When the date rolled over, very little went wrong.

Now to be fair, the high risk scenarios, systems that do navigation, financial systems, emergency services and so on were all migrated. Operating systems upgraded which covered a large swathe of problems.

So the solution wasn't just nothing happened, it was that we prepped for the worst cases and prevented them. This is exactly what we're doing now, security in harnesses, not giving AI access to everything, hardening all the software vulnerabilities that are found before they are exploited.

Biological weapons are always a risk, but this has always been the case with just humans too. The ease at which we can kill has rapidly improved, but this hasn't shown propensity for us to kill more.

AI in this respect is the ultimate enabler, I would expect a period of small events happening, but widespread end the world, much less likely (we'd more likely be wiped out by a human started WW3).

The problem in the latter case will be “who has the money to consume” as we’re living in a consumer economy. I don’t know what the answer is, but I think it best that one doesn’t accelerate things to a resolution. Society and governmental systems aren’t ready to handle tremendous disruption. It ain’t Star Trek.