To understand the prehistoric world, historians need to rely on evidence gathered by paleontology, and recreate what happened them by the objects left and the ruins that we found.
Then some civilizations in the past invented writing systems, ways to record what was happened and I think this revolutionized the way to look at the past. Although it wasn't immediate. It took some years to people in the future to look back and interpret the text and the books and the paintings to make a model of the past.
But books are objects with text that were written by humans. I know it is evidence from the past and I know that historians don't rely on single books (when they can) and the model of the past is more accepted if we found non overlapping sources to confirm that some supposition of what happened was correct or not.
But changing from nowadays, more specific to this decade and forecasting to the few days to come. We are currently adding terabytes and terabytes of information that could someday be the information that our future generation will use to understand us. In one sense, it is overwhelmingly large, but also overwhelmingly redundant. In some sense, it is stronger evidence because we have videos all over the place and not just small footage from the media, but rather from everyone. We now live in a world where the smallest contributor of society have ways to register its life and its current content.
I predict that in the near future, cellphone apps like periscope will dominate. It just takes better internet and more reliable software, actually. The internet of things is also going to boom our world with information of anywhere.
Historians in the future will have so much to search that the methodology will definitely going to change. I don't think we can know with certainty how a specific historical event that happened in south america affected eastern Europe back into the years 1000 or less... but I think the future might have a close to perfect representation of today (or our near future) and I think this will change the concept of what is make history.
What do you think about it?
No doubt, since the end of the 20th century large data projects have been an increasing feature of research projects in the social and human sciences. This tendency obviously has its own history and is properly itself an object of historical research.
Many such projects looking at historical questions currently exist, such as: Digital Panopticon, European Fertility Project, and Founders and Survivors Project. Added to that is the progressive digitization of historical publications and even archival material which has 'democraticised' the production of history.
but I think the future might have a close to perfect representation of today
A feature of a lot of the data produced today is that the state has become less central to its production. Thus, questions regarding the ownership of this data and its use will complicate the picture. Moreover, certain forms of personal data, given the passage of data protection legislation in many jurisdictions, are likely to be less readily available to future historians of the present era.
It might enable better, more plausible answers to certain types of questions, particularly, but not only, those most amenable to quantitative approaches; in certain ways, the voices and perspectives of 'ordinary' people also might be more accessible (possibly resulting in a more fragmented type of historical narrative for social and cultural history) - but the notion of a 'perfect representation' of the past is a fantasy.