3 ms·
Elon Must Worried That Google Might Accidentally Build SkyNet
- orionblastar 11y agoAI is not that smart yet to reach self awareness and know the difference between good and evil. Terminators are still a work of fiction. AI is still primitive and we don't have genuine intelligence or anything at a human level of intelligence. The real threat from AI is taking away jobs from human beings. To counter that there would have to be some sort of universal basic income and low cost housing and low cost college education to get back into better paying jobs that require more knowledge to work for those who got replaced with an AI program or robot.
- karmakaze 11y agoHowever it's not required for AI to distinguish good and evil to be harmful. It could be the very lack of this distinction that enables it to interpret a situation where the selected course of action is destructive to humanity.
- JoeAltmaier 11y agoExactly! A traffic-control AI could change lights more quickly and efficiently to encourage streams of cars to go faster, and optimize for total volume. Never mind a few more people die each day.
- orionblastar 11y agoNot to mention that the AI lacks common sense to notice that people are dying due to its decisions and make corrections so that they don't die. We are so reliant on technology and software to make decisions for us, and none of it is based on common sense.
- krapp 11y agoNot the first time Elon Musk has underestimated how complicated and hard AI actually is: https://news.ycombinator.com/item?id=9235002 https://news.ycombinator.com/item?id=9235002