
A group of researchers have just claimed that artificial intelligence could destroy humanity within 10 years. It joins a long list of other scenarios from nuclear war to designed disease. Ours has been an age facing catastrophic end-of-everything fears for some time.
I’m convinced that any of an awful lot of things could totally and civilization as we know it. Not so sure about ending humanity – people are numerous, tough, and resilient. Hard to imagine nobody making it through an apocalypse.
Unlike many of the current explanations, my take is that AI would not direct the event, but would act as an enabler for some human agency. It is easy to imagine using AI to start a war or build a plague, for example. Alternatively, systems could easily and quickly become too AI dependent, meaning any breakdown or incapacity would kill millions or more with utility, food, distribution outages.
But AI actively ending humanity? I think not. It has no hormones, after all. And this evil monster has a pretty exposed brain in its data centers and power sources. Not to mention vulnerability to a strong EMP or direct missile explosion. We tend to forget how much of our own evolution has been guided by self protection – of which AI has little.
Anyway, it really is just one other worrisome possibility, to add to those of an old guy who grew up hiding under desks for protection from World War III.









