alt Hacker News> As a thought experiment imagine you have a biochemical PhD expert on the phone with you giving you step by step instructions on how to grow some dangerous biotoxin, giving you feedback in real time and helping you troubleshoot. Would you be more successful with this expert than you'd be on your own?
I think there is some difference of degree, but not of kind. A determined terrorist can relatively easily find many ways to kill people en masse today, no AI needed. The bottleneck is usually the actual physical execution in the real world, not theoretical knowledge.
Replies
The issue with the direction of technology is it tends to make things that were once very hard to impossible. This is why there are bumper stickers mocking Libertarians by saying things like "legalize recreational plutonium".
If we are not careful and suddenly release tools with far more capabilities than we expect you go from "smart" people being able to do it, to some angsty teenager being able to pull it off in their bedroom.
The current issue with AI is we are squirting out new models faster than we can complete long term testing on them. We'll find new model capabilities long after they've been in the field. Just assuming safety is how you catch cancer from your food dye, or how you change the atmosphere around you and start burning down your planet. Now just imagine that involving intelligence and doing with a few percent of the worlds GDP to make it happen.
>The bottleneck is usually the actual physical execution in the real world, not theoretical knowledge.
And the interest or willpower too. People fall into a kind of reductive Good vs Evil mode of thinking, with "terrorists" being of course a kind of shadowy mass of pure evil lurking in the darkness. But actual real life terrorists are people too and I'd wager few of them are actually interested in trying to end humanity.