A common misconception about the evolution of viruses

I'm not sure deadliness is orthogonal to reproducibility. You're correct that the statement you provide is false, but I think I would defend a similar statement as follows:

1. Causing humans to get sick is very likely to make a virus less transmissible, as the host stops moving around as much, or dies. This generally happens in the short term, but if not then in the long term - for example, if a virus transmits solely through relatives touching the corpses of the dead, it may initially be more transmissible the more lethal it is, but once the human populations that maintain this custom have been replaced by those that don't, then killing its victims rapidly will become a disadvantage. 

2. The disadvantage for a virus in causing humans to have symptomatic illnesses in is in tension with the fact that to succeed, viruses need to make human cells stop doing what they're supposed to do, and start reproducing the virus, which is by definition going to mean our bodies working less well.

3. All viruses face both of these evolutionary pressures. Together they mean that the deadliness of a viral disease in a human population isn't random but, for a particular virus in a particular population, has an optimal level.

4. When we notice a virus starting to be transmitted between humans, and becoming endemic in the human population, we do so because that virus is more dangerous than all or nearly all other viruses currently in circulation. By virtue of the fact that we have noticed a virus, it is likely that on the "cause less/more sickness" axis it is further towards the "more sickness" end than is optimal.

5. So the new viruses that we are aware of tend to evolve to become less dangerous.

The most obvious weasel words in the above are "for a particular virus in a particular population". Given that humans evolve, and human customs and immune systems change, in response to viruses, then it could well be the case that in general the effect of viral evolution is dwarfed by the effect of humans evolving, and human societies and immune systems changing, in reaction to the presence of the virus in humans. So viral evolution might not matter much. Even if it does, other evolutionary pressures on the virus, such as avoiding the human immune response, might be far greater than the pressure to become less dangerous to humans. But I would still expect to see new viruses that we are aware of tending to evolve to become less dangerous.

I wrote about the literature around the evolution of virulence [LW · GW] previously which people might find useful for finding references on the topic.

In short, it is complicated:

1. zoonotic diseases (like COVID) don't follow a consistent pattern early in their evolution because they are so far removed from being adapted to humans as a niche.

2. As long as there is a correlation between deadly-ness and ability to reproduce, diseases will face evolutionary pressure to become more/less deadly depending on the circumstances. This is strongly influenced by transmission mode (but isn't the only factor):

   a. Disease transmitted by close person-to-person contact -> disease can't make host too sick, or host won't come into contact with others -> disease evolves to cause milder infection.

   Example: HIV has relatively few effects on hosts for years before culminating in AIDS.

   b. Disease transmission doesn't require person-to-person contact -> disease can optimize other features of transmission such as producing many copies and using host resources -> disease evolves to cause severe infection.

   Example: malaria is spread via mosquitos and doesn't require person-to-person contact in order to spread, which might explain why it is such a debilitating disease.

One thing that has occurred to me is that a less dangerous variant may end up being more infectious in practice, as more infected people are either unaware they have it or are less incapacitated, and therefore are free to expose themselves to others more.  So this is one way that a less dangerous variant may overtake a more dangerous one, even if their inherent infectiousness is similar.

(has HIV evolved towards a less severe disease?). 

How would we know? HIV is defacto less deadly nowadays than it was in the beginning. A large part of that is about having better treatments but I don't see how the world would be different if HIV in fact became on average less deadly.

Overall this post reads to me like someone talking about a subject outside of their domain experience without any citations and without being aware of the mechanism the field in question sees.