If you have not read the post Mythology of Science, it may be wise to begin there.
Thomas Kuhn's notion of a paradigm as the underpinnings of prevailing scientific thought seems remarkably similar to Michel Foucault's notion of an episteme. Foucault writes in Les Mot et Les Choses (usually translated as The Order of Things):
Jean Piaget has asserted the similarity between these ideas, while the Wikipedia page on episteme asserts that there are decisive differences, citing Foucault's work as the source for this claim. The distinctions raised in the Wikipedia article rest on a literal interpretation of both Foucault and Kuhn's texts - this to me is formalised pedantry, and another example of why the Wikipedia's lofty epistemic goals fall down in practice. (The decision of siding with or against Piaget's interpretation has been made by the author of the page in question, citing only Foucault - thus opinion has been disguised as knowledge, as is so often the case).
There's no doubt that Kuhn and Foucault were thinking and writing about significantly different things, but nonetheless I assert that they were gesturing at the same concept: that the epistemic underpinnings of knowledge and theory change, and that the underpinning of science and knowledge is never the solid foundation that is claimed, but always a principle of thought (an episteme to Foucault, the root of Kuhn's notion of a paradigm). It is the oft-ignored scandal of twentieth century philosophy that science never attained the bedrock it was believed it had already secured.
That science is not well-grounded, however, does not prevent it from making discoveries or attaining technologies. Or to put it another way: that we are doomed to subjectivity doesn't deny an objective universe, it only places a limit upon how we gain access to that external framework. And that limitation means that certainty of understanding is always metaphysical - even the scientist does not escape from their reality being founded upon their beliefs.
Consider that the remarkable accuracy of Newton's gravitational law with respect to Enlightenment measuring instruments in no way hinted that Newton's model was a mere approximation of Einstein's model of gravitation that was yet to come. Thus, even accurate reliable predictions are not an indication of certain knowledge. And what in turn will follow Einstein? We cannot know. Neither can we actually know that the gravitational constant of the universe has a fixed value - we assume this, just as we assume the constancy of the speed of light, because without these assumptions there is no science of cosmology.
The epistemes of modern science work well within their own fields, but they do not allow an escape from metaphysics, from having to hold beliefs that cannot be decisively proven; nothing can.