The speculator's name is Nick Bostrum and he is the "director of the Future of Humanity Institute at Oxford."On reading this, I realized not only that I have been out of college too long and out of touch with some delicious philosophical speculation that is apparently now in vogue. (It must be an exciting time to be a philosophy - cosmology - computer science - artificial intelligence - artificial life - logic - etc. student.) I also realized that I must not be too timid in speculating about the role in inference of something akin to medieval "signs." So I will now say a few words -- but just a very few words -- about this.
Point 2. Reference classes and, more broadly, experience (experimental data, observations, sensory information, etc.) do not "speak for themselves". That is, no algorithm or rule specifies or can specify the import or probative force or any such thing, of multiple sensory inputs, observations, reference classes, etc.
Point 3. Human beings do know things about the world (and about themselves).
Point 3A. A few such things people "know" almost immediately, from the moment of birth. But not too many.
Point 4. People learn things as they grow older and the human species gains, collectively, knowledge (in some domains) as it grows older.
Point 5. Experience with the environment, interactions with the environment, observations, experimental data, etc. are necessary for the acquisition of human knowledge. But not sufficient. (All such information is fragmentary, incomplete, inconclusive. Indeed, the behavior of reference classes is, on its own, mute, wholly voiceless.)
Point 6. The human brain and the human organism [does it follow?] must be attuned, tuned, hard-wired in a very particular and striking way: it (let us over-simplify by calling the knowing entity involved the "human brain," even though much more than the brain is involved [for example, even the retina "computes"]) -- the brain a/k/a the human organism has wired in it some sort of logic, a very sophisticated logic, that lets the human organism read what is signified by the sense experience that it has; and that in-wired, indwelling, logic is more than just some logic that allows a separated human observer to make make sense of a silent inert (or non-inert) and external mass, matter, energy -- silent and external stuff. The logic that the human organism must have must be a logic that in some way resonates with the environment, with the cosmos, with the stuff around it, etc.
So in a sense Plato was right after all. The point, metaphysically viewed, is that we human beings have to work out over time and with experience what we know about our environment (which includes ourselves); we must, let us say, develop the knowledge that we have partly within ourselves and we must must and can develop our knowledge only through experience; and through time and experience we read nature in order to make sense of it -- but nature is a "text" that over time, and with much work and experience by human beings, allows itself to be read, that can be read bit-by-bit.
But it is silly to think that we are close to consciously knowing all there is to know. My guess is that almost exactly the opposite is the case: as yet, we see and understand only as through a glass darkly.Gosh, this is almost enough to make a fellah religious. However, the stuff here is rank speculation. It may even be sheer gibberish. But, ... yet..., wait a minute! Although the thoughts written here may not hang together very well, my strong hunch is that they are not gibberish and that they point the way toward a better understanding of the many mysteries of "evidential inference."