Yes, photons are bosons / force carriers, but they interact with charged particles, and in this case produce e+ e- pairs via this Feynman diagram [1]. By rotating the diagram in spacetime, you get different known interactions: pair production (this topic), pair annihilation (same diagram running the other way in time), and, if memory serves, Compton scattering.
"Rotate the diagram in spacetime" would make for some top-tier technobabble (not expressing doubt that it is a real thing, just wow, it sounds so cool).
Ok, so high enough energy photons fluctuating to other particle/antiparticles and then being able to interact with other high enough photons is an explanation I'm willing to settle for, because I thought that multiple bosons can occupy the same space, thus they were not able to interact with each other, by the "boson" definition.
It's still blows my mind that a photon could do that -- turn into a particle-antiparticle pair, the pair then quickly gets annihilated and turns back into the same photon and continues in the exact same direction and form the initial photon was travelling.
But if this process is real, does it mean that high energy photons travel slower than low energy photons?
Because low energy photons could not transform into particles that have mass, while higher energy photons could and thus spend just a liiiiiiitle bit more time as massive particles that can't travel at the speed of light.
[1] https://en.wikipedia.org/wiki/Two-photon_physics#/media/File...