Michael Anissimov gives an update on the current state of several existential risks -potential global threats to humanity. As I've discussed in my podcast and elsewhere in this blog, the nature of existential risks prevents our usual "trial and error" approach to things. There is no opportunity to learn from your mistakes with an existential risk. There is no second chance.
It's also the nature of existential risks that there is very little incentive to mitigate such risks. It's usually expensive and, after all, we've been OK so far. This is the same attitude displayed by NASA concerning the Shuttle foam issue, but on a global scale. Since we haven't died so far, we must be OK. Even risks which we know have happened many times in the past and are guarenteed to happen again, such as an asteroid collision, are mainly ignored.
Check out Nick Bostrom's excellent paper on the subject for an extensive list of possible extinction scenarios. Some of them are rather entertaining!