame. The, just recently discovered, postcard written by Hitler in 1916 in which he tells a friend that a doctor just pulled out 19 teeth (probably exaggerated), was most likely not meant to be read by you and me.
For centuries, historians had always have to deal with having were few sources. For example: All remaining writings from ancient Greece would probably fit in your bookshelf; it would not have to be a big one. It is estimated that we possess only about 3% of all written books which we know have existed. And there is an unknown number of writings, potentially of breath-taking importance, which we do not even know the titles of. For ancient literature, the early middle ages proved to be a bottleneck. For example, we have writings from Aristotle and his scholars which only survived through one copy.
Today, every semi talented philosopher spreads his thoughts through the internet or, with clever marketing, is able to sell hundreds of thousands of copies. In times of big data, how will historians in the future make decisions about what is important and worth analyzing? Let me be clear about this, the question is not about making the RIGHT decision, but about making ANY decision.
Take an historian of 2114 who attempts to write a about the South-Western English philosophy of the web-based epoch of 1990-2040.  She will face unprecedented problems. She won’t be able to even skim read every potentially relevant book, article, facebook or blog entry, tweet (who knows how it develops?), and what not. How will she be able to make an informed decision?
What we have to do now is two things:
First, we have to acknowledge the fact that demands to historians have changed. There have to be historians with high level abilities in
programming and informatics to address these questions. Just as in the 19th century, when historians acknowledged that some of them had to become advanced economists.
Second, we must appreciate the enormous chances given my the above described developments. If you allow me, I would like to return to my example of antiquity and quote the British comedian Eddie Izzard: ‘We love the Romans, we love what they did. They murdered a lot of people, but we don’t know their names so it’s okay now.’  In fact, we also do not know the names of those they enslaved and even of the wifes of some of the most famous men. Hardly anybody in Heidelberg, where I studied, knows that there once was a Roman town, because we do not even know what it’s name was. If they had written on just one percent of all the pottery they produced, something like ‘Made in Heidelbergus’ or whatever, we would know – but they chose not to care.
By now, these problems of filtering have shifted. We do much less filtering of information than for example the Romans did in the past. I do not know anybody younger than 50 who does not have a facebook account for example. Future generations will probably have all names and hundreds or thousands of pictures of everybody from the Western hemisphere. Problem is, it is physically impossible to look at only the tiniest fraction of it – it is just too much.
If we want to help the past understanding and learning from us, we have to help them. Think of ourselves as a foreign country and ask, what questions foreigners would pose to us. Let us answer them and decide about the priorities of the remnants we want to leave behind. Let us shape our legacy.
 The picture shows the author at Guédelon, the famous project in the field of “experimental archaeology – a kind of open-air laboratory.” http://guedelon.fr/en
 Famous opening line of the novel ‘The Go-Between’ by L. P. Hartley (1953).
 Another problem: In what epoch do we live in? Epochs have been defined on the basis of culture (Renaissance), economy (Industrialisation) or politics (Cold War). In my opinion, our current epoch will (if humankind survives long enough) be labelled on basis of technology (e.g. web-based epoch) or climate (e.g. greenhouse age). The dates 1990-2040 are a wild guess.
 Eddie Izzard at the Secret Policeman’s Ball 2012: http://www.youtube.com/watch?v=X09SNFdfoZk