Not a specific vision on how things will change, but I did wrote a few years ago a paper on the potentials of Twitter as archive when doing my history bachelor. My overall conclusion was that it would of course not give and exact reconstruction of the facts, but it does give an insight in the way people responded on certain events. Therefore giving an insight in the 'zeitgeist', the cultural climate of a certain period. For researching this you would have to use other tools than standard historical research tools though, because of the vast amount of tweets. An example of one of those tools is emotion mapping: tweets cornering a certain topic can be filtered by the emoticons that are used (happy, sad, shocked etc). With that an idea of how people generally felt on certain topics can be created.

There are two points I see that would present particular problems to future historians in trying to draw anything useful about contemporary times from social media.

First, there's a real concern that little to no digital data from our time will actually be accessible to historians of the future, due to changing technology formats and compatibilities. Because of this, there are multiple efforts within the archival community devoted to trying to preserve and archive digital data, but it is an ongoing struggle and must be continuously maintained due to the constant progress of technology - new operating systems and programs often cannot interact with things designed for older ones. The most reliable way to archive digital data is actually to recreate it manually in every new system, akin to Medieval monks copying old manuscripts, which is obviously time-consuming and only done for the most important documents.

Websites like social media are even trickier, because all their data can be lost as soon as the website goes out of business, so it must be archived separately ahead of time and stored as image files, which archival communities are doing, but this means it also becomes subject to the same digital file compatibility issues as other programs.

Assuming they could be preserved for long enough to be of use to future historians, there's actually another problem. Depending on what source you trust, anywhere from 30-60% of social media accounts are actually fake accounts used for various purposes, and thus do not represent "real" opinions. So there would have to be some sort of methodology to sort between them, or future historians could arrive at all sorts of erroneous conclusions about our society - much like the people who mistake the DPRK News Service parody twitter for the real thing.

It will become easier, but also harder for historians.

Yes there will be more 'resources' available, but the accuracy of these could vary. I believe more procedures/practises, if you want to call it that, will need to be thought of. In regards to this it could be the difference of 'official' news sources or your neighbour vocing his/her's uneducated opinion. Of course this information would be good for research on what the public thought of a event etc. at the time, but say if the research was on the effectiveness/impact of vaccines, then the information base could become diluted with inaccurate studies or opinions that come from social media posting and sharing.

I guess time will tell

Please correct me if I'm wrong, but does this not go against the 20 year rule? Or is it too relevant historically that it overrules it?

I'm just curious, because it is a fascinating question and one that I haven't really seen discussed at the depth and thoughtfulness that this subreddit requires.

And what will historians think of privacy? All kinds of information and correspondence is highly protected these days but it is ok to read love letters from 300 years ago and write papers about them. What do you think when today's privacy laws are "past the statute of limitations"?