We don't need to understand consciousness to build FIVR. We already have
the rudiments of what will be needed -- cochlear implants, artificial retinas, and prosthetic hands that can transmit the sensation of touch to the brain. None of those required unlocking the deep mysteries of consciousness.
My guess is -- and this is really just a guess -- that we will have a form of FIVR by 2039, completely different from how it has been imagined. Here's how it could work:
First, a lot of the processing to build a world could happen inside the brain itself, guided by a brain stimulator and computer. The stimulation could be through "temporally interfering electric fields" or maybe via ultrasound:https://www.ncbi.nlm...les/PMC5520675/
When people dream, their brain builds a rudimentary world, that can seem very real. With advances in brain decoding, the semantic contents of dreams can perhaps be decoded; and through what's called "brain encoding", one could figure out the right stimulation patterns to apply, in order to nudge the dream in a particular direction. 2039 is a long time away, and I could see major progress made before then.
So, perhaps you would take a drug that puts you into a lucid dream state, while wearing a brain scanner and neural stimulator -- both non-invasive. Then, a computer would analyze your brain patterns, to figure out very roughly what you are thinking about. If you start to veer off course of what's on the agenda for your trip, it could stimulate and/or inhibit parts of your brain to put it back on target. The computer might not have to know the fine details of what you should be seeing each second -- just the rough outlines -- and then you brain would generate a plausible dream, given the constraints dictated by the machine.
It's possible that we may even be able to determine fine details of what is to be experienced, not just the rough outlines -- shapes, colors, smells, tastes, etc.