Any scientist or engineer knows that our scientific theories are only models, or descriptions, of reality.  These models are well known to be approximations, but are excellent and useful descriptions in order to predict behavior and make use of resources.  Models are never rejected because they do not describe reality perfectly, only when they are supplanted by models that better match reality, and often not even then.  It is true that Newton’s laws fail at sufficiently high speeds or in sufficiently strong gravity, while General Relativity describes reality much better, such as in the bending of light in gravitational fields or in predictions of the orbit of Mercury, so close to a sufficiently large source of gravity.  But it would be ridiculous to use the equations of Relativity to calculate the force on a swing set, for example.  Sometimes we use two different models that contradict each other, when expedient.  To explain and predict the discrete way that metals emit electrons when irradiated by light requires that photons be treated as particles.  But to explain and predict the way light is scattered from a slit requires that photons be treated as waves.  A better model would predict both behaviors, but the current models are elegant and simple to calculate.

Physicists freely admit that they have absolutely no idea what gravity is, they can only describe how things act under the influence of gravity.  Our equations that describe how bodies move in a gravitational field deal with the mass of the bodies; gravity does not attract particles without mass.  Einstein shook the world of physics when his theory claimed that light was affected by gravity, and claimed this could be tested by observing that the light from a distant star was bent when the path of the light passed very close to the sun.  He shook it more when the test proved that photons were bent as he predicted.  This paradox, that a massless object, the photon, could be attracted by the gravity of the sun, can be resolved by Einstein’s thought-experiment, and is so insanely clever that this result of General Relativity can be explained to a six year old.