Omicron evolution

It's too early to say for certain, and the assessment could still change as new data become available, but first indications suggest that the new omicron variant, though extremely transmissible, may cause less severe illness than earlier variants of covid-19 (more here).  That's not to say it's harmless.  Some omicron cases in South Africa have required hospitalization.  But in fully-vaccinated people the symptoms are usually mild, and I have still not seen any reports of death resulting from omicron infection, even in the unvaccinated.

If this turns out to be the case, we should not be surprised.  Higher infectiousness but less-severe disease are exactly what one would expect natural selection to produce.

In a disease microorganism, it's easy to see why natural selection would favor greater infectiousness.  A variant form which spreads more easily will, in fact, spread more, and will out-compete less-infectious variants to become predominant in the population.  But selection will also favor milder disease.

Imagine a virus producing such severe disease that it kills a person within just five minutes of being infected.  Such a virus would be an evolutionary failure.  In five minutes, it's unlikely that the infected person would get a chance to pass it on to other people, and so it would die with its first victim.

Now imagine a mutated form of this virus, a variant producing milder disease, taking five days to kill instead of five minutes.  That variant would have far more opportunities to be transmitted before the first host dies, and so would easily out-compete and replace the original form.  A further mutated variant producing even less severe disease, which doesn't kill the victim at all, would be even more successful, since it could be transmitted for however long it remained in the victim's body until the immune system cleared it out.

In fact, from the viewpoint of a virus's evolutionary success in a human population, the optimum level of disease for it to produce is zero -- that is, to not make people sick at all.  Humans understand how infectious diseases work and can take steps to minimize contact with individuals who are clearly ill.  So all other things being equal, a virus variant which produces no illness at all will spread more easily than more damaging variants, and out-compete them for survival.

So the evolution of viruses and bacteria naturally tends to mitigate the severity of the diseases they cause.  There are several diseases, such as syphilis, for which historical records show that their symptoms a few centuries ago were substantially more severe than they are today.

The delta variant partly deviated from this pattern in being more infectious than the original covid-19 but causing more severe illness.  Even so, it is so much more infectious that it out-competed and largely replaced the original version despite its greater propensity to kill its victims.  But the latter really was a liability from the viewpoint of virus survival, even if a surmountable one.

It's logical that if a new variant appears which is even more infectious and causes less-severe disease, it will out-compete and eventually replace delta, a further new stage in covid-19's evolution.  Omicron may turn out to be that variant.  If not, eventually some such variant will appear and become dominant.  Given enough time, we can expect successive mutations to steadily reduce the danger covid-19 poses.

This doesn't mean we just have to wait for omicron to displace delta and our troubles are over.  A highly-infectious disease that puts some fraction of its victims in the hospital is still a serious problem, even if it rarely kills people.  But it's worth remembering that in the long run natural selection will work in our favor.