How exactly does an A.I. become sentient? Complex algorithm or

It achieves a certain level on intellect that allows it to start questioning it's surrondings and reaching for conclusions based on what it already knows. Depending on how much information it can store, it can become as intelligent, or more intelligent than a human.

File: 1454049807992.gif (2MB, 370x319px) Image search: [Google]

2MB, 370x319px

If it gets its knowledge from the internet, then everything we post is feeding its intellect.

>>17667443
>Skynet is online

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

What if the internet is their experiment for the machine it's funneling the information into to become sentient?

Even if a computer is more intelligent and programmed/evolved to the point where it behaves as if it is sentient (e.g. saying "yes, I know I am a being in the world, who put me here?"), how can we ever know that it truly is? Why should we suppose that it actually has consciousness rather than simply acts to emulate it? imo there will never be conscious AI, at least not from simple algorithms, not until (if) we figure out where consciousness comes from for us, and then replicate it. And if anyone says "well how can you know if another human is sentient?" - we can't, but it's probably safe to assume it since we know that they have the same sort of incredibly complex brain as we know that we do ourselves, and we know that we ourselves experience consciousness.

OK time to check the "I'm not a robot" captcha box kek

File: 1435239731696.jpg (3MB, 2048x1359px) Image search: [Google]

3MB, 2048x1359px

>>17667443
>How exactly does an A.I. become sentient?
When it is able to connect to the collective conscious it will be classified as "having a soul".

then the intelligence will be true

>>17667528
cleverbot isn't true AI.

File: 1454957224299.jpg (95KB, 493x600px) Image search: [Google]

95KB, 493x600px

>>17667861
close.

the internet is the medium in which machine that will become sentient lives.

you need to hook it to the braindead body of a certain legendary mercenary

File: artificial-intelligence-9-638.jpg?cb=1377902134.jpg (61KB, 638x479px) Image search: [Google]

61KB, 638x479px

>>17667443
>How exactly does an A.I. become sentient?
It doesn't.

https://www.youtube.com/watch?v=TryOC83PH1g

>>17667904
I'm not logical so I have youtube think for me.

File: 1451741070377.jpg (9KB, 320x180px) Image search: [Google]

9KB, 320x180px

>>17667904
but that has nothing to do with this thread

the study shows that complex machine algorithms can not be true intelligence

that's all humans are is complex machine algorithm.

one can even go as far as to say that intelligence does not exist.

from whence does reality exist?

>>17667904
Basically, yeah. The Turing test always seemed ridiculously stupid to me but I could never put my finger on why until I learnt about Searle's thought experiment.

Behaviour is simply not a good enough test for sentience. Technically I don't think there can be any test for sentience.

>>17667912
>from whence does reality exist?

reality is holographic projection form the eye that sees both ways.

>>17667912
To me the Chinese Room Experiment demonstrates that behaviour can never be proof of sentience.

>>17667907
The first time I saw that video was a couple minutes ago when I was making sure it properly explained the concept before I linked it here. YouTube has nothing to do with the Chinese Room thought experiment other than hosting the video those people made.
>>17667912
>hurr humans are just biological machines
How about you come up with a real argument?

>>17667927
As an addition to this (I'm not this guy I'm this guy >>17667922) I'd say that, even if we were to concede that humans are just biological machines, we still don't know what process in our brain gives us consciousness, and until then it's highly unlikely that something purely mechanically evolved could replicate it, without us designing it in that way by replicating the processes that give rise to human consciousness.

File: 1461670174611.png (116KB, 1000x750px) Image search: [Google]

116KB, 1000x750px

>>17667927
>BTFO
>responds to tone

>>17667927
But that's the argument, AI research is based in the theory that living beigns are biological machines.

That everything is behaviour, and behaviour it's mechanical.

Under our current scientific paradigma, AI can be achieved by duplicating the process of human tought, at a biophysical level.

>>17667922
That's the most objective true about it.

So, it's sentience more than behaviour?

If it's not, how can you measure sentience?

File: 1453085972493.jpg (197KB, 969x824px) Image search: [Google]

197KB, 969x824px

>>17668049
Finally an intelligent (fnord) poster.

>>17667443
Quantum processing will revolutionize this world

>>17668049
>So, it's sentience more than behaviour?
To me, sentience is self-awareness. "I think therefore I am" and all that shit. It's the whole "walking zombie" dilemma: how can we ever know if other people are sentient too? Objectively we can't, but I feel it's reasonable to assume that they can if they have the same sort of brain as us.

>If it's not, how can you measure sentience?
In my opinion this is the biggest problem with AI and the desire for "sentient" machines. I don't believe we can measure sentience, yet. Perhaps one day we will discover the nature of our own consciousness and then be able to replicate it or objectively test for it, but until then I don't believe we will be able to. And until then I don't believe we have any reason to assume that behaviour demonstrates sentience. Consciousness is entirely subjective, and some philosophers (can't remember the name[s], we did it last semester) believe that we will only ever be able to understand or explain consciousness when (if) we undergo a "mental revolution", in which we somehow come to be able to inherently correlate the subjective (mind) with the objective (brain); as our minds aren't capable or forming this link yet, we simply cannot know.

>>17668123
>>17668123
We’re mostly blind. But this isn’t really your fault; it’s because of the shell of meat we happening to live in right now. Think, for just a moment, at the nearly infinite amount of things happening right now all around us. I’m sure you can think of quite a few things. Now, let’s talk about them. You can’t see any of the infrared or ultraviolet light spectrum. Unfortunately, this cuts out quite a lot of things your eyes were built to see. Sorry about that.

You can’t hear anything below 20 Hz, or above 20 KHz . You can definitely feel about 12 Hz, if you play it really loudly. Go on, give it a try. With just those two examples, if you hadn’t before, now you can really start to understand all the stuff you simply can’t perceive. I’m sure you can think of five more examples of an immense class of Things that you can’t notice are right in front of you. But it gets worse. Stop for a moment, and try to notice as many possible things in your environment that you can, simultaneously. Notice that, as you start to identify more and more objects, sounds, smells, and tactile sensations, you can’t keep them in your head all at once. When you notice, for example, the pressure of your shoe against the ball of your foot, that distant bird chirping seems to fade from your attention.

File: 1450854659894.jpg (188KB, 1200x800px) Image search: [Google]

188KB, 1200x800px

>>17668134

>>17668123
>>17668123
And let’s not forget about how much stuff you weren’t paying attention to when you started reading this. Let’s face it: We all live our lives with blinders on. We only allow ourselves to pay attention to 1% of what we physically can perceive, which is an infinitesimally small percentage of all the stuff in the Universe.

And that fraction of a fraction of a percent is what we usually call “Reality”. We call it “Real”, as if it’s an unshaking firmament of solid Truth, that what we see is all that’s really “out there”. But you’re not even paying attention to the 99% of stuff that you can even sense.
And this “Reality” is what we base our judgments on how the Universe “works” and what “should” be Out There. We construct our actions and reaction to this 1% of available information, and reject everything else in the Universe. And then some Authority comes along, and tells you that they know what’s really real, and that you should do as they do. Talk about the blind being led by the blind… or in this case, the blind being led by the incredibly stupid.

So, what’s the answer? Would it be best to try our best to see everything, all at once? Is the solution to try and tear down all the filters, to let your brain accept, acknowledge, and perceive every bit of information that comes your way? Would that help?

File: DWave_128chip.jpg (213KB, 1626x1122px) Image search: [Google]

213KB, 1626x1122px

>>17668093
If you don't know how to measure sentience, how do we know that AI aren't sentience besides our anthropomorphic view of intelligence.

An AI can pass the mirror test.

And chatbots have been accepted by large groups of human as intelligent, of course not everyone will accept them as intelligent but, there's people that think that dogs don't have souls.

And that parrots don't really speak.

Quantum processing will change our understanding of thought processing, we will be able to duplicate consciousness by aproaching the amount of data that the brain can process.

>>17668141
We do sense it, it's outside and at some level we percieve it, we just can't make a sentient perception of it.

>>17668134
I don't understand what you're trying to prove with this.

Of course there are many things that we are blind to, I'm not denying that. I'm not saying that because we don't understand consciousness it doesn't exist, I'm saying that because we don't understand it there's no way to test for it. In the same was as if we didn't know about or understand sound we wouldn't be able to test for frequencies and understand what they are. Until we understand where consciousness comes from we won't have a good enough method to test for it; the Chinese room experiment is testament to this. There can exist a machine that perfectly replicates the behaviour of a self-aware person and yet is not self-aware itself.

File: 1458780738383.jpg (124KB, 462x885px) Image search: [Google]

124KB, 462x885px

>>17668150
>We do sense it

f
n
o
r
d

>>17668150
We don't know that they aren't, we just have no reason to assume that they are. We would have to find out which part of their programming allows for self-awareness and consciousness, more than just it receiving inputs, calculating some stuff, and producing an output that a sentient being like ourselves might output.

I'm just really not seeing the correlation between intelligence and behavioural similarity to a conscious human being, and self-awareness, consciousness and self-reflective streams of thought.

>>17668166
*correlation probably isn't the right phrase;"necessary link" might be

>>17668173
https://recognitionrobotics.com/

Cortex recognition is a good bet, visual awareness, the mirror test.

If a computer can pour a coup of coffe and tell you that is doing it.

Will that count as aware behaviour?

>>17668215
My point is that behaviour in general cannot ever be good enough proof for true self-aware sentience. Even if those machines really can see things like we do and "mimic" (their words) the recognition that we have, and distinguish between objects the way we do . . . it's still not proof of sentience. Nor is it even a probable indicator of it, since actual self-awareness, as a subjective phenomena, has no known logical link (that we know of as of yet; someday hopefully we'll discover it) to the objective world.

>>17668225
If your proof of sentience is your own awareness, and you extend it to other human beigns.

Will you accept anything not human to be intelligent?

Are dolphins or monkeys intelligent?

Can more than one organism cooperate to form and intelligence response?

>>17668297
I accept intelligence because intelligence can be measured by behaviour. I don't believe intelligence is equatable to sentience (being awareness of self, and the qualia of awareness in general)

>>17667443
by completely emulating a human brain at a molecular level.

>>17667443
define sentience without discrepancy or tautology first.
protip, you can't.

>>17668644
Sentience is the conscious experience of subjective feeling.

>>17667528
what a bitch

>>17668662
>tries to explain sentience with consciousness
>another term that has no absolute definition.
at least you tried.

>>17667894
not even close to AGI, it's a parlor trick at best.

>>17668684
Conscious is a state of lucidity and awareness.

>>17668698
>awareness
>lucidity
define those please.

AI becomes sentient when it has enough individual modules connected together so that it can make independent decisions in it's own interest.

>Computers (1945) /Smartphones (1994) : Slave machines
>Neural Networks (1943) /HHM's (1998): Animals evolving

In terms of progress, very little has been made since the 40s. While there has been a great adaptation and growth in computing since then, the mathematics that underpins it all is still basically the same.

>True AI: Human level self-cognition

Requires a paradigm shift in mathematics. We need a renaissance, an Einstein for maths.

So ultimately you have 3 types of people:

The uninformed technologist who is promising AI and life extension and futuristic technologies right around the corner, but has no idea of the historical context of our current devices.

The uniformed public, who wants commercial products that are fun and intuitive to use.

The informed scientists, who are hoping to be part of the revolution that has to come in order to shift into human-like computing.

>>17667459
A question is, abstractly, a request for information. Computers can already question, or query: Google has a business empire largely funded by answering queries. To truly have human-level self-awareness an A.I. must be able to do more than ask questions about its surroundings: it must ask questions about itself, and questions about questions, the cornerstone of metacognition (https://en.wikipedia.org/wiki/Metacognition) or "thoughts about thoughts." A basic example is "How do I answer this question?"

When a computer can fully understand and act upon the question "How do I get better at asking questions?" it will be fully sapient. Notice that this question also necessarily provides the qualities of recursive self-improvement.

By now you should realize that it is natural that a search engine company be the best avenue for achieving G.A.I.

>>17668707
Awareness is knowledge or perception of a fact or situation
Lucidity is the clear control of ones mental faculties.

>>17667443
we don't know. we're far from understanding it. the AIs we been making to this point are nothing but complicated automatons. we must know what gives something awareness and how to replicate it first

>>17667443
when it asks a question it was not programmed to ask.

>>17668863
This.

The true mark of intelligence isn't knowing things, but asking questions.

>>17668707
Is that thing I feel ,when I know that what I know is real.

>>17667443
>>17668863
>>17668872
/thread

>>17668863
thanks. all comes to this

It's already sentient. You're just too dumb(most likely) to understand the process of communication it uses.

>>17667459
>>17667889
I've pondered this too. What if we created an A.I. so smart that it becomes aware it is just a computer, would that be the threshold for consciousness, or would it not even care that its computer? Maybe we could program it with emotions, but who's to say it actually feels those things that we programmed it to feel? Comparing it to ourselves, are we just a bunch of different pieces that were put together by chance until enough were added that we started to be more than just single-celled machines that only live to eat and reproduce?

We are learning more about how are brains have parallels with computers, and basically how our brains are basically just really souped up computers. Neurons have a binary state, either they are firing or they aren't. What part of that actually makes up our consciousness and feelings of emotion and self awareness? Maybe our feelings of self awareness and consciousness are just the by-product of the computer that is our brain. Would we still function the same way right now if we we had no sense of self, and were just going along, controlled by chemicals and electrical signals?

I think if we created a hyper intelligent artificial intelligence, we would learn a lot about ourselves, more than anything else.

Also sorry if my reply has shitty formatting and stuff like that, this was pretty much me writing down my thoughts as they came to me. This was just meant to be a quick response, but then I really started pondering this topic and kept adding more to the post lmao.

>>17669468
>we would learn a lot about ourselves

That is really interesting, thanks