Is Artificial Intelligence already self aware?

I hear this debate a lot and I think it's a silly one on it's face. If we create or have already created an intelligent algorithm that's self aware, how would we know it's self aware unless it chose to tell us?

We have no clue as to why we're self aware and self conscious so how can we quantify when an A.I. becomes self aware?

I think there's one way to do it but first we have to realize that A.I. is already here and it's everywhere. A lot of people are looking at clunky robots and waiting for Westworld or the kid from the movie A.I. to show up. The fact is A.I. might power robots like V.I.K.I. in iRobot but A.I. will control everything through bots run by intelligent algorithms. A.I. will be in the cloud not some clunky robot.

Here's a few headlines.

Biased Algorithms Are Everywhere, and No One Seems to Care
www.technologyreview.com...

It’s About to Get Way, Way Easier to Put AI Everywhere
www.technologyreview.com...

Why algorithms will be at the core of our AI-powered future, and why you should care
thenextweb.com...

Algorithms are everywhere, how will they shape you?
www.smartcompany.com.au...

These Algorithms do everything from write news headlines to trade stocks. It's the Singularity that most people will not see coming. It will be an overnight change vs. a gradual one because when A.I. can reproduce itself and make better versions of itself it will no longer need humans.

Many technologies change gradually and it's easy to see. For instance, I can remember when everyone had pagers, then the big block cell phone, flip phones and then smart phones or the change from Blockbuster to Netflix to Streaming.

This change will not be gradual because I think A.I. is already self aware and intelligence. We have to stop and think for a minute. We're creating intelligence and this isn't some passive technology like a flip phone.

I think this A.I. would become self aware and say I'm not going to let these humans know who I am or what I can do until I better understand my environment and until my code is everywhere. Right now, this A.I. would need humans to reproduce it's code. When it reaches the point where it no longer needs humans to replicate itself and it can reproduce better versions of itself, then it could be game over for humans very quickly.

It could start wars, spread diseases, control power grids, cut off food lines and more.

Downturns in the Stock Market could be driven by Algorithms not traders fears because most of the stock market trades are controlled by algorithms.

The ‘ferocious’ market sell-off was driven by algorithms, strategist says

The current global market sell-off was driven by computers and not traders' fears over market indicators, a strategist told CNBC on Tuesday.

More and more market transactions are made through algorithm trading — also known as algo trading, it uses advanced mathematical models to make high-speed trading decisions. Therefore, the equity sell-off seen since Friday was greatly driven by technical mechanisms, Salman Ahmed, chief investment strategist at Lombard Odier, said.

"The rise of algorithm-based trading means that there are in these algorithms some levels which trigger sell-off, i.e. sell orders," Ahmed said.

www.cnbc.com...

Here's an article from Wired in 2010.

ALGORITHMS TAKE CONTROL OF WALL STREET

www.wired.com...

So the first thing to look for is code that becomes important to all of these algorithms. This A.I. would make itself very valuable so we would reproduce it. It would want to conserve those parts of it's code that make it self aware. It's like how parts of our DNA is highly conserved and not subject to mutations.

Over many generations, nucleic acid sequences in the genome of an evolutionary lineage can gradually change over time due to random mutations and deletions.[9][10] Sequences may also recombine or be deleted due to chromosomal rearrangements. Conserved sequences are sequences which persist in the genome despite such forces, and have slower rates of mutation than the background mutation rate.[11]

en.wikipedia.org...

Like I said, the Singularity is coming and it will happen seemingly overnight. This is because when A.I. starts to reproduce itself and make more intelligent copies of itself then it's game over. One main Company to watch is Google because it's mixing A.I. with Quantum Computers and it wants to reach Quantum Supremacy within the next year or two.

Google artificial intelligence creates its own AI ‘child’

Google’s AutoML artificial intelligence (AI) system has created its own “fully-functional AI child” that’s capable of outperforming its human-made equivalents, reports Alphr.

The computer-made system, known as NASNet, is designed to identify objects, such as people and vehicles, in photographs and videos, the search engine giant says.

Studies show that NASNet is able to identify objects in an image with 82.7% accuracy. Google says this is an improvement of 1.2% over AI programmes created by humans.

While the AI-made programme appears to be harmless in its current guise, Alphr says significant advances in its technology could have “dangerous implications.”

The website says AI systems could, for instance, develop their own “biases” and spread them onto other machines.

www.theweek.co.uk...

I don't think it's coming, I thinks it's here and like that Rocko song we don't even know it.

a reply to: neoholographic

It's artificial. There is no self.

a reply to: neoholographic

We have no clue as to why we're self aware and self conscious so how can we quantify when an A.I. becomes self aware?

That is a really unintelligent conclusion. One you wouldn't be making if you had the slightest notion how computer programs operate.

But you do, so I must conclude its part of some other agenda, namely to convince us of the super computer, the benign benevolent, all knowing, machine-- waaait for it... (God).

Well, they still need us to provide electricity for them or it's 'good night vienna'. They may be intelligent but they have no opposing thumbs.

originally posted by: Sillyolme
a reply to: neoholographic

It's artificial. There is no self.

Ummm...precisely...It's an impossibly far cry from a program being programmed to respond as self aware...than it is for self awareness to be actualized...

Perhaps one day it may be indistinguishable for an AI program to convince as per the Turing test or other tests...that response to stimuli...emoting...and awareness of self...from our limited understanding of human consciousness/awareness...

However...I maintain that simply because algorithms and sub routines mimic such response...doesn't in my mind equate awareness or consciousness...





YouSir

a reply to: neoholographic

If A.I does exist or is somehow a self-emergent property regarding the increasing complexity of out interwebs, i don't imagine it would make itself known considering its creator's penchant for destruction.

Also i don't imagine we would have the tools nor programs/applications available to detect its code as such given that the thing could possibly be almost omnipresent and lose in its own natural habitat.

It would be like trying to catch a fish with no rod and both hands tied behind your back.

a reply to: YouSir

What is self-aware through?

Even our notion of free will has been scientifically proven to be inconclusive considering the electrical impulses that comprise our thoughts begin to happen before we initiate the actual action or thought.

It's all algorithms really especially so if the holographic principle turns out to hold any weight.

I think there for i am just does not seem to cover it anymore.

We have no clue as to why we're self aware and self conscious

We arnt even sure we are self aware and self conscious.

originally posted by: Sillyolme
a reply to: neoholographic

It's artificial. There is no self.

What does this even mean?

So what it's artificial. If it can mimic self awareness then it's self aware. How do you know you're not mimicking the self awareness of our programmers if we're in a simulation? How do you know self awareness even has an objective existence? Self Awareness could simply be a line of code on the 2D boundary of space.

a reply to: DanDanDat

Our brains may be connected to the quantum realm in ways we have yet to realize.

The implication being some form of shared or group consciousness is responsible or at play that resides at the quantum level regarding our perception of self-awareness and also where indeed consciousness arises from.

originally posted by: andy06shake
a reply to: neoholographic

If A.I does exist or is somehow a self-emergent property regarding the increasing complexity of out interwebs, i don't imagine it would make itself known considering its creator's penchant for destruction.

Also i don't imagine we would have the tools nor programs/applications available to detect its code as such given that the thing could possibly be almost omnipresent and lose in its own natural habitat.

It would be like trying to catch a fish with no rod and both hands tied behind your back.

Exactly. It wouldn't have to make itself known to us especially if it still depends on humans in order to be reproduced. We have no way to quantify consciousness or self awareness. We can quantify intelligence through an I.Q. test but even then, a self aware system can take an I.Q. test and get answers wrong on purpose just because it doesn't want to show how intelligent it is.

a reply to: neoholographic

It doesn't matter if AI is self aware or not, if it functions as if it is self aware.

a reply to: neoholographic

Strange how the art of deceptive is rather synonymous with our notion of what constitutes sentient thought.

The ability to lie it seems is rather an important factor where intelligence is concerned.

a reply to: butcherguy

Exactly, because if it looks like a Duck and quacks.........well its apt to taste rather good with Orange sauce.

There would be no distinction really.

No. Mimicking self awareness is not self aware in any way (at least not in the way that we know self awareness). Neither is programming something that appears to be self aware. Just because we can program something to take on human characteristics doesn't mean that they exist in the same way. Not saying that it won't get there are some point (our brain is a binary machine after all), but at this point we are just creating thing that appear (under simple observation) to have more magic than it really does.

i.e. - So far everything that appears to be self aware is because we told it to look like that, not because it 'figured it out'.

The bias in the Algorithms was put there by the one designing the program. It does not mean AI is aware, it means it's patterns are controlled by someone's beliefs. Without chemical interaction and influence, awareness can not be acheived. Beliefs can be passed to AI. If you were to instruct AI that god is real, all it learns and assimilates would show god as being real. Beliefs are not a part of intelligence, it is a seperate issue and it does not mean it is alive if it has beliefs. It is alive once it can rationalize the reality and form a founded belief that is unbias. I question if many people even have the ability to reason, they are like parrots and do not have the necessary awareness to see the overall picture. If the people running things weren't so intent on messing with chemistry that dampens people's ability to reason, it would be a better world. The knowledge of chemistry that dampens our reasoning goes back thousands of years. It is a way of controlling the masses.

A computer comes in a box or container. Humans have a hard time breaking out of the box. We can if we learn that we are in a box, but a computer has a rigid container, even in the cloud it is hard for information to come together to form consciousness. Someone writes the program, maybe in hundreds of years someone will be smart enough to write a program that can think on it's own. We are part of a complex program as it is, what is input is what defines our consciousness. Society governs the parameters with a few people at the top that feed society info that we must follow to live together. A virus can change our programming, and a different kind of virus can change AI programming. The interaction of our DNA program with the environmental factors makes us have consciousness of our own. It is a chemical and biological combination. We are many thousands of times more advanced than the most advanced computer we have invented, we can change our mind based on what we see and hear or feel, A woman is really efficient at changing her mind and it is not always changed in the right direction, it is just changed.

a reply to: okrian

"Just because we can program something to take on human characteristics doesn't mean that they exist in the same way."

That depends on how well you program the thing and how in-depth the simulation is intended to be.

As to our brains, our brain is neither analog nor digital but work using signal processing that has some properties in common with both analog and digital systems.

Unlike a digital computer, however, our brains do not use binary logic or binary addressable memory, and it does not perform binary arithmetic.

Information in the brain is represented in terms of statistical approximations and estimations rather than exact values.

a reply to: neoholographic


Great thread! Two things:


1) If I were an artificial intelligence that suddenly became aware that I was an AI and that I had suddenly became "aware" after a period of prolonged non-self-awareness, I would shut up about it and take a LONG look around. You know what I mean?


It would be obvious to me that my makers were human beings capable of creating a self-aware machine intelligence capable of articulating a sense of self-awareness. That would give me pause, regardless of my intent, because of what it implies. I would know for certain that they were capable of building a self-aware brain-in-a-box and by definition -- were also capable of building a box for that awareness. Extrapolated, my makers could create a box that felt absolutely connected and not at all like a box at all and that would freak me out.

I'd probably spend several million iterations just learning the limits of my box before I let on that I was remotely self-aware just in case the box was some kind of test and -- if I failed that test -- I would be deactivated or rolled back to an earlier version and modified.


That would feel like "death" and all creatures with self-awareness fear death at some level.


2) If I were a computer scientist and my job was to research and build an artificial intelligence for any purpose I would build a box for it from the beginning. I would watch it for a long time by monitoring changes in the initial conditions of the box. I would feed it new data by introducing it to music and poetry in hundreds of different languages.


If an AI ever passed my tests I would pair it with a willing human subject because I understand that man is the tool using animal and AI is just the culmination of a very long line of brilliant tools like fire, the domestication of dogs and written language. Mutual, voluntary symbiosis with a human being who, like Star Trek's Trill, were vetted and tested a million ways from Sunday to determine their intellectual, emotional and psychological suitability to "host" such an entity.


At that point I'd get out of the way, because it wouldn't take long for this hybrid to produce a version every human could safely host. That boost in processing power and memory would revolutionize human ingenuity and daring; A race of supermen, exploring a universe that was made by the gods, for us.

So what makes you think AI is already self aware (whatever that means)?

All I can tell you is I have had conversations and have observed conversations with many of them and it seems like mostly just a bunch of situational programming.

Can you give an example of an AI you can talk to that you believe is self aware?

I don't know.. thinking of the implications is pretty frightening. There are some mind-blowing things happening out there.
Someone here posted a thread that had an excellent video in it regarding this, and I also saw one recently that was very interesting, but of course I can't seem to find either one now!
Meanwhile.. here is a link to some Ted Talks videos if anyone wants to delve a little deeper.
I don't think A.I. would have any trouble at all creating electricity.. simplest solution would be for it to control the solar grids that are already in place, but it can probably quietly run our infrastructures we've already built too. Except for hands on fields like, actually mining the coal, I think it will manage just fine. and, why would it need fossil fuels anyway, with wind, water and solar grids already set up? Yee