Is gravity a force or not? - pendulum example Let's imagine a pendulum and think about Newton's law about action and reaction: If gravity is not a force, what is the stress in thread balanced with?
 A: If you model the universe as existing in curved 4-dimensional spacetime, gravity is not a force, it is a manifestation of how objects travel in straight lines in curved 4-dimensional spacetime. Hence what we call the force of gravity is in fact a pseudoforce like centrifugal (pseudo)force, the pseudoforce that throws you forward into your seatbelt when a car you're in comes to a sudden stop, or the pseudoforce acting on the thread in Gandalf61's reply. It is a manifestation of an object's inertia as measured by a non-inertial frame tracking the movement of that object.
If you model the universe as existing in flat 4-dimensional spacetime, gravity is a force with the following character:

*

*It just happens to work exactly like a pseudoforce because it has the unique character among real forces of acting equally on everything at once.

*It just happens to make the apparent flow of time measured by an observer near a large mass change exactly the way that was predicted by the curved spacetime model.

Models that rely on a lot of "just happens to..." statements and require inelegant mathematical representations are considered less true than models with elegant mathematical representations with strong explanatory character.
That being said: almost nobody, even in technical and scientific careers, has any practical use for the curved-spacetime model. For terrestrially-relevant accelerations and distances, you can almost always approximate an accelerated frame with pseudoforces as an inertial frame in which the psuedoforces are real forces that do real Work. The approximation that the surface of the Earth is an inertial reference frame and gravity is a real force that "just happens to..." is more than good enough. $\vec F_g = GmM\hat r/r^2$ and it just happens to... is true enough to build a skyscraper, a submarine, a supercomputer, or a rocket ship.
The only terrestrially relevant use for the curved-spacetime model that I can think of is coordinating the clocks of satellites with surface-level computers.
A: Put the pendulum in a spacecraft far out in space. There is no stress in the thread. Now accelerate the spacecraft to create "artificial" gravity. Now there is stress in the thread. What balances this stress ? Nothing.
The answer is that there does not need to be a force to balance the stress in the thread if the pendulum is not in equilibrium.
