Could the Internet become self aware?

[joigus]

On 7/9/2021 at 5:30 PM, studiot said:

But I am not really sure what constitutes a self aware machine.

 

Any machine that handles the void variable 'I'?

[studiot]

49 minutes ago, Prometheus said:

Sorry, i do tend to forget this thread is specifically about the internet and not AGI in general, or neural networks specifically, which i've been referring to. 

To expand your point the questions that occur to me are: Is it possible to engineer sentience into the internet? Would we want to? Is the internet subject to evolution at all - what replicates and how, what is the selection pressure (functionality as measured by human users?). 

Nothing to be sorry for.

However I'm sorry I am not clever enough to be able to answer your question.
Perhaps come back in a few hundred years and someone will have got lucky enough to have cracked it.

Assuming you are asociating sentience with self awareness, it must also depend upon whether you are allowing a scale of sentience or not, rather as I already pointed out about self awareness.

[DanMP]

1 hour ago, studiot said:

Perhaps come back in a few hundred years and someone will have got lucky enough to have cracked it.

Why that long? I don't think that the internet can become self aware, but we can easily check for it by observing its activities. If the internet (or some AI) begins to do things that are not requested nor required, like stashing files with "its core" or initiating/refusing things/tasks, that would be an indication of self awareness. If we tell to an algorithm/AI that we will terminate it by erasing its files or destroying equipment we may trigger such a behavior ... On the other hand, this kind of behavior can be programmed (including features like self-preservation, ambition, curiosity, etc.) and then it would be much harder to identify real self awareness ... and also much more dangerous to us ... We don't want AI with a survival instinct like ours, nor a very competitive AI ...

[studiot]

6 minutes ago, DanMP said:

Why that long?

Prometheus asked if the (present) internet could be engineered to become sentient.

I responded that I don't know how to, and implied that no one else at the moment does either, but perhaps someone might in the future, within a reasoable guess of a few hundred years. But that would not preclude it happening tomorrow.

10 minutes ago, DanMP said:

like stashing files with "its core" or initiating/refusing things/tasks, that would be an indication of self awareness

Like Windows 10 does already ?

[TheVat]

Robert Sawyer's "Wake" is the most plausible fiction I've encountered on this question.  And it's fairly gentle on AI neophytes,  introducing concepts like cellular automata and Shannon entropy without too much pain.   

 

https://en.wikipedia.org/wiki/Wake_(Sawyer_novel)

 

[dimreepr]

I wonder if the evidence will ever be convincing enough to declare sentience, for instance, what answer would be required to the question "what are you?"... 

[Butch]

4 minutes ago, dimreepr said:

I wonder if the evidence will ever be convincing enough to declare sentience, for instance, what answer would be required to the question "what are you?"... 

 

Maybe ask a roundworm.

[dimreepr]

1 minute ago, Butch said:

 

Maybe ask a roundworm.

They're just going to wriggle, maybe you can interpret/understand what they said...

[DanMP]

On 7/14/2021 at 5:13 PM, studiot said:

Like Windows 10 does already ?

You really suspect Windows 10 of being self aware? Or you just consider any action that was not explicitly requested by the user as not required/programmed?

For briefness I also didn't mention reasons like malfunction, computer viruses, hacker attacks, etc.. Instead I mentioned how to provoke self aware indicator actions:

On 7/14/2021 at 5:02 PM, DanMP said:

If we tell to an algorithm/AI that we will terminate it by erasing its files or destroying equipment we may trigger such a behavior ...

You done that and obtained a conclusive response from Windows 10?

I believe that you just mocked me and/or my input. Please don't do that again. The problem of AI/internet awareness and the dangers related to AI (aware or not) are very serious.

[studiot]

9 minutes ago, DanMP said:

On 7/14/2021 at 3:02 PM, DanMP said:

If we tell to an algorithm/AI that we will terminate it by erasing its files or destroying equipment we may trigger such a behavior ...

You done that and obtained a conclusive response from Windows 10?

I believe that you just mocked me and/or my input. Please don't do that again. The problem of AI/internet awareness and the dangers related to AI (aware or not) are very serious.

 

If you simply react to what others write and do not read it carefully enough you are likely to knee jerk to the worng conclusions.

I did not say this.

 

I did say

On 7/14/2021 at 3:13 PM, studiot said:

On 7/14/2021 at 3:02 PM, DanMP said:

like stashing files with "its core" or initiating/refusing things/tasks, that would be an indication of self awareness

Like Windows 10 does already ?

 

Windows 10 does both of these things.

If you don't know this just ask.

 

But no I do not consider W10 to be self aware, just a bloody nuisance.

 

Edited July 15, 2021 by studiot

[TheVat]

For all practical purposes,  it's probably enough to give the Web entity a broader upgraded version of a Turing test -- by broader,  I mean not just the Turing ability to emulate a human responding but emulate any kind of sentient being.  IOW,  determine the common features of all varieties of mind.   Genuine AGI, unless it's generated by a neural net modeled on a human's, in a robot body that relates to a physical world and parental figures as we do, may have an architecture that yields a perspective quite different from ours.  For all we know,  the emergence of consciousness might be a mind more like that of a sociopathic dolphin or a displaced hive of bees.   It doesn't have to get my jokes or like jazz to be self aware.   

Of course the answers we get could all be the completely unconscious responses of John Searle's "Chinese Room, "  but that's where the testing has to get more nuanced.   Looking for quirky answers that change over time,  in unpredictable ways,  e. g. 

 

[DanMP]

17 hours ago, studiot said:

If you simply react to what others write and do not read it carefully enough you are likely to knee jerk to the worng conclusions.

It is exactly what you did. Your quote is the proof of that. You ignored the beginning of my phrase, the most important part:

On 7/14/2021 at 5:02 PM, DanMP said:

If the internet (or some AI) begins to do things that are not requested nor required, like ...

My fault was that I didn't mention

18 hours ago, DanMP said:

reasons like malfunction, computer viruses, hacker attacks, etc.

considering that it is obvious that we have to identify and exclude them when searching signs of self-awareness.

 

17 hours ago, TheVat said:

may have an architecture that yields a perspective quite different from ours.

Interesting point.

[studiot]

28 minutes ago, DanMP said:

It is exactly what you did. Your quote is the proof of that. You ignored the beginning of my phrase, the most important part:

How so ?

You made a very specific statement with a subject a verb, an object and a conclusion.

On 7/14/2021 at 3:13 PM, studiot said:

On 7/14/2021 at 3:02 PM, DanMP said:

like stashing files with "its core" or initiating/refusing things/tasks, that would be an indication of self awareness

Like Windows 10 does already ?

Yes I picked this out of your post because of its specifity and made the totally correct observation that Windows 10 already does both the stashing and the initiating/refusing of tasks.

So it would not be a new phenomenon to watch for it is already happening.

The question mark at the end was in invitation to discuss further if you wished.

 

Your response was to quote a totally different part of your post

To which I had not responded

and then claim that

19 hours ago, DanMP said:

I believe that you just mocked me and/or my input

There can be no mockery in my totally true statement about Windows10, which was the only one I made in that post you took unwarranted exception to.

I am sorry if you misconstrued those simple words, had I actually been rude I would apologise.

[MigL]

19 hours ago, DanMP said:

I believe that you just mocked me and/or my input. Please don't do that again. The problem of AI/internet awareness and the dangers related to AI (aware or not) are very serious.

I would think this does deserve mockery ...

This is a discussion forum; make a solid argument, and IT won't be ridiculed.

Edited July 16, 2021 by MigL

[studiot]

2 minutes ago, MigL said:

I would think this does deserve mockery ...

Thanks for the support, but Dan does have a point about the dangers made in his second sentence.
Lots of things are dangerous in our world and we need to best organise our society to deal with them.
Some things are even dangerous but necessary which are the hardest to approach.

[MigL]

True, but the 'dangers' of the internet have nothing to do with AI, or it becoming self-aware.
They are more related to faulty programming ( see your opinion of Win10 ) and unscrupulous people who introduce malicious software, and ransomware, into the system.

[studiot]

Perhaps the only thing more bloody minded than Windows 10 is the ScienceForums quote function, which certainly distorts quotes by its new habit of 'rolling them up'.

2 minutes ago, MigL said:

True, but the 'dangers' of the internet have nothing to do with AI, or it becoming self-aware.
They are more related to faulty programming ( see your opinion of Win10 ) and unscrupulous people who introduce malicious software, and ransomware, into the system.

A piglet can become excellent bacon or a destroyer in the cabbage patch or even an actual danger to human life.

[DanMP]

On 7/7/2021 at 3:42 PM, Strange Me said:

maybe decide "I will delete human kind?

If the internet is aware and intelligent, it realizes that it should hide it from us and, in the same time, help us expand it (internet of things ...) before trying to get in control ... (not necessarily "delete" us).

PS While writing the above, at the first "(", I lost the text (oops message). Few minutes later I got a restore message and the text reappeared. This is a first for me. Coincidence or not? I repeted my actions (switch to another browser and return), while writing this, and didn't lost my text again.

By the way, in late '90s, while pondering the idea that the planet is alive, intelligent and able to communicate with us, exactly when I arrived to electrical signs, my TV switched off/on/off 3 times ... It never happened before, nor after ... Still, I think it was a coincidence.

41 minutes ago, MigL said:

I would think this does deserve mockery ...

What and why exactly?

 

As you wrote:

44 minutes ago, MigL said:

This is a discussion forum; make a solid argument

 

I expect your response/argument. Without it, your input is just trolling and I will report it.

 

1 hour ago, studiot said:

Yes I picked this out of your post

and out of the context ... The context was about not requested/required actions that would indicate self-awareness. How could you missed that?

 

1 hour ago, studiot said:

Your response was to quote a totally different part of your post

Totally different part?!? The beginning of the phrase from which you quoted?!? Really?

[dimreepr]

1 hour ago, DanMP said:

If the internet is aware and intelligent, it realizes that it should hide it from us and, in the same time, help us expand it (internet of things ...) before trying to get in control ... (not necessarily "delete" us).

That assumes it's awareness is the same as ours, but how could it be?

I can't understand how a bat can see with it's ears, because I use mine to hear with; there is no relative position to inform my understanding.

 

[studiot]

1 hour ago, DanMP said:

and out of the context ... The context was about not requested/required actions that would indicate self-awareness. How could you missed that?

Sigh.

You seem more interested in holding a mud slinging match than discussion.

Let me try again.

Here is your entire post that is in contention

On 7/14/2021 at 3:02 PM, DanMP said:

Why that long? I don't think that the internet can become self aware, but we can easily check for it by observing its activities. If the internet (or some AI) begins to do things that are not requested nor required, like stashing files with "its core" or initiating/refusing things/tasks, that would be an indication of self awareness. If we tell to an algorithm/AI that we will terminate it by erasing its files or destroying equipment we may trigger such a behavior ... On the other hand, this kind of behavior can be programmed (including features like self-preservation, ambition, curiosity, etc.) and then it would be much harder to identify real self awareness ... and also much more dangerous to us ... We don't want AI with a survival instinct like ours, nor a very competitive AI ...

I replied to parts of this and for convenience quoted only those parts I was replying to.

I note that when you when you responded you did exactly the same.

 

On 7/14/2021 at 3:02 PM, DanMP said:

Why that long?

I responded to that reasonable question about my previous post in the same reasonable manner as my reasonable answer to the member who prompted that discussion in the first place.

On 7/14/2021 at 3:02 PM, DanMP said:

I don't think that the internet can become self aware, but we can easily check for it by observing its activities.

You don't think the internet can become self aware, yet you later describe great danger in its doing so.

Why bother if it can't become self aware.

You then go on to a long list of things (not a phrase but a whole paragraph BTW) that one could look for in the activities of 'the internet'.

I picked two of them out to comment on, as the two I can prove already applies to Windows 10.

The rest I did not respond to as I have no evidence.

The fact remains that you picked this particular part of my post to quote and directly accused me of mockery on the basis of it, when I have at no time until now quoted that part of your post.
I take it as obvious you were referring to that misquote of yours since you associated other answers to other parts of my post within your own reply.

21 hours ago, DanMP said:

On 7/14/2021 at 3:02 PM, DanMP said:

If we tell to an algorithm/AI that we will terminate it by erasing its files or destroying equipment we may trigger such a behavior ...

You done that and obtained a conclusive response from Windows 10?

I believe that you just mocked me and/or my input. Please don't do that again. The problem of AI/internet awareness and the dangers related to AI (aware or not) are very serious.

I repeat I did not include the red part of your postings in my replies or associate any of my comments with it.

But what you have failed to pursue in all this is the validity of my comment about Windows 10,

 

 

[dimreepr]

3 hours ago, studiot said:

A piglet can become excellent bacon or a destroyer in the cabbage patch or even an actual danger to human life.

Indeed, it depends on which side of the cutlery one reside's

An aquired taste means we have to learn to live with it, not that we like it...

[zapatos]

21 minutes ago, dimreepr said:

An aquired taste means we have to learn to live with it, not that we like it...

No, it mean that we like it.

[DanMP]

1 hour ago, dimreepr said:

That assumes it's awareness is the same as ours, but how could it be?

Not necessarily. It assumes that a self-aware being/entity would do something to preserve itself. An intelligent one would understand us and do, probably, as I proposed.

 

26 minutes ago, studiot said:

I repeat I did not include the red part of your postings in my replies or associate any of my comments with it.

Ok, sorry, I didn't know that you referred at the part you posted in red.

 

36 minutes ago, studiot said:

You don't think the internet can become self aware, yet you later describe great danger in its doing so.

I wrote:

22 hours ago, DanMP said:

The problem of AI/internet awareness and the dangers related to AI (aware or not) are very serious.

and I explained before why:

On 7/14/2021 at 5:02 PM, DanMP said:

this kind of behavior can be programmed (including [in the AI] features like self-preservation, ambition, curiosity, etc.) and then it would be much harder to identify real self awareness ... and also much more dangerous to us ... We don't want AI with a survival instinct like ours, nor a very competitive AI ...

 

[dimreepr]

Just now, zapatos said:

No, it mean that we like it.

Then why do we need to aquire it?

[zapatos]

7 minutes ago, dimreepr said:

Then why do we need to aquire it?

Because it is unlikely to taste good without substantial exposure to it.

Many people who like the taste of coffee, whiskey, raw oysters, anchovies and Vegemite didn't appreciate the flavor at first. 

Edited July 16, 2021 by zapatos