Skip to main content

AI might kill us, but not the way you think

One of the big hypothetical doomsday scenario is rogue AI that will end up killing the whole human kind for a reason or another. But we don't need a malicious machine to finish us all, we can just use it to our own demise.

AI can be really good at coming up new thinks to kill people. Sure, the public ones have all kind of safeguards to prevent them to be used for such things. But those protections aren't perfect, especially as even the creators aren't quite sure how their creations works. There was just an article in the news how somebody had used one of the latest LLMs to create 40 000 new neurotoxins!

But wiping out humanity doesn't even need to a conscious effort. We have already pushed the climate to the brink of making the Earth unlivable by humans. AI is just another power hungry tool for us to spew us new wonders while we feed it our future.

Of course there could be the optimistic side to all this, either the machine help us kill ourselves without harming the environment and leave it livable for other species or even better - it could help us find a way to save us.