I could go on and on about the failings of Shakespeare and the constitution and Stradivarius violins, and at the bottom of this post I do*, but really I shouldn't need to: the Bayesian priors are pretty damning. About half of the people born since 1600 have been born in the past 100 years, but it gets much worse than that. When Shakespeare wrote almost all of Europeans were busy farming, and very few people attended university; few people were even literate--probably as low as about ten million people. By contrast there are now upwards of a billion literate people in the Western sphere. What are the odds that the greatest writer would have been born in 1564? The Bayesian priors aren't very favorable.
This argument is fantastically bad; in Bayesian theories of evidence, priors can't be 'damning', by definition; the whole point of a prior probability is that it will change (be updated) with new information, so you can't dismiss anything based only on its prior, by definition. Indeed, famously we know that, allowing certain easy assumptions, no matter how bad the prior probability for a claim is, if it is larger than zero, there is some new evidence that could in principle require an update to a high probability that the claim is true. What is more, Bankman-Fried does not assess the prior probability correctly; the prior has to take into account all relevant information already in possession, but Bankman-Fried has deliberately refused to include as part of his evidence the widespread testimony to the excellence of Shakespeare. Third, he misconstrues even the basic probability problem. The question is not, "How probable is it that the greatest English writer up to now was born in 1564?" The actual problem requires a more interesting reflection. The probability that candidates for the greatest English writer up to now were born at some point in the past is one; it is absolutely certain that anyone who can be considered one of the greatest writers up to this moment has written before this moment. It is also certain that any candidate for the greatest English writer up to now wrote at some point after the development of something identifiable as English that has literary remains; the earliest of these (and the first great works in the history of English literature, like Beowulf) are in the seventh and eighth centuries. Therefore, we know for a fact that the greatest English writers up to now, who must by definition already have been born, lived somewhere during those thirteen centuries. A title like the 'greatest writer' is going to depend on how many great writers there are, not directly on the general literate population; even if we assume that there are more great writers in the twentieth century than in prior centuries, the probability that serious candidates for the greatest writer would have occurred at some point in the twelve centuries before the past century rather than in the past century (which is what Bankman-Fried is really arguing against) is actually very good. And, again, this is all before we have even considered the fact that he constructs his prior probability incorrectly by deliberately ignoring evidence.
Huemer, unfortunately, does the same thing in his blog post:
As SBF said, it would be surprising on its face that the greatest writer should have been born in the 1500’s, when (a) there were a lot fewer people in total than today, (b) people had much lower literacy, and (c) they had much less education.
It is not in fact "surprising on its face". As with the above, the relevant probability problem is not "How probable is it that the greatest writer should have been born in the 1500's?" but "How probable is it that the greatest writer should have been born at some point before the recent past?" Also, as noted above, this judgment of the claim being "surprising on its face" deliberately ignores the evidence of widespread testimony that it is so, and therefore is only getting the surprisingness by gerrymandering the evidence to get it.
Huemer gives additional arguments, all of which are bad. The first:
People have a bias towards old things in other areas. E.g., if you ask who the greatest thinkers are in almost any area (unless it’s a new field, like computer science), people almost invariably name figures from long ago.
If you ask who the greatest thinkers are in any area at all, people will always give someone whose work is not in the future; that is to say, the question is specifically about the past. There is no given measurement for "long ago"; is it Aristotle-long-ago, or Beowulf-long-ago, or Shakespeare-long-ago, or Mark-Twain-long-ago? Without knowing anything more specific, all it means is an indefinite "non-recent past"; and since the argument explicitly identifies a kind of exception, it reduces to claiming that people will say the greatest in a field is from the non-recent past, except occasionally when they will say the greatest in a field is from the recent past. There is nothing in this that is capable of identifying any kind of bias.
“Being the greatest writer” is not a property that you can just see. There is no proof of someone’s “greatness”. It’s a matter of highly subjective judgment, which is exactly the sort of thing that would be very easily influenced by bias.
"Being the greatest writer" is obviously comparative, and obviously in this context is not put forward as a merely subjective but an intersubjective judgment; as people in aesthetics have repeatedly noted, aesthetic judgments of this sort are claims about what will tend to be judged by people familiar with the points of comparison, especially (since it answers Huemer's bias point) where those people have different backgrounds. Appreciation for some authors is mostly confined -- to a place or time or culture -- whereas for others, like Shakespeare or Austen, once it picks up it becomes stable and widespread, and also crosses cultural lines. To say that someone is a great writer in English is to say that they can be appreciated by Anglophones who enjoy reading, Anglophones of all sorts, Anglophones whatever their cultural assumptions. This does not eliminate all possible bias, but it makes it relatively insignificant, in the sense that the judgment is tested in a context that means that its dependence on at least a very large range of possible biases is not throwing off the results. As to any biases that it does not cover, they are going to have to be extremely common, and it is not in fact an argument against Shakespeare being the greatest writer of English, a human language, that vast numbers of human beings of all sorts of human backgrounds are inclined for human reasons to value what he has written, even if you do call the reasons for it 'bias'.
Huemer's third argument is the closest to a good argument; he argues that the reason why the 'greatests' tend to be from the "distant past" is that reputation builds. The thing of it is, sometimes it does, and sometimes it doesn't, as previously noted. Some reputations fluctuate wildly, some diminish and never really recover. But the building reputation of an author means that, at the very least, people who like reading keep discovering things about their writing that they like, and pass them down to people who, not finding such assessments inconsistent with their own experience, pass them down. Again, Huemer is taking the intersubjective nature of aesthetic evaluation as if it were proof that it was more biased rather than (as it really is) proof that at least individual biases are not actually behind the evaluation. The evaluation is still defeasible, perhaps, but nobody has ever claimed that they had demonstrative proof of it from first principles. It's a very well established evaluation, one people of all kinds and backgrounds keep making in all sorts of contexts.
And all of these arguments are essentially irrelevant, in any case. The best criterion for greatness of writing is that given by C. S. Lewis in An Experiment in Criticism: if lots of people who like reading enjoy reading and re-reading a work, and keep doing so, it is a great work. Likewise, an author who delivers multiple such works is a great author, and if such people keep finding some such authors to be most excellent in this way, they are candidates for being among the greatest of their kind. Whether Shakespeare should strictly be considered 'the greatest writer in English', I don't know; there are different dimensions of writing, so it's likely that different authors can be 'the greatest writer' depending on the dimension one primarily considers. But none of these vague and poorly formulated probability games are very relevant to the question, and they certainly aren't enough to give definite conclusions, whereas widespread testimony is at least good enough for a presumption.