Re: security concerns
128 posts • joined 5 Jan 2015
AI is a mind-lever, as opposed to an independent consciousness. Beware not the pawnish Terminators but (the mindset of) the chiefs of Skynet.
Push-AI gets programmed (eg by rules or training data/regimes). But pull-AI (adapt-to-thrive bots) in the political world becomes whatever thrives best in that world. Like scaremongering and threat-creation. Like human politicians, just amplified. Oh, we already have one.
I discussed this kind of application (at a space conference) with one of the major ISS designers (prior to ISS happening). He agreed it was a good application in principle, e.g. avoiding the “structural rattle” associated with (more impulsive) traditional gas-expansion attitude-control options.
But the aerospace industry is renowned for its conservatism. Who needs wheels when there is so much accumulated experience with skids? And risk assessors love accumulated experience.
One reasonable question (addressable) regards possibility of ionisation and hence structural charge.
Biting the hand that feeds IT © 1998–2021