AV1 is unlikely to ever be practical for "muggle" encode use, at least in this decade. It will only be worth committing that much compute workload to making a smaller file if the recipients will number in at least what, millions?
I'd be really curious what a hardware realtime AV1 encoder would even look like. How much silicon would that take? That kind of chip would have be be colossal even if it sacrifices huge amounts of efficiency to spit out frames at reasonable time (in the same way hardware hevc and vp9 encoders kind of suck).
You were probably thinking about hardware decoding though.
They're complementary options rather than competing. Each does something well, and it sounds like you want the thing that software encoding does.