AI Zone Admin Forum Add your forum

NEWS: Chatbots.org survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

Will an AI become self aware?
 
Poll
When will an AI be self aware?
In my lifetime (less than 25 years) 10
Less than 150 years 6
More than 150 years 1
It will never happen 2
Total Votes: 19
You must be a logged-in member to vote
 
  [ # 16 ]

Additional food for thought:

The mirror test is used to judge self awareness in animals.

Wasps recognize other wasps faces. Do you believe in self-awareness in insects?

The way people recognize faces may have more to do with culture than nature. Did you know, some people do not have the ability to distinguish faces, face blindness is similar to color blindness.

 

 
  [ # 17 ]

> http://www.youtube.com/watch?v=TphFUYRAx_c

Check out this historic video, “QBO Robot in front of a mirror”....

 

 
  [ # 18 ]

Less than 25 years? Maybe I’m overly optimistic, because I feel no doubt it will happen in less than 25 years, like *far* less. I would have created poll choices like: 1 year, 5 years, or 10 years, and I probably would have voted for 1 year! Remember, this technology is *long* overdue, and quite likely rests only on breaking the “creativity barrier” rather than breaking any technology barrier.

Thanks, Merlin, for that interesting info about the mirror test. I was unaware of that test, and how it related to various animal species. My complaint about “self-awareness” as people are using the term to mean in this thread is that they are assuming it means self-awareness of body, but there also exists self-awareness of mind, which to me is the more interesting question. As that Qbo video shows, it is a fairly routine task for a vision system to recognize its own physical appearance, since it’s just another physical object in its environment. The real trick would be for a vision system to know that the physical object it recognizes as itself is performing the processing that is allowing itself to make that recognition, or better yet, to take an objective look at some of its own processes while those processes are going on, and to recognize their relationship to itself and its conclusions.

I had a friend who regarded the “Cogito ergo sum” insight as the ultimate in self-awareness, and that is the kind of self-awareness of mind to which I referring…

http://en.wikipedia.org/wiki/Cogito_ergo_sum

Yet even self-awareness of mind doesn’t seem like much of a hurdle to me, since a memory test or hardware test in an existing digital computer could allow a computer to look at its own processes to a certain degree. It’s possible the only difference is that such a hardware test lacks only a more global understanding of the relationships between the specific digital computations and their implied conclusions. It could be just a matter of using a larger number of internal sensors!

 

 
  [ # 19 ]

I’ll have a $100 bet with you that self aware AI doesn’t happen next year. Although, that would be a bet I would love to lose.

 

 
  [ # 20 ]

Steve, I’m surprised you didn’t offer a £100 wager, instead. smile

Frankly, given the scope and gravity of the challenges involved, I don’t see the emergence of a self aware AI entity happening within the next 5 years, let alone 1. That said, however, Like Steve, I’d dearly love to be proven wrong. smile

 

 
  [ # 21 ]

Mark, I think there might be machines that already qualify as “self-aware” by your definition. (Especially with regards to machines that monitor their own processes.) However, I think for most people, “self-aware” is a shifting goal post. Animals suffer from the same bias—any example of self-awareness (mirror test, etc.) only leads to a re-definition of what “real” self-awareness is, such that it appears solely human again.

I think it’ll only be settled on when machines are sufficiently versed in rhetoric to be persuasive enough to convince us. And that, I bet, is still a ways away. smile

 

 
  [ # 22 ]

Great topic and great insights. Its a complex question and one of the reasons that RICH was created. This is excerpted from a paper that I’m working on so its a little long winded (caveat in place!) but it raises an interesting (I believe) question regarding the mirror test. No doubt its been raised before, so if anyone can add to it please do!

The Dictionary definition of “self” is ;

Noun; A person’s essential being that distinguishes them from others, esp. considered as the object of introspection or reflexive action.

However do we truly understand what our human “self” is, and are therefore able to recognize it in something other than our…“selves”? The part of “me” which is sometimes referred to as the watcher, recognizes that I am thinking.  There is the thought, and something that is aware of that thought and is independent of it. Is that something also a thought? If I program a sub routine that is aware of what my main program is doing and can comment on it independently (an internal dialogue), have I created “awareness”?

Regarding the mirror test, here’s an interesting hypothetical question. If I was put into a coma, and had complete facial reconstruction without being aware of it, would I recognize “me” in a mirror when I awakened? Based on the answer to this question can we still consider the mirror test valid as it relates to self awareness?

I have a feeling that we will discover that what we call the “self” in ourselves will be found to reside outside of normal space/time and we will be combining these discussions with thoughts on Quantum Mechanics in the near future. In the meantime some of the tests we will be running with RICH in the next few weeks should be…entertaining at the very least.

Looking forward to hearing the groups thoughts on the Mirror test question

Vincent Gilbert

 

 
  [ # 23 ]

Thanks for your comments and thoughts, Vincent.

Brain maps explain a lot of what you’re talking about, and brain maps are the source of some strange anomalies when those maps get messed up for some reason. Here are three interesting examples…

(1) body ownership

The patient demanded to know whose left arm was lying in the hospital bed with him. “He would pick it up and throw it out of bed. The arm would come back and hit him in the chest,” recalls Dr. Kenneth Heilman, an American Academy of Neurology fellow. Here’s the kicker: It was his own arm.
...
He was experiencing the symptoms of hemispatial neglect, a neuropsychological condition that means the patient is unaware of anything on one side. It’s normally the result of damage to the brain’s right hemisphere, which results in that lack of awareness of anything left of center.
http://bodyodd.nbcnews.com/_news/2011/11/01/8383479-brain-damage-makes-some-blind-to-the-left?lite

(2) phantom limbs

After amputation of a limb, an amputee continues to have an awareness of it and to experience sensations from it. These phantom limb sensations are also present in children born without a limb, suggesting that perception of our limbs is ‘hard-wired’ into our brain and that sensations from the limbs become mapped onto these brain networks as we develop.
http://www.wellcome.ac.uk/en/pain/microsite/medicine2.html

(3) tickling yourself

Scientists may have unraveled a mystery which has puzzled them and millions of children for years — why is it impossible to tickle yourself?
...
The Daily Telegraph said today the secret lies in the cerebellum, a region at the back of the brain which predicts the sensory consequences of movements and sends signals to the rest of the brain instructing it to ignore the resulting sensation.
http://abcnews.go.com/Health/story?id=117038&page=1#.UHDyz1E-0qA

In summary, the brain keeps a map of its own body, so the brain knows about the body that houses its own brain. In my opinion, the mirror test merely adds a visual component to this somatosensory/proprioreceptive map. Therefore in the scenario you mentioned of waking up with a new face (or even body), I’m fairly sure within a few months somebody in that condition would have mostly adjusted to their new face/body, the same as they would if they injured a limb for a long period of time and needed to adjust to a different gait or to a different habit of reaching. I don’t believe there’s any need to appeal to metaphysics or quantum mechanics for an explanation of such things.

I’m also working on my own AI architecture that relates to self-awareness. In fact, if I’m not mistaken, self-awareness of its own processing activity arises immediately and naturally as a result of its architecture, without having to be specifically built into it. That’s one reason why self-awareness doesn’t particularly interest me, and that’s why I predicted we’re only one year away from self-aware machines: if nobody else designs a self-aware machine within the next year, I intend to! Unfortunately, I’m not at the stage where I can prove my claims, I don’t want to disclose the architecture before it’s published, at any point I could realize I’m completely wrong, and my planned demo won’t be very convincing of the self-awareness aspect since so far it’s aimed at demonstrating other abilities that I believe are more important.

We teach our children that humans have five senses: sight, hearing, touch, smell, and taste. We really have more. Vision is more like three senses—motion, color, and luminance (black-and-white contrast). Touch has pressure, temperature, pain, and vibration. We also have an entire system of sensors that tell us about our joint angles and bodily position. It is called the proprioreceptive system (proprio- has the same Latin root as proprietary and property). You couldn’t move without it. We also have the vestibular system in the inner ear, which gives us our sense of balance. Some of these senses are richer and more apparent to us than others, but they all enter our brain as streams of spatial patterns flowing through time on axons.
(“On Intelligence”, Jeff Hawkins with Sandra Blakeslee, 2004)

 

 

 
  [ # 24 ]

Excellent observation. I worked for a company some years ago that did a bit of research on giving an artificial limb a “kinesthetic” sense and I agree that this is part of it. My suspicion is that a person in the circumstances described would know immediately that they were looking at themselves even if the image that they saw was entirely different.  One of the questions that I hope to ask with RICH is “Do I ;

A. recognize my image and therefore know that I am the one looking.”
B. know that this is me and therefore know that the image I’m viewing must be what I look like.”

My belief is that we will discover that there are other factors that come into play, but certainly spatial awareness must be a part. Perhaps there is a Meta Spatial Awareness at work. I have heard Inertia described as a “body with mass having an innate sense of where it is in the Universe and resisting any effort to move it”. Also key I believe will be the sensing and recording of entropy as it relates to a Kinesthetic sense and Spatial awareness.

I am looking forward to seeing your findings published.

Vincent Gilbert

 

 
  [ # 25 ]

I do agree that you would recognise it was yourself in the mirror no matter what the reflection looked like. Common sense tells us that if we are in front of a mirror, the reflection is ours. This can be confirmed by doing actions such as opening and closing our mouths and watching the reflection do the same.

 

 
  [ # 26 ]

I think that, given the above described circumstance, I would come to realize that it was me in the mirror, but I think that our all too human dismissal/denial of uncomfortable truths that we don’t want to believe would seriously kick in in this situation, and I would probably deny the facts for as long as I possibly could, and quite possibly even try to destroy the mirror, as if it were it’s fault. smile

 

 
  [ # 27 ]
Dave Morton - Oct 8, 2012:

and quite possibly even try to destroy the mirror, as if it were it’s fault. smile

It is the mirror’s fault. After all, it’s giving you a reversed reflection of reality. That no good, lousy, lying instrument… grin

A couple more thoughts…

As Steve noted, it would be fairly easy to recognize oneself in the mirror, regardless of how one’s face or body had changed, since reaching out to touch one’s face would correspond perfectly to the motions in the mirror and to the sensation of touch at the corresponding places. Even if you had a new appendage, say an elephant’s trunk where your nose used to be, the situation would be almost the same, assuming your brain map included that new appendage for its sensory inputs. I think it would be strange to get used to having a body like an octopus, though.

All of this suggests an interesting possibility for the future: just as our neocortex grew as a layer on top of the older, more primitive brain, in the future we may be able to grow a new layer of skin over our existing skin, skin that would receive sensory inputs from sensors external to our body and that would feed to an extra layer of sensory input atop the existing cortex (a “neo-neocortex”?), so that we could be able to actually feel things going on in our environment, such as the depressions on the ground of approaching footsteps, slipping tectonic plates, the pain of others, damage to plants, damage to soil health, and so on. That’s one logical extension of our existing sensory system and would be a humanist’s and environmentalist’s dream.

 

 

 
  [ # 28 ]

All good points. If you were to attempt to test this the experiment would have to be designed so that it would nullify the effects of bio motor feedback, or optical data feedback.

VLG

 

 
  [ # 29 ]

Then, to me, it becomes more of whether you, me, or a robot is able to “Realize” rather than recognize. This is so especially with the previous “facial reconstruction surgery” scenario. Though you might not Recognize the face as the one you previously had for so many years, you will (aside from brain trauma), “Realize” that the moving, talking, emotive face is now yours.

The elephant with the chalked X on it’s forehead might only have REALIZED that there was a mark on it’s forehead that wasn’t Gray like the rest of it’s skin and not that it was looking at or recognizing itself, per se.

Just some thoughts.

Note:
If someone said / voted: < 150 years and it happened tomorrow, you’d still be correct!

 

 
  [ # 30 ]

Art, that was exactly my point, too. I’m glad that we can agree on this bit. smile

 

 < 1 2
2 of 2
 
  login or register to react