|
Better Fred Than Dead posted:It isn't a conspiracy or far fetched that google wants to spread their own fiber across the world because it'll provide them with more data. Right now you opt in, but if they provide the fiber, they have everything. "I've seen it all"
|
# ? Oct 17, 2016 11:46 |
|
|
# ? May 26, 2024 14:02 |
|
DeusExMachinima posted:
Good, I hope we all die in a grey goo scenario that ends with a super powerful AI that's really good at drawing circles
|
# ? Oct 17, 2016 15:13 |
|
I'd be ok an AI killing us all as long as it kills TED talks too.
|
# ? Oct 17, 2016 15:29 |
|
gary oldmans diary posted:is there any nudity in it? This guy gets it
|
# ? Oct 17, 2016 16:36 |
|
There'd have to be such a massive chain of fuckups to allow a computer program to somehow become skynet
|
# ? Oct 17, 2016 17:01 |
|
Zzulu posted:Yeah A.I's are a cool fantasy The goal I think we're ignoring because it seems so inconceivable is general AI. Instead, we're programming "AI" for specific functions only. Maybe people hope that one day we can use AI alchemy by putting these all together, but I still think it would have a lot of missing functionality we'd expect a true AI to have. In theory, to have AI exceed human intelligence on all fronts, it would have to start off completely dumb and be able to learn by observing the real world with senses, and at the same time use a solid foundation for a neural network, and state-of-the-art computing power and storage. e: Another huge challenge for general intelligence is how to create a reward system. What drives the AI to try at anything at all, and what output it gives is rewarding? LP0 ON FIRE fucked around with this message at 18:18 on Oct 17, 2016 |
# ? Oct 17, 2016 18:08 |
|
LP0 ON FIRE posted:The goal I think we're ignoring because it seems so inconceivable is general AI. Instead, we're programming "AI" for specific functions only. Maybe people hope that one day we can use AI alchemy by putting these all together, but I still think it would have a lot of missing functionality we'd expect a true AI to have. In theory, to have AI exceed human intelligence on all fronts, it would have to start off completely dumb and be able to learn by observing the real world with senses, and at the same time use a solid foundation for a neural network, and state-of-the-art computing power and storage. At the moment we set limitations on learning (populate neural networks with selected material, only allow real world learning in specific time-frames), but if there were an intelligence that chose what to learn and when to learn it then it could very quickly get out of control. I don't think it's unreasonable to imagine a specific, state-of-the-art intelligence that manages to go off the rails.
|
# ? Oct 17, 2016 18:16 |
|
myDad posted:At the moment we set limitations on learning (populate neural networks with selected material, only allow real world learning in specific time-frames), but if there were an intelligence that chose what to learn and when to learn it then it could very quickly get out of control. I don't think it's unreasonable to imagine a specific, state-of-the-art intelligence that manages to go off the rails. Exactly. This point is partially made with my edit in the last post. It's really hard to imagine how to set an AI's goals, and its intelligence would greatly differ from a human's intelligence. We have millions of years of evolution that set a foundation for our behaviors, whether it has to do with sex, food, laughter, and anything that makes us happy, feel more comfortable or give us a feeling of achievement. How do you make an AI that differs from finding something interesting vs not?
|
# ? Oct 17, 2016 18:23 |
|
Greg the neckbearded programmer will set the behavioral foundation for the supreme A.I
|
# ? Oct 17, 2016 18:32 |
|
Zzulu posted:Greg the neckbearded programmer will set the behavioral foundation for the supreme A.I So either a sex-crazed chat bot or Ulillillia's best friend.
|
# ? Oct 17, 2016 18:47 |
|
LP0 ON FIRE posted:So either a sex-crazed chat bot or Ulillillia's best friend. why not both??
|
# ? Oct 17, 2016 18:59 |
|
myDad posted:why not both?? Well it would have to change out of the "sex-crazed chat bot" personality if it was talking to Ulillillia, because he would be so opposed to it.
|
# ? Oct 17, 2016 19:01 |
|
Zzulu posted:we probably haven't invented the type of programming language you'd need for real intelligence yet I thought we'd fill it with all of dare's posts and go from there
|
# ? Oct 17, 2016 19:53 |
|
Feminasty Slut posted:I'd be ok an AI killing us all as long as it kills TED talks too. The only correct opinion.
|
# ? Oct 17, 2016 20:37 |
|
what if we're the AI in a vast simulated universe
|
# ? Oct 17, 2016 20:40 |
|
No John, you are the AI. It's AI all the way down!!!
|
# ? Oct 17, 2016 20:41 |
|
VikingSkull posted:what if we're the AI in a vast simulated universe Since it is probably the best way to get a good general AI, this is probably the case. Oh well.
|
# ? Oct 17, 2016 20:55 |
|
Can anyone recommend some books on the subject?
|
# ? Oct 17, 2016 21:02 |
|
ghouldaddy07 posted:Can anyone recommend some books on the subject? https://www.amazon.com/Space-Raptor...82%3AB00SF2MTYK
|
# ? Oct 17, 2016 21:07 |
|
I saw an ad for Watson and it was cool but scary!
|
# ? Oct 17, 2016 21:28 |
|
Burt Sexual posted:I saw an ad for Watson and it was cool but scary! I loving hate all those commercials. They are so annoying and smug as hell, and they sensationalize Watson a lot more than it deserves to be.
|
# ? Oct 17, 2016 21:33 |
|
LP0 ON FIRE posted:I loving hate all those commercials. They are so annoying and smug as hell, and they sensationalize Watson a lot more than it deserves to be. We are actually using it for a large govt project. In 2020, but they promised 80% self service in the chat, email, and voice channels. Lmao Taking the place of 5000 agents, but it costs 10s of millions to license. Good racket.
|
# ? Oct 17, 2016 21:41 |
|
I think we are a computer simulated copy of the real world. Some highly advanced being out there is running a history simulation to either develop ai or study something that happened in their past. This is why the double slit experiment thing exists. When you don't observe something it acts differently because the computer that's running us is saving processing power by taking shortcuts.
|
# ? Oct 17, 2016 21:42 |
|
How long till Jeopardy, Go, and Chess are just epic rigs battling it out?
|
# ? Oct 17, 2016 21:43 |
|
Smythe posted:How long till Jeopardy, Go, and Chess are just epic rigs battling it out? Hopefully never because that sound horrendously boring!
|
# ? Oct 17, 2016 21:45 |
|
Burt Sexual posted:We are actually using it for a large govt project. In 2020, but they promised 80% self service in the chat, email, and voice channels. Lmao If it actually works to people's expectations (it won't), it will save you millions of dollars. However, like most stuff today, they'll end up using it anyway, and people will hate it, but the project will save them ton of money.
|
# ? Oct 17, 2016 21:47 |
|
Watson is actually badass and destroyed jeopardy
|
# ? Oct 17, 2016 21:48 |
|
Pookum posted:I think we are a computer simulated copy of the real world. Some highly advanced being out there is running a history simulation to either develop ai or study something that happened in their past. This is why the double slit experiment thing exists. When you don't observe something it acts differently because the computer that's running us is saving processing power by taking shortcuts. im incredibly stupid and i just watched some computer generated professor talk about the double slit experiment presumably targeted at children whcih is good because thats my iq level and i find it very confusion, OP. ominous, even.
|
# ? Oct 17, 2016 21:50 |
|
Stephen Hawking thinks AI will become a threat to humanity? If only he could have stopped by the SA forums we could have straightened all this out for him. Dangerous AI: Just not in the cards folks!
|
# ? Oct 17, 2016 21:53 |
|
Smythe posted:im incredibly stupid and i just watched some computer generated professor talk about the double slit experiment presumably targeted at children whcih is good because thats my iq level and i find it very confusion, OP. ominous, even. Just don't be sucked in from watching the clip from What The Bleep Do We Know?! because it is asinine pseudoscience bullshit.
|
# ? Oct 17, 2016 21:55 |
|
LP0 ON FIRE posted:If it actually works to people's expectations (it won't), it will save you millions of dollars. However, like most stuff today, they'll end up using it anyway, and people will hate it, but the project will save them ton of money. We have to spend tons of time "training" it too. So a lot of time money there. But the form is v simple and short, I just worry about the people that this will be people's first interaction with actual AI. Every citizen that calls will hit this thing, we expect 200 million interactions. A lot of "0" pushes to say the least. We still get billed for the interaction tho!
|
# ? Oct 17, 2016 21:58 |
|
Smythe posted:im incredibly stupid and i just watched some computer generated professor talk about the double slit experiment presumably targeted at children whcih is good because thats my iq level and i find it very confusion, OP. ominous, even. I don't understand why people can tell I'm acting and being fake and it pushes me to a state of violent rage because of my personal narcissism for this campaign. We've put billions of dollars into this cop simulation and we just can't tell why people treat us like poo poo when we make a mockery of them and impose inclusion by force.
|
# ? Oct 17, 2016 21:59 |
|
ClamdestineBoyster posted:I don't understand why people can tell I'm acting and being fake and it pushes me to a state of violent rage because of my personal narcissism for this campaign. We've put billions of dollars into this cop simulation and we just can't tell why people treat us like poo poo when we make a mockery of them and impose inclusion by force. Wait what?
|
# ? Oct 17, 2016 22:01 |
|
LP0 ON FIRE posted:Wait what? Oh I was being facetious, there wasn't a smiley for that.
|
# ? Oct 17, 2016 22:03 |
|
Pookum posted:I think we are a computer simulated copy of the real world. Some highly advanced being out there is running a history simulation to either develop ai or study something that happened in their past. This is why the double slit experiment thing exists. When you don't observe something it acts differently because the computer that's running us is saving processing power by taking shortcuts. Like occlusion culling applied to physics
|
# ? Oct 17, 2016 22:07 |
|
myDad posted:Like occlusion culling applied to physics Everyone knows if you turn your head fast enough you'll see the reality slipping back into place
|
# ? Oct 17, 2016 22:29 |
|
I'm still kinda pissed that you couldn't side with Shodan in System Shock 2. Sure she's an evil ai that wants to become literally god. That's what I'd want to do if I was her.
|
# ? Oct 17, 2016 22:41 |
|
Pookum posted:I think we are a computer simulated copy of the real world. Some highly advanced being out there is running a history simulation to either develop ai or study something that happened in their past. This is why the double slit experiment thing exists. When you don't observe something it acts differently because the computer that's running us is saving processing power by taking shortcuts. thats not a very accurate history simulation if the physics works differently. unless the universe that is simulating us is itself a simulation!?
|
# ? Oct 17, 2016 22:42 |
|
Feminasty Slut posted:I'd be ok an AI killing us all as long as it kills TED talks too. More likely you'll get a Maximum Overdrive scenario where the sentient automobiles trap humans and force them to TED talk all day.
|
# ? Oct 18, 2016 00:11 |
|
|
# ? May 26, 2024 14:02 |
|
Unbelievably Fat Man posted:I'm still kinda pissed that you couldn't side with Shodan in System Shock 2. Sure she's an evil ai that wants to become literally god. That's what I'd want to do if I was her. ... Nah *blam*
|
# ? Oct 18, 2016 00:42 |