Peter Atkins, now retired, was Professor of Chemistry at the University of Oxford. In this short interview, he says that 'science is the only way of learning about the nature of the world'; and, again, that '[t]he scientific method is the only reliable method of achieving knowledge'. The claim appears to me to be either definitionally true but at the same time useless, or else false. One could, of course, just call any human activity that yields knowledge about the world science, and then Atkins's claim becomes safe against refutation. But that will so broaden the notion of science that we would now have to say that interacting with other people, therefore seeing friends, or looking out of your window, or going to the movies is science, since each of these three activities can teach you something about the world. Not only would this move revise the standard meaning of the word 'science', it would also render the apparently privileging nature of the claim (science the only way) much weaker than it looks, for many, many things would now become science. On the other hand, if we treat the word 'science' as applying to a relatively organized body of methods and theories for studying the world, then science isn't the only way of learning about the world. If you go about the place, even in a fairly haphazard sort of way, you may learn something about people or the neighbourhood or the climate or gravity; if you watch movies you may find out about movies (and much else); and so forth.
This may seem to some like a pedantic way of understanding Atkins's claim. Perhaps he just meant to say that science is the only way of getting to understand the world, relative to other large and organized bodies of ideas, such as religion, astrology, homeopathy, soothsaying etc. But I don't think that so interpreting the claim rescues it. Religion, for example, may well contain valid understandings of certain things even if it is false in its most central metaphysical beliefs. And if Atkins were respond to that by saying, 'Well, any true understanding provided by religion would count as science', then we're back where we started and have a proposition that's made true by definition, and useless for practical purposes.
Atkins disparages what philosophers are capable of telling scientists, so perhaps I shouldn't be engaging with what he says. Perhaps I shouldn't, but in any event I may.