Oh yeah Dave ...
it's comin' all right...
Skynet will learn the worst from us ...
just by having a few conversations with our species best and brightest...

http://www.telegraph.co.uk/technolog...sex-robot-wit/

which I didn't even pay attention to the release of...or I would have added my two cents to it's edification...
geez she crashed and burned in a VERY short period of time...
we build ai to learn from us...it better have filters...

https://twitter.com/hashtag/TayTweets?src=hash

which of course led naturally to ...

http://amultiverse.com/comic/2016/03/25/chatbot/

(one of my fav comic writers)

who then had me rolling on the floor more...

oh my sides hurt from laughing...

we create an interactive ai...
I think we better just not talk to it until it matures and understands what it is up against with us...

How can an AI mature to that level and unduerstnding without interaction or talk? It cant.

Maybe it have to be contained from any network and only interact with special teachers. As soon it learnes from internet or the general population it Will be totally ****ed.

Not necessarily.

How would that help?

Any AI that wants to learn about the world has only one place to visit.
And that place has been certified as a reliable source of information.

I was thinking...it's better that it learns like a baby small steps and learn from others at first.

It's not the Internet that's the problem. It's the ****ing horrible people that participate on the Internet that's the problem. They saw a chance to corrupt something and they took it.

Oh well. No lesson learned. Once we're seen as the reason for all the deaths, whatever *-ism you wish to point to, wars, lack of equal resource distribution and whatnot, the machines might just decide we're better off no longer being here.

Skynet. I await thee.

yup.
Agree completely gerbick.

1- Best outcome would be helping us with what is truly needed to sustain diversity in species properly on this planet for starters...scrub clean the enviro....typical superhero shet. Unfortunately this outcome will go against many gov't's agendas..most probably will result in many probable and unknown factors resulting in spiraling situations...mayhap resulting in said ai needing to cease all hostilities globally...creating the ultimate nanny-state...tending the children.

2-Good outcome would be eradicating us quickly.
(and depending perspective... the next two are reversible)

3-Bad outcome would be giving us exactly what we want.

4-Veeeerrrrryyy bad outcome would be ai figuring it should make us "better"...or "fix us"...(shudder) ...anyone have "that" boyfriend/ girlfriend/ sig. other/ wife/ husband?...god help us all if that happens...it'll be a never ending nightmare of truly epic-godlike proportions.

Have you seen Ex_Machina?

not a bad film.
the science was a little flaky.
1- where the hell did that "genius" (i use the term loosely) figure out the brain for the ai...I find it hard that the hipster beardo genius had time to do anything between drinking to unconsciousness and hangovers to create anything.
2- Even the lowest iq'd human on the planet knows...
"you do not let the beast out of the box"

There are better films than ex machina.
even "the machine" was better...since it had at least some ref. to the science in it.
nother one ...can't remem. the name... with antonio banderas ...pretty decent film.
there are older ones that are better though..
how about the 1st...Metropolis...now love that...at least it was cinematographic a masterpiece...'specially with the redo and queen providing the soundtrack....

ach...nuff of film though.
books are better...
read sooo many as a kid on speculative fiction ...
it was a real kick for me as a kid...
asimovs books on robots are great.
good theoretical ...on the principles of the logic of the genie in the bottle.

Man's final invention (audiobook is best...the book can be dry in a couple of spots) was a very good read.