Israel aims to be 'AI superpower', advance autonomous warfare
Source: Reuters
JERUSALEM, May 22 (Reuters) - Israel aims to parlay its technological prowess to become an artificial intelligence "superpower", the Defence Ministry director-general said on Monday, predicting advances in autonomous warfare and streamlined combat decision-making.
Steps to harness rapid AI evolutions include the formation of a dedicated organisation for military robotics in the ministry, and a record-high budget for related research and development this year, retired army general Eyal Zamir said.
-snip-
He named GPT (Generative Pre-trained Transformer) and AGI (Artificial General Intelligence) as deep-learning realms being addressed by civilian AI industries which could eventually have military applications.
These, Zamir said, potentially include "the ability of platforms to strike in swarms, or of combat systems to operate independently, of data fusion and of assistance in fast decision-making, on a scale greater than we have ever seen".
-snip-
Read more: https://www.reuters.com/world/middle-east/israel-aims-be-ai-superpower-advance-autonomous-warfare-2023-05-22/
GPT-type AI are LLM (large language model) AI and hallucinate, or make things up, often so convincingly their mistakes aren't caught quickly. Which is why GPT AI like ChatGPT come with lots of warnings not to trust their results.
AGI is the level at which a lot of experts warn AI can become uncontrollable and extremely dangerous.
Prairie_Seagull
(3,329 posts)'Wait but that's going to be way in the future'. I am thinking maybe not so far.
Ray Bruns
(4,098 posts)marybourg
(12,633 posts)an show up at your door? 😂
Sin
(472 posts)Orrex
(63,215 posts)XorXor
(621 posts)Then things are going to get pretty scary. Someone is going to eventually do it. Whether it is Israel, The US, China, or Russia (or all). Good luck trying to convince any country to unilaterally stop all defense development like this, as they know their adversaries won't be ceasing it.
LudwigPastorius
(9,155 posts)...or, have they?
The manufacturer says "no", but the South Korean government isn't saying one way or the other.
XorXor
(621 posts)We also see missiles like LRASM that is touted as being able to use AI to determine which ships to target once they get close. Although, that's more of an incremental increase since it's my understanding that even legacy missiles and torpedos do the same in less sophisticated manners.
I'm not edumacated enough to make any suggestions on where the line that open Pandora's box is at. I'd be curious to hear some thoughts of any experts who may happen across this.
sakabatou
(42,155 posts)BrinksHomeSexcurity
(14 posts)Are more like 'The Culture' than Skynet or the Matrix.