Both possibilities seem unlikely. For a start, it's not necessarily a historian's job to find sources, although that is part of it; it's a historian's job to interpret sources, and assign meaning to them. We already have a surfeit of source material for the 19th and 20th centuries - far more than any individual historian could ever see, let alone use. The massive outpouring of source material that our contemporary world produces will mean more material (assuming most of it survives for posterity - which it might not), which will keep historians in jobs for many centuries - selecting, interpreting, and debating sources.

As to your second point, such a computer model would require far, far more data than we have on, say, the Elizabethan era; there is a huge amount we simply don't know about any given period because it's not recorded in detail in sources - and of course there's the possibility that there are things that were so basic and fundamental to the understanding and methods of thought of people producing sources that they simply didn't record them - we must therefore reconstruct and interpret those things as best we can. Therefore any computer model will require extensive input from historians and will be shaped by the interpretations of whichever historians worked on them.

So both the possibilities you raise will essentially keep historians in work, not put them out of it.

To answer your immediate question directly, not really no. This comes from the fact that the field of history is not just about recreating narratives. In some sense, I think your question presupposes that the job of historians is to cobble together all the information they can find and present it chronologically. Following from that assumption, one could conclude that with publicly-accessible, detailed records the role of historians would be less important or replaceable by advanced algorithms. However, I think that conception of the historians profession is somewhat flawed (as the other response to this question indicated as well). Historians are in the business of making arguments. Although the phrase 'the evidence speaks for itself' is sometimes taken as a truism, it's actually simply false. The evidence doesn't speak for itself, the historian speaks for the evidence (by making an argument). In other words, if you amassed some kind of ultimate timeline of all things and all people in all places across the entire time humans have existed, you would not really have a "history" of humanity the way that academic historians use the word history. Even ignoring the the sheer size of such a dataset and how we would actually define its precise parameters, history is more than just a record of the past.

I think the most succinct version of this that I've ever read comes from E. H. Carr's classic "What is History" in which he says that history is "an unending dialog between past and present." In this sense, history can never be finished, it will never be complete, even if we had that fantastical record posited above. That is because that timeline would have different meanings to different people in different places and times, and that meaning would be up for discussion. History is, in this sense, a discourse. It is a discourse that is based in verifiable evidence (aka "sources"), but it is a discourse nonetheless.

With regard to your discussion of the mass amount of data available to historians of the future, I am reminded of a post I made a couple years back here:

https://www.reddit.com/r/AskHistorians/comments/1y2f9d/is_twitter_oral_history/cfgro2s

I'll quote it at length because I think it's particularly relevant.

Historians are going to have an unprecedented amount of data/sources to work with when they/we study this era. I'm sure there will be methodological debates over what exactly to do (they've honestly already started, especially in the Digital Humanities.) There are a host of issues with working with digital sources that we've really only begun to fathom.

Benedict Anderson gave a talk called "Letters, Secrecy, and the Information Age" in which he talks about a great many things, not the least interesting of which was the idea of working with letters vs. e-mail both from the perspective of the writer and the perspective of the historian. He laments the loss of letters as historical sources personally. He is particularly concerned with the idea that letters have a personal quality and leisurely, careful pace, while e-mails sort of demand action from the user which leads to a different quality of the content.

The question is ultimately, though, what do we do with all of these digital "sources"? I'm not sure - especially given mass anonymity - how historians will use them for building narratives. Their use in trying to understand discourse seems valuable, but perhaps by the time historians are looking back at the 2000s serious discourse analysis will seem a quaint methodology of the past.

In this sense I think the future historians are going to have a methodological problem to solve, but will not face some fundamental crisis in the field. Historians and other scholars in the humanities have already begun thinking about ways to use computation to deal with huge data sets. This has sometimes been called the "computational turn" in methodology. My feeling is that this way of dealing with data is still in its infancy when it comes to understanding history and I frankly look forward to historians finding new and interesting ways to incorporate it into their work. I also remain at least a little skeptical that enough is being done to preserve all this data that it will, in practice, be available to future historians. It certainly might be, and I hope it is, but there are only, at best, vague attempts to systematically keep all of this data available in perpetuity.

On the question of computers, historical recreations and what I think ultimately amounts to question of whether or not computers can replace historians, it seems to me that history is not in any particularly special place here. I'm not really sure what dataset you would use to create a perfect VR simulation of victorian England, even assuming you had a program/algorithm theoretically capable of doing it, but even assuming you had it, I think that would provide, at base, an abstracted data set that could also be worked with and wouldn't rid ourselves of present bias. To put it another way - if you put me into a perfect VR simulation of a particular time and place, I would still be experiencing that simulation as someone from the 21st century. The question of what it was like for a person living there in the first place would still be up for debate and would rely on traditional sources.

I'm open to the idea that AI is going to come along eventually (I'm in the camp that thinks far-future is more likely than near-future, but I could obviously be wrong), and it's possible that AI will surpass humans in intellectual capability. But I think that given the definition of history that I've tried to lay out at the top of this post, this is what we'd have to wait for before human historians become obsolete. Even then, we'd have historians - they'd just be AI historians. I also suspect that any AI capable of doing the work would bring its own bias to the table, so I don't think the question of bias comes off the table when it comes to AI.

I would offer up a challenge to the assumption that our born digital source material will necessarily offer a vast pool of source material for future historians. The topic is debatable but it is quite possible that we are currently living in a digital dark age (see https://en.wikipedia.org/wiki/Digital_dark_age) where the records of our communication will be lost to future generations. I don't wish to detract from the intent of the question too much here, but there is a real chance that a modern historian (say in the year 2060) may have the problem of too many sources to peruse while the historian in 2160 will have virtually no source material for our present age.