Discussions with historians today about the "future" history of now are pretty speculative, but it isn't at all clear that the rise of digital communications is going to make that for the better. Why? Because much of that data is 1) proprietary, 2) ephemeral, and 3) lacks contextualization.

What do we mean by that? Proprietary means that a lot of the data being generated is owned by private companies. This means access is not straightforward, and such datasets often die with the companies themselves. They rarely end up in archives.

Ephemeral means that the data being generated is often not being done with any eye to the long term. It is not being stored on media with an eye for preservation, it is only being stored as long as it is being used and useful. Consider how difficult it is to access a webpage you saw 5 years ago — half of the time you're going to find it isn't there, or the site revamped its design and buried the page, or whatever. Maybe Archive.org will have it, maybe not. Now take that problem that multiply it by a million for any kind of content that is not being systematically archived.

Lacking contextualization means that we have a lot of "raw data" that doesn't tell you much about who made it, the conditions under which it was produced, or even, at times, how to interpret the data itself. There is a tremendous amount of opacity with regards to the data being generated by social media, for a variety of reasons, including but not limited to privacy, proprietary standards, and just a general lack of interest in preservation by the companies involved in its production.

These are separate from the question of whose data is being generated versus whose is not (it has always been much easier to replicate the lives of affluent people in affluent societies; will these forms of data generation merely replicate and magnify those source biases?), or whether any of this kind of data will answer the kinds of questions historians will have about the past.

Separately, we might also ask about the forms of media that are basically being subsumed by digitization trends, and the increase ephemerality of communication. Paper lasts for centuries with very little upkeep, if kept under good conditions. When you communicated to someone in much of the recent past, you did so by writing down your communication into a very easy to preserve format. Today you use e-mail, which often gets wiped with new computers or accounts. Or worse, SMS messaging, Snapchat, or whatever it is the kids are using today. These things are unlikely to be preserved for centuries.

Historians of science, I know, worry about this latter problem a lot. We know communication between researchers is a crucial part of how science as an organized activity "works." Correspondence of famous scientists serves as a "bedrock" for many historical claims about how they did their job. In basically a generation that has moved into very ephemeral media. E-mail also mingles formal with informal communication to a very high, arguably embarrassing degree. Are scientists of the future going to save their e-mails and make them accessible to researchers? We don't know.

Lastly, we might consider that the overwhelming volumes that come with these Big Data sources are not necessarily going to be useful for answering the more particular, individual questions one might have about the past. Wading through petabytes of data isn't easy, much less doing so in a way that answers the complex questions historians like to ask.

I don't want to make this all doom and gloom — I'm sure historians of the future will figure it out, and there may be opportunities in these data sources that are unknown to us as of yet. (And hey, I benefit from digitization a lot! Digitized archives and newspapers make up a huge amount of my research sources, and I don't long for the days of microfilm in the least!) But I do want to push back on the narrative that this will somehow make things easier or better. There are always ups and downs to every kind of technological change.

(For a nice critique of Big Data, much of which can apply to this question, dana boyd and Kate Crawford's "Critical Questions for Big Data" is pretty useful.)

The rise of social media, the internet, and technology in general allow us to capture about life than ever before.

It's worth noting that the selection bias involved in this, already, is incredibly misleading and distorting for the more bleeding edge "histories" of recent events that have come out in the digital age.

I'll give you an example that I'm familiar with: the Arab Spring in Egypt, 2011-2012.

There were some hot takes published at the time that called what happened in Egypt a "Twitter Revolution." (Including of course Thomas Friedman who is apparently constitutionally unable to resist shitty technobabble hot takes). This was and is nonsense. The population of twitter users in Egypt at the time never represented anything more than a fraction of the population of the opposition to Mubarak, let alone the Islamist voices in the country and even less so the forces of reaction that we now would eventually reassert control.

Before ISIS got savvy about twitter my old boss used to joke that if the population of the Arab states held the views of the Arabic speaking twitter user population then they would be the most secularist and liberalist states on the planet. An exaggeration of course, but this reading of what was happening on the streets as a reflection of what was going on in social media just completely distorted perceptions of what was going on, and I think to some extent still does.

Among other things it totally leaves out the role of the Muslim Brotherhood, which by design did not want a prominent public leadership role in the events but was no less a driving part of what unfolded. They didn't organize through facebook or twitter, but through the mosques and their usra cells. So we may well never know what precise role they played, but it can be plausibly speculated that they may actually have done the brunt work of toppling the government by putting the kind of actual headbreakers out on the streets who attacked police stations, and mounted mass jailbreaks, and set government buildings on fire and that kind of thing.

It also ignores the way that actual, real life social networks and political activism played a role. Sure there were protest hashtags, but far wider reaching were the mass-SMS messages calling people out to organize. People going through neighbourhoods with bull horns. Graffiti. Egyptian domestic "legacy" media, which is still robust enough to put out morning and evening editions of some of their daily newspapers. And so on and so forth.

None of that is captured in social media.

---

And yet for all of that I would be willing to wager that the history of those events 100 years from now still keeps big sections for twitter and hashtags and facebook. Why? Because the material is available, easy to access, and can be saved. I think the sheer volume of available material is going to privilege those kinds of sources in a way that will be just as much of a historiographical challenge as the lack of sources has been in other eras.