Exactly. We're not talking about rational agents anymore, it's about lowering the barrier of entry both financially and technologically so ordinary (arguably insane) people can become Bond villains on a shoestring budget. There are a lot of possible motivations why a person might do this, for example being a religious nutjob, or because date night took a turn for the worse.
Unfortunately, in reality, such attacks are likely to cause huge amounts of suffering which muddies the water.
But if there was a magic button to press that'd instantly eliminate all life on earth, I think it'd only be fair to push it.
Or another way to put it, why is the pope-mobile bulletproof?
If you don't believe me, hear it from the horse's mouth. Here is a video from a Muslim peace conference in Norway: http://www.youtube.com/watch?v=bV710c1dgpU#t=45s It will take five minutes of your time, but I think it shows best how sincere these people are.
And the memeset I'd personally put my money on being the proximal cause is "humans are a blight upon Mother Gaia and should be removed at all costs". We don't have to imagine the existence of people who would push the Kill All Humans button if given a chance... we need merely read the right bits of the Internet. Sure, most of them would at least think twice... but there's enough that wouldn't.
I'd imagine the threat of release would make for a pretty compelling bargaining chip. Potentially, you could gain a lot of leverage without having the backing of a sophisticated military-industrial complex (unlike nukes).
It's kind of like a DIY WMD for the millennial, "maker"-terrorist generation.
[1] http://www.cnn.com/2013/12/04/world/americas/mexico-radioact...
However, I don't think that it is outside the realm of possibility that relatively small entities, on the order of a small-cap corporation or private laboratory, could eventually create a weapon that has far more global destructive power than a radioactive dirty bomb.
This would afford the entity more leverage at a global bargaining table- as opposed to a negotiation with a single government. Combined with the relative anonymity one may be able to maintain when releasing the virus, these entities might see a higher expected value in engineering a virus.
It would be "DIY" in the sense that you wouldn't need the backing of a large nation state to produce an ICBM, or opportunistically steal some radioactive material to create a dirty bomb.
Also, a virus is implicitly cheaper to distribute: with a longer incubation period and an exponentially increasing rate of propagation, a virus could reach critical mass before there's time to react.
In a sense, this makes them more effective than other types of WMD- which may have lower singular chances of successful detonation, have high per unit capital costs, tend to have local area effects only, and are generally difficult to produce far removed from military industry at scale.
Personally, I think the world is more likely to end with a bug than a bang. I honestly don't know which is more terrifying to me.
The risk comes from widening access to a given technology, and the amount of harm it could do. It's hard to calculate the risks, but bio weapons potentially have wide access and great harm.
And "bad guys" are not necessarily required for a global catastrophe. Human error can do the job just as well.
That said, I think biotech is an amazing force for good in the world, we just need to be responsible with how we deal with it. Which means defining and justly implementing 'responsible' is a core tech challenge.
With bioweapons, the perpetrator is much harder to track down. And if you don't know who your attacker is, you can't retaliate.
Take this researcher, Bruce Ivins: