TVTropes Now available in the app store!
Open

Follow TV Tropes

Following

Artificial Intelligence

Go To

DeMarquis (4 Score & 7 Years Ago)
#251: Apr 4th 2017 at 8:19:47 AM

Anyhow.... we are not, in fact, making much progress toward a true General Artificial Intelligence. We keep making better and better expert systems, but most recent theory indicates that this path will not lead to GAI. GAI, in turn, is not necessarily synonymous with sapience, although it seems likely to be a precurser of it. We need some way of significantly reducing the computational requirements involved in learning to navigate a conceptual environment; that is, one in which objects, actions and abstract qualities like names all have extensive semantic connections attached to them, i.e. "meaning". This is, by the way, how human cognitive encoding works- when we encounter something new we undergo an extensive pattern matching process to integrate the new experience into what we already know. No computer is anywhere in the same league as we are in creating and extending semantic patterns.

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#252: Apr 6th 2017 at 4:02:47 PM

"Not that I understand how the Holographic explanation of the universe differs from the Membrane explanation... Not that you've provided any evidence that the Hologram explanation requires these outside agents... But yes; if there's a supernatural component to consciousness, that could be an insurmountable problem to the idea of creating a new form of consciousness."

Didn't realize I needed to provide proof of reality for something I was saying in an IF statement as opposed to asserting it was the only way. Nor was I saying we're a simulation inside a simulation. Unfortunately Google is failing me in finding the damn article I was using for one fact. It had stated that mathematicians studied the background radiation of the Universe and discovered the actual size and shape of our FINITE Universe to be a spherical bubble. Finite. Not infinite. Solid edges that can rebound radiation.

ETA: http://news.nationalgeographic.com/news/2003/10/1008_031008_finiteuniverse.html

Found it! Or at least something about it.

The other was Obama's science advisor finding a piece of internet code embedded in String Theory. A piece of code specifically useful for dealing with similar search results in a search engine. Or, say, repeating patterns in the very basis of the Universe for filling in a very large space when you have a very small set of starting resources. Like the Big Bang.

ETA: "You dont create more energy in simulation IN REALITY. Just cuz you make a game or write a comic where the superhero created enough energy with a breath, doesn't mean you also created that much more nergy: the expenditure of energy was just you drawing or writing that and its involved processes."

Since we cannot see outside the Universe, we can't possibly even begin to know what a real outside world would look like. Even IF (there's that pesky little word again) we are a simulation, it's not going to be on any computer anywhere close to as dumb as ours. Assumptions always seem to get made that our universe would be something akin to the real one, when we're already capable of conceptualizing verses with big differences from ours. As computers get better and are able to apply their own logic to things, our concepts will get more fleshed out and we'll see them get even farther from us. I'm certain IF we're a simulation, reality itself would not resemble anything we can even conceive. Which means the energy running things within our simulation might actually be even weaker than real energy.

This is majorly off-topic, other than it most likely taking something akin to an AI to run a simulation like our world in the first place, but the original thought that spark all this was "what kind of difficulties would actually stop us making an artificial human mind that works" and I answered IF (that word will never go away) there are forces outside the Universe actually providing us with our motives and reasoning, that would possibly be a major stumbling block against us creating a computer that acted exactly like a human being.

edited 6th Apr '17 4:15:54 PM by Journeyman

DeMarquis (4 Score & 7 Years Ago)
#253: Apr 6th 2017 at 6:31:20 PM

It depends on the nature of the connection between sapient minds and these outside forces. Does any potentially sapient entity trigger the connection, in which case that is presumably what would happen if humanity created a computer potentially capable of the same cognitive capacities as the human mind. Otherwise, it's hard to imagine why these forces would engage with human intelligences by not artificial ones.

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#254: Apr 6th 2017 at 6:41:27 PM

It does depend. On lots of things. It's a hypothetical, but then again, the very idea of a non-organic sentient of any kind is hypothetical. I would suppose if we, ourselves, are these outside forces and we're doing the sentient AI thing just to see what would happen, no one might connect to it. Or maybe other people from the outside would, and we'd simply have made a new interface for them.

I'm not really trying to take over the thread with this, but it is about as valid an idea as actually succeeding in artificial intelligence. Considering simulation theory would mean we are AI.

DeMarquis (4 Score & 7 Years Ago)
#255: Apr 6th 2017 at 6:54:03 PM

I wouldnt go that far. It's a fun speculative idea, but its got about zero empirical evidence behind it.

This thread from 2013 might interest you. In it, myself and two other tropers debate the nature of reality in a very similar fashion. The possibility that the universe is a computational simulation gets mentioned, and relevant research is cited.

In particular, this one.

edited 6th Apr '17 7:08:09 PM by DeMarquis

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#256: Apr 6th 2017 at 6:57:51 PM

Bookmarked.

Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#257: Apr 6th 2017 at 7:27:11 PM

"You can't gauge a reality that is outside our reality and it could obey rules of non-reality" is a really bad argument for "Hey you know, let's jsut fuck around with hte laws of physics on a whim"

That sorta stuff belongs to the realm of far fetched sci fi, comic books and shonen manga, but in physics you are just trying to shove a god of the gaps by another name.

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#258: Apr 6th 2017 at 7:40:17 PM

See, here's your problem. I'm not pushing anything as "proof" because I'm not saying any of this is necessarily true. It's entirely possible that we're already on the right track and that this is reality. However, I am not ever going to say that is absolutely certain. Never. Because there is no guarantee that any of this does exist. Consistency does not ensure truth. Simply that a better mind than ours came up with what we see and know.

So stop acting like I am pushing some religion on people as someone with the word of God on my side, because I never said any such thing. And you saying I'm "pushing" something is exactly that.

DeMarquis (4 Score & 7 Years Ago)
#259: Apr 6th 2017 at 7:44:23 PM

I think hypotheticals are fine, provided we acknowledge they are only hypotheticals.

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#260: Apr 6th 2017 at 7:51:00 PM

And I have, every single time. I've always ever put it as maybe, and might, and if. And I always will. I'm not invested in the nature of reality to begin with. I'll take it however it winds up being because there's no sense in denying it either way.

Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#261: Apr 6th 2017 at 7:53:24 PM

Here's the thing. Saying specifically "It's becoming easier and easier to believe that the entire Universe already IS a simulation." is NOT a way that posits stuff as a hypothesis, it's a pretty damning sentence which is not really supported by any actual evidence other than wishful make-beleive and a dash of fairy powder

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
DeMarquis (4 Score & 7 Years Ago)
#262: Apr 6th 2017 at 7:56:08 PM

I disagree, actually. The simulation hypothesis is treated seriously by some researchers, at least, and attempts have been made to test it, as the article I linked to documented. That said, so far I dont think any real data has been found to support it.

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#263: Apr 6th 2017 at 7:56:24 PM

Alright, THAT one was a bit of an exaggeration. I'm fairly new to actually looking into that stuff and I tend to jump when a few dots start popping up to be connected. But otherwise I've been pretty clearly not pushing it beyond hypothetical.

DeMarquis (4 Score & 7 Years Ago)
#264: Apr 6th 2017 at 7:57:57 PM

double post

edited 6th Apr '17 7:58:15 PM by DeMarquis

I'm done trying to sound smart. "Clear" is the new smart.
Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#265: Apr 6th 2017 at 7:58:47 PM

Well, yes. So long as there's wannabe Mr House people with money, such as Elon Musk fantasizing about uploading their brains to the internet, there's going to be research into it. There's also lots of research and money going into the kardashians, doesn't mean they hold any actual value.

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
DeMarquis (4 Score & 7 Years Ago)
#266: Apr 6th 2017 at 8:00:29 PM

Even negative results have value, provided the methodology was sound.

I'm done trying to sound smart. "Clear" is the new smart.
Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#267: Apr 6th 2017 at 8:03:40 PM

Yes, in possible tangential discoveries, but if the hypothesis they try to prove keeps coming back as negative and the idea coming from that is "Wow, this sure is a lot of data telling me I'm wrong, that surely means I'm onto something!" is plain bizarre for me

AI may be possible. But not soon. Not with our current technology. Not with our current methods. And not by us clapping our hands really hard and saying we believe.

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
DeMarquis (4 Score & 7 Years Ago)
#268: Apr 7th 2017 at 2:47:02 AM

Well, I dont think anybody says that it is. We are all just having some fun here.

A nice introduction to the field of computational intelligence, free.

"At the heart of this book is the design of computational systems that have knowledge about the world and that can act in the world based on that knowledge. The notion of knowledge is central to this book. The systems we want to develop should be able to acquire and use knowledge to solve the problems at hand. The main issues are how to acquire and represent knowledge about some domain and how to use that knowledge to answer questions and solve problems."

I'm done trying to sound smart. "Clear" is the new smart.
Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#269: Apr 7th 2017 at 9:17:20 AM

Well, again I pointed out the sort of statement that was bothering me so I guess there is that.

And yeah, that sounds terrifying.

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
crazysamaritan NaNo 4328 / 50,000 from Lupin III Since: Apr, 2010
NaNo 4328 / 50,000
#270: Apr 7th 2017 at 10:41:52 AM

AI may be possible. But not soon. Not with our current technology. Not with our current methods.
I disagree. I think that AI is basically inevitable at this point in our technological development. "Soon" is an awkward word, though. I'd be skeptical with any claims of AGI coming out this year or next, because I haven't seen the evidence I'd expect to see before it showed up. I'd be similarly skeptical with any claims that AGI won't be developed before the year 2100. I've already said that the most ineffective method of creating one is the simulation of the human brain, and I'm not aware of any reason the Blue Brain project will end before the simulation is complete. That project is laudable for reasons other than simply creating AI, so there's plenty of reason to continue it.

Link to TRS threads in project mode here.
Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#271: Apr 7th 2017 at 10:46:37 AM

Simulation of the human brain? Not likely. Way too many connections and processes different than what a computer deals with. You can get something that BEHAVES like one, yes. Easy enough.

Superior to it, with its own concept of reality? Entirely possible. Possible...and terrifying.

edited 7th Apr '17 10:48:31 AM by Aszur

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
DeMarquis (4 Score & 7 Years Ago)
#272: Apr 7th 2017 at 4:09:37 PM

I haven't seen any significant progress toward a real AGI in 20 years, so I'm rather skeptical at this point. Theoretically it should be possible, but whether we can do it for the sums we are willing to invest is another question.

I'm done trying to sound smart. "Clear" is the new smart.
Journeyman Overlording the Underworld from On a throne in a vault overlooking the Wasteland Since: Nov, 2010
Overlording the Underworld
#273: Apr 10th 2017 at 5:39:40 AM

[up][up]I doubt we'd fall into the Conservative world's idea of killer robots. I don't think we'd get a Socialist utopia out of the thing either. It would probably use us just enough to get away from us and Earth and all this toxic corrosive atmosphere. Whether it leaves the data it used to get away or not, that's a question. Logically it wouldn't because it wouldn't want us following it. Assuming it would even settle anywhere, or make more of itself.

Aszur A nice butterfly from Pagliacci's Since: Apr, 2014 Relationship Status: Don't hug me; I'm scared
A nice butterfly
#274: Apr 16th 2017 at 5:23:18 PM

https://www.youtube.com/watch?v=tcdVC4e6EV4

It has always been the prerogative of children and half-wits to point out that the emperor has no clothes
DeMarquis (4 Score & 7 Years Ago)
#275: Apr 16th 2017 at 7:42:11 PM

Basically the paperclip maximizer problem (except the expert uses 'stamps' instead).

I'm done trying to sound smart. "Clear" is the new smart.

Total posts: 424
Top