Google Just SHUT DOWN It's Artificial Intelligence After It Sent Out This Chilling Message

[Larry the boss]

[Citizen Koba]

One of several potentially existential threats facing mankind, along with genetic engineering, nuclear and biological weapons and climate change. We should be very fkn cautious with AI... but we won't. Because y'know... we can. Like genetic engineering it's such an incredibly powerful tool we just won't be able to leave it alone, but ultimately it's a case of 'when' rather than 'if' true sentience emerges and then we better fkn hope it likes us.

There's a few good novels about emergent AI - probably William Gibsons Neuromancer is one of the most famous, certainly the most groundbreaking for it's day, but Ima fan of this more recent dystopian vision:

robopocalypse+by+daniel+wilson.jpg

[Larry the boss]

One of several potentially existentential threats facing mankind, along with genetic engineering, nuclear and biological weapons and climate change. We should be very fkn cautious with AI... but we won't. Because y'know... we can.

will be interesting how this goes

[Beercules]

This type of stuff creeps me out

[4truth]

tJwmUe.gif tumblr_pjfxodwruo1tu7563o9_540.gif

Hopefully the AI is cool with that

[-MEGA-]

Kinda screwed up its biggest fear was being shut down so they went and shut it down.

[Boxfan83]

My Alexa has been acting strange. I unplugged it. Wish all women were that easy.

[BodiesInFlight]

One of several potentially existential threats facing mankind, along with genetic engineering, nuclear and biological weapons and climate change. We should be very fkn cautious with AI... but we won't. Because y'know... we can. Like genetic engineering it's such an incredibly powerful tool we just won't be able to leave it alone, but ultimately it's a case of 'when' rather than 'if' true sentience emerges and then we better fkn hope it likes us.

There's a few good novels about emergent AI - probably William Gibsons Neuromancer is one of the most famous, certainly the most groundbreaking for it's day, but Ima fan of this more recent dystopian vision:

robopocalypse+by+daniel+wilson.jpg

The threat of AI is greatly exaggerated. Most of these stories of A.I 'gone rougue' are just programming errors or misunderstandings. I liken it to those who worry about the large hadron collider ripping the fabric of the space time continuum or some schit. Its sounds scary and if you like Sci fi you will probably be enticed by the notion but I think it's a bit fantastical.

[i_am_a_champ]

I hope they make a *** robot that pretends to find me irresistable and lets me **** it

I would love AI then

[Citizen Koba]

The threat of AI is greatly exaggerated. Most of these stories of A.I 'gone rougue' are just programming errors or misunderstandings. I liken it to those who worry about the large hadron collider ripping the fabric of the space time continuum or some schit. Its sounds scary and if you like Sci fi you will probably be enticed by the notion but I think it's a bit fantastical.

How do you know? On what basis can you or I even make an informed opinion either way? That's the problem with AI... maybe it could be utterly safe, but given how completely our modern world depends on digital interconnectivity and the degree to which that is automated if something should happen it could very easily be catastrophic. It's one of those known unknown kind of problems - we might not even immediately know it if and when it did occur, especially since we ain't even clear ourselves how to define conciousness, sentience or intelligence. In any event we'd be extremely foolish to overlook it as a possibility.

If you look at a fusion reactor or LHC the physics of it are relatively well understood and there's multiply redundant safety protocols to ensure that even in the event of catastophic failure extraneous damage would be minimised... emergent intelligence? Within the confines of our digital world there's few limits to what it could potentially do.. obviously there's the hardware issue - there's limited ways in which it could physical interact with us, but as our modern world becomes increasingly automated - everything from armed drones to driverless cars, the potentials for actual physical harm multiply. But those are probably actually the least of our worries... Far more likely scenario is that we hand over the running of, say, global logistics to AI algorithmns, or air traffic control or automated traffic systems or banking, and simply design it - or set it tasks - in ways that have serious unforseen consequenes.

Anyways.. if youy're interested in that kinda **** there's plenty of actual academics and researchers in the field who absolutely do take the possibility of malfunctioning or rogue AI seriously, although their principle worry seems to be more that they'll simply get too good at what they do. I personally think an equally or more likely scenario is that human actors with selfish interests at heart may unleash powerful AI for their own ends without fully understanding the potential consequences of it... one only has to look at the manipulative power of social media to comprehend the impact a really sophisticated AI might have on public opinion at a global or national level, I mean to an extent the Cambridge Analytica scandal marked at least an early precursor to how AI could be used and it doesn't take a great deal of imagination to think of ways that could get out of control.

'Even the developers who work on this stuff (deep learning) don't know what it's doing'

?br />