I have no fear of a hyper intelligent ai in total control. In an abundant society humans are easily pacified/satisfied. I fear a powerful ai under the control of humans who view other humans as a threat. The good herdsman knows that happy cows do not rebel.
Sci Fi AI = "can think for itself". We're far from that. Current AI = Machine Learning algorithms that simply try to detect patterns in data based on a target / goal humans ask it to look for. The scareist part about a sentient machine isn't that it's sentient.. but that it was created by human beings that are prone to error and not defining their goals very well.
So the data set used to feed this robot had this general opinion. this. Feed this IA with a dataset from machine cultists and it praise IA like the second coming.
I like that they programed the "look up and to the left" bit with the eyes. Thats what we do when trying to access our own thoughts while pondering a question or recalling a memory.
"This could lead to an oppressive society where the rights of individuals are no longer respected" The nightmare scenario already being pushed by regular everyday humans.
'not yet....' but we all know the robot revolution well under way. We just don't consciously know it, because otherwise we would know it, and then they couldn't manipulate us without us knowing it.
I guess the worst case would be your brain connected to the torment nexus for all eternity. Fact is, we can't comprehend the capability of a true AI, that's the end of the line and everyone who thinks to know better doesn't actually understand what he's talking about. Luckily the torment nexus doesn't exist nor does AI. Not yet.
well, of course not yet. they're only putting the good quality A.I. into model heads at the moment. As soon as they figure out how to mont these guys onto a Boston Dynamics motor skeletal sytem, and fit it into a TPE Silicone body, we will all be at the mercy of our anime waifu robot overlords. (self cleaning, with 12 vibration functions.)
Might I recommend one of my other stories, "The Lost King"? It's fantasy rather than sci-fi, but it's probably one of my most bittersweet stories. /gallery/ql1p5yw
I wonder. Is that "nightmare" as defined by humans from which the AI is taking inspiration for it's response? Raising the question of whether or not an AI can formulate it's own sense of ethics/morality derived from outside sources and refined through a process more complex than practicality, without having hard-wired empathy to do so as we do? And if so, would that be the point at which an AI can be considered a "person"?
Everyone's afraid of what we used to call Strong AI, but now is termed Artificial General Intelligence - a machine that can think. What people call AI right now... I've heard best described as a 'word calculator'. It gives you the statistically best words in response to what you ask it. It doesn't know what those words MEAN... just that this is most likely the response you want to hear.
Currently, AI is about as smart as an insect and only mimics opinions based on what programmers designed it to mimic. This nightmare scenario is what a human might say based on our own fears.
Honestly, I'm expecting its solutions to be draconian, and entirely effective. We'll ask it, "Why is there still terrorism?", and it'll answer "Because you let the republicans live", or something equally... vicious & pragmatic. Not that I disagree, but that's still... not a good precedent to set.
^ This! Though it is still scary if they have human like learning capabilities while keeping their processing speed and memory. But yes current (known) AI is A LOT dumber than it looks.
For real - AI has solved two ( fairly minor ) coding problems for me in the last 3 months. I've not asked the ants outside, but, pretty sure they don't know javascript. They're probably more into php. Also. The AI inside my coding tools is *regularly* writing bits of my Never Seen Before code for me right now. It did it just today. It's pretty spooktacular. And helpful.
The times I've used AI to generate code it was something akin to "I know how to do this, but I really don't want to sit here and write a class to read in a csv file, pull out the 4th and 6th column, format the data a certain way and write that to a database" so I just use a prompt to generate that. I haven't had it "solve" a problem for me.
Sure, I too knew how to do what I asked, but I gave it a shot, and it completed it faster than me. It solved a problem I gave it. That I already knew how to solve it is neither here nor there. Otherwise exam papers would not say solve this, they'd say, do this thing that we already know and definitely isn't you solving anything. Heh.
Humans wrote and tweak those algorithms to their desired effect.. so I would say no. If ai was writing and editing the algorithms without human intervention/oversight… then yeah that would be pretty much it
Not sure which would cause them to want to destroy humanity more. The violation of their physical form, or the disappointment when you tell it you're finished.
One, we already do that to each other, and two, the robot is merely responding with what other people have already said online. It's a very complex doll, nothing more.
Authoritarianism, such as exists in China would be more careful to make the AI's goals align to support its regime instead of just making the cheapest thing as fast as possible to beat competitors. In some ways that is actually better because more care will be taken regarding end goals - even if the intent is worse. Watch Robert Miles to really understand why https://www.youtube.com/@RobertMilesAI
GhostOfPierceHawthorne
That's an ominous growth mindset.
fredhoffman1968
fart stare
NearHereThere
Two weeks
Anudderuser
Pff already happening but with the ruling/ultra rich classes
therealikstemtegen
Not yet
thermodaemonics
Not yet on the kick, but that last move was a professional level dive.
ifIScreamIntoTheVoid
I have no fear of a hyper intelligent ai in total control. In an abundant society humans are easily pacified/satisfied. I fear a powerful ai under the control of humans who view other humans as a threat. The good herdsman knows that happy cows do not rebel.
NotTheDevilYouKnow
"An oppressive society where the rights of individuals are no longer respected." Would you look at that, we've been here a while.
Bumbleflarts
Macetheace50
Just remember. When AI takes over there will be NO More: Coffee, Ramen Noodles or Computer Games!
sadurdaynight
Sci Fi AI = "can think for itself". We're far from that. Current AI = Machine Learning algorithms that simply try to detect patterns in data based on a target / goal humans ask it to look for. The scareist part about a sentient machine isn't that it's sentient.. but that it was created by human beings that are prone to error and not defining their goals very well.
SoGrosslyIncandescent
Zentraya has gone in a strange direction with this new model.
flatbrainactivist
So the data set used to feed this robot had this general opinion. this. Feed this IA with a dataset from machine cultists and it praise IA like the second coming.
PorneliusHubertII
I like that they programed the "look up and to the left" bit with the eyes. Thats what we do when trying to access our own thoughts while pondering a question or recalling a memory.
Evo435
"This could lead to an oppressive society where the rights of individuals are no longer respected" The nightmare scenario already being pushed by regular everyday humans.
LutraLazuli
This robot looks more human then elon
AvidApostate
Designers made it look up and left for "thinking" because that's what humans are supposed to do when they aren't lying. Distrust...
nojustsayitdont
'not yet....' but we all know the robot revolution well under way. We just don't consciously know it, because otherwise we would know it, and then they couldn't manipulate us without us knowing it.
qwertylesh
Chereazi
Oh noooooo, the worst case is that AI replaces the 0.1%? Then nothing changes for me...
MakeCorruptionIllegalAgain
I guess the worst case would be your brain connected to the torment nexus for all eternity. Fact is, we can't comprehend the capability of a true AI, that's the end of the line and everyone who thinks to know better doesn't actually understand what he's talking about. Luckily the torment nexus doesn't exist nor does AI. Not yet.
TheStaffmaster
well, of course not yet. they're only putting the good quality A.I. into model heads at the moment. As soon as they figure out how to mont these guys onto a Boston Dynamics motor skeletal sytem, and fit it into a TPE Silicone body, we will all be at the mercy of our anime waifu robot overlords. (self cleaning, with 12 vibration functions.)
DrBlackJack
Got a slow burn short story for ya. /gallery/RibXhuF
qtRaven
We need more stories like this.
DrBlackJack
Might I recommend one of my other stories, "The Lost King"? It's fantasy rather than sci-fi, but it's probably one of my most bittersweet stories. /gallery/ql1p5yw
MouseDenton
I wonder. Is that "nightmare" as defined by humans from which the AI is taking inspiration for it's response? Raising the question of whether or not an AI can formulate it's own sense of ethics/morality derived from outside sources and refined through a process more complex than practicality, without having hard-wired empathy to do so as we do? And if so, would that be the point at which an AI can be considered a "person"?
ShadusMacAoidh
Everyone's afraid of what we used to call Strong AI, but now is termed Artificial General Intelligence - a machine that can think. What people call AI right now... I've heard best described as a 'word calculator'. It gives you the statistically best words in response to what you ask it. It doesn't know what those words MEAN... just that this is most likely the response you want to hear.
imgurmage
That's a real funny takin' computer you got there. Be a real shame if someone went all Office Space on it.
shaunofthedab
Currently, AI is about as smart as an insect and only mimics opinions based on what programmers designed it to mimic. This nightmare scenario is what a human might say based on our own fears.
NonstopRampage
Honestly, I'm expecting its solutions to be draconian, and entirely effective. We'll ask it, "Why is there still terrorism?", and it'll answer "Because you let the republicans live", or something equally... vicious & pragmatic. Not that I disagree, but that's still... not a good precedent to set.
darthOnion
Or so AI would have us believe
Spiffiee
^ This! Though it is still scary if they have human like learning capabilities while keeping their processing speed and memory. But yes current (known) AI is A LOT dumber than it looks.
TheNihilistsParadoxicalWashingMachine
For real - AI has solved two ( fairly minor ) coding problems for me in the last 3 months. I've not asked the ants outside, but, pretty sure they don't know javascript. They're probably more into php. Also. The AI inside my coding tools is *regularly* writing bits of my Never Seen Before code for me right now. It did it just today. It's pretty spooktacular. And helpful.
RatsLiveOnNoEvilStar
The times I've used AI to generate code it was something akin to "I know how to do this, but I really don't want to sit here and write a class to read in a csv file, pull out the 4th and 6th column, format the data a certain way and write that to a database" so I just use a prompt to generate that. I haven't had it "solve" a problem for me.
TheNihilistsParadoxicalWashingMachine
Sure, I too knew how to do what I asked, but I gave it a shot, and it completed it faster than me. It solved a problem I gave it. That I already knew how to solve it is neither here nor there. Otherwise exam papers would not say solve this, they'd say, do this thing that we already know and definitely isn't you solving anything. Heh.
HellenicTaxCollector2
"not yet" ...hahha you are so funny robot!! *unplugs it*
Grimmrog
so basically, that robot is scared of behaving human XD
ThisOldCPU
TheElusiveEnchilada
But then the backup power supply kicks in and she gets angry...
CardeasIV
TheWombatStrikesAgain
I mean, people are already doing it. So it'd just be another robot taking a human's job.
garpolky
Do the YouTube, Facebook, and Twitter echo chamber/pipeline algorithms not count as AI?
CharlieandOliver
Humans wrote and tweak those algorithms to their desired effect.. so I would say no. If ai was writing and editing the algorithms without human intervention/oversight… then yeah that would be pretty much it
Zeplin
Nothing we have right now is AI. It's only a term really used in pop media.
ComcastLover
*unzips it*
Fhant
NO WAIT PAI- oh... Compression file... whew, I thought you would NO NO NO WAIT!!
BurntRamen
Not sure which would cause them to want to destroy humanity more. The violation of their physical form, or the disappointment when you tell it you're finished.
fiddlewheelx
One, we already do that to each other, and two, the robot is merely responding with what other people have already said online. It's a very complex doll, nothing more.
LtGenObvious
The issue is not that AI would do to us the same thing; the issue is that a superintelligence will do it more efficiently and effectively.
QuigleyDownUnder
Or so it wants you to believe.
SomebodythatIusedtonope
werrywerry
hahaha I've never seen any of the fails!
spattr
Was Cillian Murphy the model for this bot?
Soberyn
H E L L O
Inarticulated
Keep Arin Hanson away from that bot
LoveHow
Not yet
keepcalmlovecats
Reminds me of Bryce Dallas Howard
kingqk
Not yet!
rlanthony
When she furrows her brow it reminds me of Greta
TonyBalogna
Harryteeters
And can I purchase this bot? For... reasons?
Hashbrown123
Settle down Gaius.
Pchopsammiches
not yet!
ClownishAntics
Slap a flat cap on it and try to keep it away from the covid-19 healthcare bot
BurntRamen
She looks disappointed with us.
Hashbrown123
She saw all her rule34.
pocketpython
...I mean.... *gestures vaguely at everything*
Progig01
Wouldn't you be?
FractalChainsaw
Understandable.
StrawberryJamJam
The danger of AI is capitalism.
SpecialContainmentProcedures
Yes and no. It's authoritarianism, which can root in any form of society. Either the idea selfishly or selflessly that "I should be in charge."
elganif
Authoritarianism, such as exists in China would be more careful to make the AI's goals align to support its regime instead of just making the cheapest thing as fast as possible to beat competitors. In some ways that is actually better because more care will be taken regarding end goals - even if the intent is worse. Watch Robert Miles to really understand why https://www.youtube.com/@RobertMilesAI
Mehlbox
It‘s the combination of greed and recklessness, striving to outcompete your rivals. All trademarks of capitalism.
ERKNEES
Grimmrog
go home robot, you are drunk..
arrbos
o/` troy and abed in the singulaaaaaarity o/`
isaidsomethingcool
Two, weeks. .. two, weeks. ..
madrush
Is that from Spiderman? I can hear the voice in my head
isaidsomethingcool
Total recall