It depends on what you mean by "infinity", "exist" and "the physical world". Sorry if that sounds like a coput, but I'm a mathematician.
You know, I do happen to have a math degree myself (though I'm probably not nearly as knowledgable as someone like yourself), so I'm somewhat interested in this discussion. Speaking as a physicist, though, I might make a slightly stronger statement, and say that infinity does not exist in the physical world. In means much the same thing in science as it does in mathematics: it refers to a limiting process.
Also, one could claim that the universe has infinite length. As in it has no start or finish. We do not have any way to prove or disprove that at the moment either.
Actually, it's commonly believed today that the universe is finite in size. There are even estimates on the mass of the universe. So it's entirely within the realms of cosmological theory that the universe is finite in size. As far as I can tell, this is in fact the most likely possibility.
You know there is no such number that is smaller than any real number. I thought you said you're mathematician...
That's why I said "they're not in vogue". They had to be posited and used in the early development of calculus, but even then mathematicians realised they weren't really a tenable concept. That's why there was a century of refinement in order to reformulate the foundations of calculus in terms of limits.
Interestingly, I happen to have an old calculus textbook (=from the sixties) called "Calculus: An Infinitesimal Approach." As Marcus said, many of the early pioneers of calculus, such as Newton, Leibniz, the Bernoulli guys, etc., genuinely believed in infinitesimals. In the next century, Weirstrass formulated the rigorous "epsilon-delta" limit definition, which eliminated the need for infinitesimals, and in fact modern calculus courses do not even entertain the idea. But this textbook of mine revives the idea, and introduces the concept of infinitesimal numbers. Apparently the idea
can be rigorously formulated.
Not that physicists care about rigorous formulation. All this time we've been under the impression that the familiar
"dx" from calculus is a bona fide number.
