I wonder sometimes where vacuum tube development would have gone had transistors never been invented. I suspect we'd see some forms of vacuum integrated circuits at some point, constructed similarly to VFDs, and vacuum tubes the size of modern transistors or smaller.
Well, physically speaking, semiconductors are inevitable. On the most basic level, they work by promoting charge carriers across a potential barrier, and magic happens; in vacuum, this is the work potential, and electrons are the charge carriers. In semiconductor, this is the band gap, and electrons -- or the absence thereof (holes) -- are the charge carriers. Whereas thermionic or field emission is required in vacuum (and so far, no one's yet devised a reliable field emission cathode, I believe is the problem?), thermal "emission" at room temperature suffices in semiconductors.
Indeed, there are semiconductor analogs of vacuum tubes: of the pentode, the FET (minus the screen current and secondary emission "kink"); of the triode, the SIT (static induction transistor). And the operating principle of the thyratron for example, isn't far from that of the thyristor -- in both cases a plasma forms, in the sense of abundant freed charge carriers crossing a potential barrier, the production of which is ~proportional to current flow (hence the low and ~stable voltage drop, and latching operation). (Although I'm not sure that a thyratron analog of a GTO exists; perhaps a grid electrode could block charges very close to the cathode or anode (in the "dark" barrier around either), while being insulated in some way so it doesn't draw current itself, but that seems like several big "if"s.)
So, if it's any consolation... they still glow, just at far IR?
Tim