See, this is what happens when you prevent/restrict the use of nuclear weapons. If we would just recognize the effectiveness of hypersonic, ballistic re-rentry, multiple warhead nuclear munitions and deploy them in conflicts instead of conventional weapons we wouldn’t need to worry about AI mis-identifying non-combatants one by one. [taps forehead]
I’m not sure how I feel about it, but I heard the argument made that AI maybe would be better for killing. People mess up all the time and misidentify threats which causes collateral damage in conflicts often. AI in theory could be much better at this identification process.
It’s kind of the same argument with self driving cars: we freak out whenever they get in an accident, but people causing thousands of accidents a day doesn’t cause an outage.
Not saying I necessarily agree with either argument, but they do make me question how we think about and evaluate technology.
See, this is what happens when you prevent/restrict the use of nuclear weapons. If we would just recognize the effectiveness of hypersonic, ballistic re-rentry, multiple warhead nuclear munitions and deploy them in conflicts instead of conventional weapons we wouldn’t need to worry about AI mis-identifying non-combatants one by one. [taps forehead]
I’m not sure how I feel about it, but I heard the argument made that AI maybe would be better for killing. People mess up all the time and misidentify threats which causes collateral damage in conflicts often. AI in theory could be much better at this identification process.
It’s kind of the same argument with self driving cars: we freak out whenever they get in an accident, but people causing thousands of accidents a day doesn’t cause an outage.
Not saying I necessarily agree with either argument, but they do make me question how we think about and evaluate technology.