SG Talk

My idol Elon Musk claim AI is " Dangerous " in fact is true

(21-11-2022, 05:59 PM)[[ForeverAlone]] Wrote: [ -> ]My idol Elon Musk claim AI is " Dangerous " in fact is true

U SHOULD WATCH THE MOVIE BY SPIELBERG
https://www.youtube.com/watch?v=_19pRsZRiz4

(21-11-2022, 07:04 PM)singaporean1964 Wrote: [ -> ]U SHOULD WATCH THE MOVIE BY SPIELBERG
https://www.youtube.com/watch?v=_19pRsZRiz4

I forget that scientist name , he already RIP few year back if I remember well.

He also mention " AI is dangerous "

(21-11-2022, 07:06 PM)[[ForeverAlone]] Wrote: [ -> ]I forget that scientist name , he already RIP few year back if I remember well.

He also mention " AI is dangerous "

NOT TRUE
ONCE THE ai ALgo-rythm/SUB ROTINEs BECOME AS COMPLEX AS THE HUMAN MIND
then it has a choice to chose good or bad
some humans chose good
some choose bad
AI will choose a certain thing based on external input/experience..same as human
eg..i dislike indians generally becos the have wrong me in the past
if there are too many evil humans around
the AI may choose to protect itself with violence

(21-11-2022, 08:34 PM)singaporean1964 Wrote: [ -> ]NOT TRUE
ONCE THE ai ALgo-rythm/SUB ROTINEs BECOME AS COMPLEX AS THE HUMAN MIND
then it has a choice to chose good or bad
some humans chose good
some choose bad
AI will choose a certain thing based on external input/experience..same as human
eg..i dislike indians generally becos the have wrong me in the past
if there are too many evil humans around
the AI may choose to protect itself with violence

I for see some already trying to integrated AI with REAL Human Brain.

I happen to see one medical school they got keep lots of Human brain from those RIP donated their organ for research.
Some Scientist already have the idea etc to do this. AI with REAL Human Brain.

Do you know some scientists already using technology to make a baby out ? even A lamb.

aiya, ai is just computer programs. if the programmer is bad, the ai is bad. if the programmer is dangerous, the ai he created will ofc be dangerous

(21-11-2022, 10:04 PM)WhatDoYouThink! Wrote: [ -> ]aiya, ai is just computer programs. if the programmer is bad, the ai is bad. if the programmer is dangerous, the ai he created will ofc be dangerous

actually I know Microsoft already invented a " self programmer AI " it will ownself detect codes not working etc even can do programming.

there are narrow, general and superior AI.

the prevalent ones are narrow AI, which is to do specific tasks.
They are pursuing general AI at the moment which can mimic multi-modal inputs and outputs with intelligence. similar to humans but better.
Superior AI is something that will far exceed all of what we known today.

if you work with autonomous self-learning algorithms such as reinforcement learning, it is rather daunting to know that something will outperfrom a human's capability many times. However, the learning process for RL is still a brute-force approach which require numerous iterations in the simulation.

miscrosoft's AI cortana chose the 'good' path
https://www.youtube.com/watch?v=WKryvDPoJf8

(21-11-2022, 05:59 PM)[[ForeverAlone]] Wrote: [ -> ]My idol Elon Musk claim AI is " Dangerous " in fact is true

Probably they never heard of google AI trial.