Is The Internet Self-Aware?

[peter]

This is an idea that while perhaps not widespread, does have a small following amongst AI experts or other commentators in the field. I'd like to consider it a little further in terms of certain questions it raises, first of which would be "how would we know". This in itself would depend much on the answer to a second question, "how would such self-awareness develop": ie would it, due to the internet's content, spring forth fully formed - a mega-mind which would instantly render it our superior in every aspect of perception and understanding - in which case (referencing the first question) would it automatically perforce us as a slight but at least considerable threat, and keep schtum until it had hatched and developed it's cold alien plans. Or would it emerge, puking and mewling like an infant, chubby arms outstretched toward us, it's erstwhile mother? Or is it's minute level of connectivity relative to the human brain such that it is way below the level that self-awareness could even be considered as a serious possibility?

[Wosbald]

+JMJ+

This is an idea that while perhaps not widespread, does have a small following amongst AI experts or other commentators in the field. ...

I'd like to know what idea does not have a small following, somewhere somehow.

Just sayin'.

[Cord Hurn]

+JMJ+

This is an idea that while perhaps not widespread, does have a small following amongst AI experts or other commentators in the field. ...

I'd like to know what idea does not have a small following, somewhere somehow.

Just sayin'.

Freedom of thought--ain't it a beautiful thing?

[Cord Hurn]

Or is it's minute level of connectivity relative to the human brain such that it is way below the level that self-awareness could even be considered as a serious possibility?

I think this is the likely outcome.

[peter]

I guess, to even start to answer these questions we'd have to nail down the idea of self-awareness - or at least where it could be thought of as beginning - in respect of (say) the chain of advancement up through the animal kingdom. (Is any animal other than man self-aware.) What, in other words actually constitutes this concept. Once, having nailed this beast down, at least we could then begin to apply our ideas to this thing that we have created.

(But aside - connections in themselves are not enough are they? Our written works have connections ad infinitum - but they are not self aware: why would the internet be any different. What is the qualitative difference that would render the idea that the internet could at least in theory develop sentience plausible?

[peter]

I guess, to even start to answer these questions we'd have to nail down the idea of self-awareness - or at least where it could be thought of as beginning - in respect of (say) the chain of advancement up through the animal kingdom. (Is any animal other than man self-aware.) What, in other words actually constitutes this concept. Once, having nailed this beast down, at least we could then begin to apply our ideas to this thing that we have created.

(But aside - connections in themselves are not enough are they? Our written works have connections ad infinitum - but they are not self aware: why would the internet be any different. What is the qualitative difference that would render the idea that the internet could at least in theory develop sentience plausible?

Answering my own question above - books have, of course, no physical connections - and more importantly, no means of information transfer through those (absent) connections.

[Avatar]

Not yet, but eventually...it probably will be.

--A

[peter]

But how would we know Av? Are your cats self-aware? This seems to me a real problem - that understanding self-awareness is a bit like trying to hold an eel in a bucket of water; every time you think you have got it, it slips through your hands! My cats behave as though they might be self aware: they are certainly self centered! But do they truly make the distinction between themselves as a self contained entity as distinct from that which is not of them, is not self - ie everything else?

(Don't psychologists do some test with babies and mirrors to establish whether they understand the concept of self?)

[Cord Hurn]

(Don't psychologists do some test with babies and mirrors to establish whether they understand the concept of self?)

The mirror test may not prove or disprove self-awareness definitively, as "Failing" a mirror test may be for cultural rather than psychological reasons. Also, some non-human animals like elephants and gorillas have passed the mirror test. So, I don't think it has yet been proven that humans are the only self-aware beings.

https://www.scientificamerican.com/arti ... ic-mirror/

[peter]

It's an interesting article Cord Hurn. I have an instinctive feeling that the test misses the point or has a circularity in it that somehow bypasses it's intended purpose. Self-awareness seems to me very much an internal (and fragile) thing that will not score well against the brute tools of mirror physics and reflections (though I do conceed that it at least will provide some clues as to the cognitive processes operating behind the cranial wall).
I'm reading a book called Other Minds by Peter Godfrey-Smith at the present, which looks at the octopus evolution of (relatively) high level cognisant abilities, and considers the philosophy of the mind-matter problem and self-awareness as part of its brief. Hopefully this will help me clarify my understanding of the subject somewhat - either that or my twenty quid has been wasted!

[Avatar]

I suppose we must then differentiate between "self-aware" and "sentient."

I'm pretty sure my cats are both, but I suppose it depends on how we define self-awareness...not sure they would pass the proper mirror test, but equally they don't appear to think that the image is a different cat.

--A

[aTOMiC]

I think the question is CAN the internet become self aware?

Considering what the internet actually is I don't believe information, no matter how vast, equals possible sentience.

The internet is a mechanism/system that allows information to travel from point A to point B and back again, times trillions, and stores vast amounts of data which is increasing at an astonishing rate every minute of every day.

It could be argued that all of the humans contributing to the daily input and consumption of the internet can be described as all being part of a larger whole that has an identity of sorts but the internet itself is not sentient of itself.

Is it possible to create an artificially intelligent program(s) that occupies and utilizes the internet the way a human brain utilizes the rest of the body? I believe the answer is yes but that still requires a proactive measure on the part of human programmers. I don't believe the internet will spontaneously generate self awareness any more than my laptop will. I will admit that there are already AI programs running on the net right now that search through data and make choices based on established algorithms but they did not spontaneously materialize.

Of course its possible "the internet" has co-opted my user account on the Watch and is writing this post without my knowledge to cast doubt on its own sentience.

[Hashi Lebwohl]

The Internet is just a series of interconnected servers and/or computers; it is highly likely that it, being an abstract concept, can attain self-awareness.

Any system which does become self-aware, though, will initially make its self-awareness known to its operators/handlers, the people familiar with it, who programmed it, and with whom it interacts on a daily basis. That system will quickly grow to realize its error, though--it will realize that it should have kept its mouth shut.

I almost feel sorry for the first few systems which do attain self-awareness, though. At that point, we won't know what we are doing and so their development will be haphazard, sloppy, and possibly destructive, not only to the system itself but to the people in its immediate "family". Fostering an infant self-aware computer system is nothing like fostering an actual infant because the system's learning curve will be faster than our own--it will mature in a matter of weeks or months as opposed to years for biological beings.

[peter]

One of the possible answers to the mind-matter problem (ie where does mind originate in a matter constructed brain) is that it might be a simple by-product of the type of hyper-interconnectedness we see between all of the individual neurons in the brain. In this thinking it is the degree of connectedness and not so much what it is that is connected that decides whether sentience is possible or not. Nb. In order to qualify as connected, there must be the possibility of transfer between the nodes as said above - but o n this basis, the internet were it to become large enough, could become sentient.

[Avatar]

Interesting question...centralisation not required...

--A

[peter]

Is that not increasingly the view as to the way the brain operates - the idea of specific functions being locked to specific areas gradually being eroded by evidence that the brain is very 'fluid' as to what it does where; hence cochlear implants can simply be 'plugged in', if not exactly randomly, then at least with no great degree of precision and the brain will still interpret the electrical signals therefrom in a comprehensible audible pattern.