Seems to me that the default is "I don't know what's going to happen" and if you're making a confident prediction, bring evidence.

Scott makes a Lindy effect argument which is plausible, but don't let that fool you, we still don't know what's going to happen.

I'm pretty sure that gaming benchmarks can continue indefinitely.