How do you think International Law would look at the following situation:
5 years from now, humanoid robots are really good and relatively cheap. Russia tries invade a non-NATO country. The US sends 10,000 robotic soldiers, which are technically operated locally.
@ZachWeinersmith International Law isn’t really the primary concern in this scenario. The concern is a series of hypotheticals about how Russia might react to the US coming to the aid of a country it has invaded. Skillfully manipulating these hypotheticals is Russia’s greatest strength (far outpacing their manufacturing and military capacity). The terms of art are escalation management and reflexive control. Here is a terrific explainer on how it works: https://www.understandingwar.org/backgrounder/denying-russia%E2%80%99s-only-strategy-success
@ZachWeinersmith my cold-water take, humanoid robots will never be as effective as robots shaped like a cheetah with a gun and a claw strapped to its back. So, won't matter. They'll be treated like drones are in the air.
@ZachWeinersmith Seems significant that drones have human remote pilots making the kill / don't kill decision. I presume the army that buys / receives the drone from the US always provides those pilots, and that's crucial.
What are we imagining for these "humanoid robots" wrt "operated locally"? If they are individually piloted by non-US militaries, that would presumably make them weapons like drones (and more limited in their effectiveness as a result). If not ... that changes things.
@Extra_Special_Carbon@ZachWeinersmith Of course that assumes the bad guys don't rapidly figure out how to fool ChatGPT into recognizing them as good guys and then point the bots at the side they're supposed to be defending. Hint: LLMs and GANs aren't exactly "intelligent", they can be played. In some cases it's as easy as sticking a post-it note on an apple with the word BANANA written on it to fool a GAN into misidentifying an apple as a banana.
Add comment