So, in destroying human viruses, you’re speculating that we might also destroy transposons? Would this really wipe out the human species, or would it just make us less diverse? (And thus more prone to being wiped out by something else.)

I don’t think your scheme would be possible. The RNA tag needs to be transcribed, not added later (since then it would be added to viral RNAs transcribed in the cell), so it actually needs to be inserted into the DNA. That means modifying every single transcribed open reading frame in the entire genome. You additionally need to add a means of detecting this tag, so that means perhaps adding a novel protein or even whole set of proteins to the translation machinery–good luck with integrating that. This massive genomic overhaul would have to be done on every human on the planet in one fell swoop, because otherwise if a person without the overhaul bred with a person with the overhaul, their offspring would either die or have major systemic problems. Then there’s the addition of many opportunities for new disease if a mutation drops the tag on a gene. Add in the additional problem of evolution being more clever than you (Orgel’s second rule) and the eventual acquisition of this tag by a virus is inevitable. If this were remotely possible, evolution would have invented it already!

Lateral gene transfer is not terribly important for multicellular organisms. The only cases of viral transfers of genes that we know about are oncogenic (causing cancer). Viruses have been instrumental in causing mutations, which can be fuel for evolution, and we actually domesticated one viral gene in our evolutionary history that is active in placental development.

Transposons also have been useful in history, but for their role in adding DNA (fuel for evolution) and screwing up genes, sometimes making new ones. Transposons cannot transmit DNA from individual to individual because they do not move outside the cell.

siRNA is just small interfering RNA. In invertebrates it is generated from double-stranded viral RNA (normal RNA is single-stranded) by the action of a RNA-cutting molecule called Dicer, and the short RNAs are then used by the RISC complex to degrade complementary viral RNA. However in mammals the response to double-stranded RNA goes through Toll-like receptor 3 and the RISC complex is actually used for making micro RNAs that help control genomic gene expression.

There are several major classes of viruses that have very different lifecycles, and there is no magic bullet that can destroy every single one.

“With an altered DNA transcriptase, a leading key sequence could be added for every transcribed mRNA. The gene that codes for ribosomes would also need to be altered to accept only mRNA with the leading key sequence.”

The problem is RNA viruses transcribe their mRNA in the cell, and this would add on the tag. You might mildly inconvenience positive-stranded RNA viruses until they could generate more transcripts from their genome, but negative-stranded would proceed about their business as usual. Additionally you don’t touch retroviruses and DNA viruses.

“I’m not talking about evolution here, I’m talking about engineered genomes.”

Again, Orgel’s second rule says evolution is more clever than you. Genetic engineering is a way to jump ahead of evolution, but evolution always catches up.

“No, evolution produces good enough solutions, not the best solutions.”

You’re proposing a solution that would use two simple (according to you, according to me way too complicated to ever work) modifications and you think evolution would not have hit upon this already if it were possible? The types of tags you talk about are in use for many purposes in cells, and viruses have co-opted these.

“Correct. I was suggesting that viruses, human viruses, or multicellular organism viruses were the biproduct of viruses that were useful to our single celled predecessors.”

I think viruses were primarily parasitic from the beginning, although sometimes being subverted.