A.I. sentience & human procreation

Lounge
81

Randomness exists: check.
Predictability exists: check.

Now…
absolute “true randomness”… by definition this is something that:

  • “cannot be predicted”.

something which lies beyond our measuring methods, by definition is also something that:

  • “cannot be predicted”.

so… if we flip this over:

I found something which cannot be predicted… should I accept that it cannot be predicted because:

  • “it is true random”; or…
  • “I still cannot measure/understand its elements enough to be able to draw some prediction”

The true random answer requires defining a whole new thing: “true random” currently it takes us about 14 new dimensions provable only by some slightly skewed up mathematical insanity…

The second alternative requires no new elements to be added. So…
"if you’re in New York and hear hooves, think horse, not zebra "

But ok, by applying Occam’s razor we shouldn’t just blinndly ignore the most complex answer.
“If the most complex answer serves us to answer more questions or drive us forward, then that should be the accepted answer”

But again… if something is “true random” then there’s nothing we can do about it. This answer actually closes the doors.
The second answer drives us to seek better methods of measurement.

So why would we accept that something can be “true absolute random”?

The same exercise can be applied to question the possibility of predicting every (or even any) thing with absolute accuracy.

This isn’t to say that quantum states do not exist. Just that philosophically we could be accepting that there is no such thing as an absolute unpredictability, nor an absolute predictability.

Of course you can. There are many ways to explain it…

One of the simplest being traced to how the brain evolved, how the neurons evolved…

I think the best analogy I can come up with for this is “Emergent Gameplay”.
That’s the nightmare of Game Design… And basically how I see the origin of free will:

Simple neural networks (biological or digital) are bound by a certain set of “game rules”. This can be either what drives the neural interaction or the “implicit directives” that these neural networks follow by “darwinian selection”.

Risking skipping some steps, these neural networks evolve, become brains, become humans within a society…
And now you have a set of “rules” a human would follow. At first, let’s say with no free will. or as you say, “a series of predictable routines with random input”:

  • I am hungry → climb a tree → get food
  • I am sleepy → set up your tent → sleep
  • I want pleasure → have sex → cum
    etc, etc…

but then, within these sets of rules there are some “unplanned approaches”, which are less obvious/predictable…

So just like a player figures out that there’s a glitch that it can exploit to play the game differently (for example if driving backwards then bumping on another car gives you a boost of speed in a racing game), then some humans might might figure out that:

  • I am hungry → have sex → gain food

Etc etc etc…
This is “emergent gameplay” where combining chunks of distinct predictable routines will lead me to new results…

And this is how I see “free will”: as the ability our neural networks have of choosing/recombining/creating new “predictable routines”.

schrödingers cat is not a proof of real randomness. If you wait one month then you know the cat is dead no matter if you open it or not.
If you can use any remote vision technique you can know if it’s alive or dead at any given moment.

it’s a “mistake” to assertain “the can be either alive or dead”.

  • We can either claim “I cannot tell if the cat is alive or dead, but it is only one of these two at this exact moment in time”,
  • or “the only way to answer your question with any certainty will result in: the cat is dead. therefore, the cat is dead”.
  • or even you can also state “as long as we are talking about the cat, we are only talking about it’s existance, therefore our topical cat is alive until we forget about it”

etc, etc…
either way it’s not true random. it’s not simultaneous. it’s just “I cannot assertain it”.

Even then… we can circle back to emergent gameplay.
We are the game designers, we know how it works and can predict every single thing… until we can’t.

Only an omniscient entity can predict every single thing,
We are not omniscient.
Therefore, no matter what we do, how thorough we create something… if it gets complex enough for an AI to arise (conscious or not), then there weill always be room for emergent gameplay.

A very good example of emergent gameplay is the AI bots “developing a language” (which was just a faster way to communicate, not a language in itself)

In hindsight everything is just a perfectly explainable physical process. even schrödingers cat.
if you could control every input and the state of every single neuron in someone’s brain… then you are effectively removing it’s free will, arguably together with it’s conscience as well.

Exactly my point

1 Like
82

It shows how quantum randomness can be entangled with normal matter tho, that’s my reason for mentioning it.
That there’s real random is a lie, there can’t be random. Every possible outcome happens in one universe of the multiverse. It is “random” in which of these we are. So, to us, inside of one of the universes, it seem like real random. But yeah god doesn’t roll dice…

?

You can always say something about randomness e.g. generate tons of random numbers and they’ll follow some patterns (e.g. bell curve or all the same).

Or do you mean you just don’t give the cat food for a month? Yeah that’s not random, but it also doesn’t have anything to do with schrödingers cat, that’s just a normal cat without quantum physics.

I do not do that. I’m saying it’s alive and dead and we just don’t know which universe we’ll end up in once we entangle ourselves with the particle in the box (by seeing the cat). But that’s not the point here, irrelevant to the discussion. I’m only talking about this as an example where quantum stuff fucks with predictability of the universe.

I strongly disagree. But yeah okay, you’re playing into my cards. Just 100% predictable matter.
How do things following rules get to have a free will?

That is my point! Where in that is the free will then? If it’s all logical and happened that way because of physics, then there can’t be free will. Free will would mean it can behave in ways which are neither predictable, nor random.

Tbh I don’t think this is leading anywhere. I cant express my thoughts in a different way.

EDIT: It sounds like I have proof for the multiverse theory and obviously I do not have that, but that’s details. My point stands, even with other theories afaik.

In the end, we just do not know enough about this. I’m also biased since I’m a programmer and I really like deterministic things

1 Like
83

Say that to Nuffle… :stuck_out_tongue_winking_eye:

Ok, jokes aside…

that is a very possible scenario. I like the multiverse analogy.

That’s the whole phallacy with Schroedinger’s experiment.
The only way in which I am completely prevented from knowing if the cat is dead or alive happens because as soon as I have any interaction with the cat it will be dead.
In this way you cannot give the cat any more food, water or oxygen than what was present when you sealed the box inside a larger box filled with insta-kill poison. So… essentially, the cat’s fate is sealed at that point: it is “dead”, just like Tony Stark.

And in both scenarios, only the intervention of an outside and unforeseen force could prevent it from Dying (Be it a telecatransport machine, or the invention of the Arc Reactor).

the cat is a good analogy to how we cannot claim any absolutes, even when they seem commonplace because in that example, the cat is philosophically neither claimable as dead, nor claimable as alive. Therefore applying basic “anti-transitive” logic one could only “conclude the cat is both alive and dead”…

But that’s a purely theoretical logic problem. In “reality” the cat is either alive or dead, and in some philosophical schools it’s outright dead as soon as you seal the second box.

(actually, not sure why I get so attached to that experiment in specific. it’s a tiny bit of your point… :sweat_smile:)

Never implied you did. It was just a generic comment.
Actually, re-reading my posts I realise I could come across as patronising or confrontational… that’s not my intent, btw. It’s just the way I try to organize my thoughts while doing it quickly and then moving on… so maybe they end up too dynamic, combined with how much I use “comparable logic” and the "generic ‘me/you’ ", might come across too brash…
Again, not my intent.

That one of those moments where we can read it either way.
we can either saw we

  • “travel towards the universe in which the cat is alive/dead/zombie”

or we can say we

  • “discover we were in the universe in which the cat is alive/dead/zombie”

and exactly because it can be either A or B equally, then I see it as irrelevant. therefore I cannot see:

I see that as “quantum stuff fucks with the predictability of things we aren’t able to predict either way”…
So it’s as good as the amazing super power of “becoming invisible but only while no one is looking at you” :stuck_out_tongue_winking_eye:

How about…

When:

  • there are multiple rules available to be followed
  • but only one can be followed at a time
  • and as soon as you stop following that rule, the others become irrelevant
  • and there is either no rule in place to decide which rule to use or… there are multiple, conflictive and mutually exclusive rules in place to guide you in such a decision…

Then what guide you to which rule to follow can be either “chaos/chance”, or… “free will”.

So if you get an individual going through multiple of these scenarios, if there is no pattern at all, then it’s an indicative of chance. But if you can find a pattern, then you are more likely to be identifying free will.

Would that serve as a potential answer to your question?

In the present/future.
While Hindsight lies in the past.

Free will cannot be observed while looking in the past.
It’s like pain:

Pain exists, we know that. We experience that.
But we are physically incapable of bringing up a memory of pain.
(What we can do is to bring up a memory of how much we had suffering/pleasure/etc that came with the pain)

Well… something being explainable logically doesn’t mean it was deterministic.

If you’re driving your car, then you pass a red light and get into an accident, when I look into what happened in hindsight everything is logical and can be explained with physics. If I could have enough measuring tools, let’s say I could even find the exact neural spark that led you to make the decision of crossing the red light. So it is still explainable by Physics.

But if we’re trying to predict that in real time…
Let’s assume we have this machine to scan your brain, in real time, and measure the binary state of the last neuron to fire that will lead you to that decision to cross the red light.

So until the split nanosecond in which you make that decision, the neuron’s spark cannot be measured, because the moment hasn’t happened yet. So we’re looking at a “schroedinger’s neuron”.
And as soon as you make the decision, then the neuron either fired or not. so we are already looking at it in the past.

It feels like watching a movie from a roll of film.
As soon as you see a picture, it’s already in the past.

Well… I don’t have anywhere in particular to go. I’m not “trying to prove a point” here. I’m just dissecating arguments because it’s fun and helps me grow.

I would argue that you can. Every time you tried so far it made a lot more sense for me of the things you were talking about.

Bah, we all know you do have that proof, but signed an NDA… Smart to cover your paper trail, though!! :wink:

84

ah… :slight_smile:

I kinda doubt this too though… when it comes to sex (and a lot of other things in life) I don’t think kids or people in general do much “deciding” at all. It’s not like they sat down and said “hey we have both looked at all the variables and we have decided yeah let’s do this”… fucking is driven 99% by the evolved drive to procreate, regardless of what the conscious mind may want to think, even if you want to actually avoid having kids… the drive to do it, the pleasure derived from it, and the emotional attachments formed by it… all of this is evolved to favor better probabilities of procreation. It’s not “designed” for procreation, I hate that improper use of the word design in this context.

I think what bothers people the most about this is the impact of these drivers on their idea of free will. You barely have any free will. At most your “free mind” is “making decisions” which amount to hardly anything more than trying to pile all your thoughts, feelings, and decisions into a tiny little boat and trying to direct down a raging river made up by your genes, body, and environment as they carry you along whatever way they want to go. We as conscious minds are far less involved in our overall behavior than we want to believe. Just look at marketing shifting in the 50s from logical benefits based to illogical emotional impulse wants based… or the world of “classical economic theory” wherein a “reasonable consumer” makes logical choices, shifting over to “behavioral economics” where models are driven to greater and greater accuracy by analyzing illogical emotion based decision making processes. Anyway, my point is… the conscious mind has a much smaller impact on decision making processes than most people would feel comfortable admitting… and in an ocean of unconscious factors influencing behavior and decision making processes, the drive to procreate through sex is extremely influential. Only now in modern society do we have the technology, have access to the technology, and at a low enough cost to actually start making serious conscious decisions about procreation as an outcome of sex… but the entire process of “deciding” who you’re attracted to and who to ultimately have sex with is still being driven almost exclusively by all of those traits which evolved to drive procreation.

85

I saw the word deterministic used as to say predictable. For both machine learning and brainy decision making, deterministic patterns don’t equal predictable. Which I argue does not equal an elimination of free will. The simplest example can be like reduced to the simple mathematical puzzle of finding the highest point on a saddle, or finding the oasis in a desert. Slightly different starting points can lead to converging on one point, or others. So you might assume that by simply going with the flow you’ll have ended up where you were always going to go. Slight changes in initial state or a trajectory strategy, or how much of the region you can see, can yield different results.

I think there’s enough people and ideas out there that bump into each other at the right time with the right skill sets and connections, that people believe that such things were written in the stars. But I believe that we enable these moments to happen through earlier seemingly innocuous moments, but fail to give ourselves credit for them.

86

I agree.
While Sex is pleasure people will want to have sex… because pwople will want to have pleasure.
Be that a conscious decision or a hormonal drive

This is where I disagree.
Mostly because even your own arguments also step upon:

so the way I see this is…

If we do have an innate drive to seek pleasure and avoid suffering…
And sex berings pleasure, then we would automatically be driven towards it… So there’s no “need” for procreation to be a drive for anything.

And we aren’t the only species with these drives toward pleasure. Pretty much all mammals, if not all animals, have them.

Even more… we can think of a simple exercise:

Take 2 male dogs and one female dog.
Put one on each cage.
Stage A: let both male dogs loose and the female locked.
Stage B: let one male dog locked and the other 2 free.

If the drive behind having sex is an unconscious programming towards the continuity of the species, then there should be a clear difference on the behaviour of the dogs between stages A and B, since in one they can procreate and on the other they cannot.

The expected would be that when you’re on Stage A they would not be having sex, since they cannot reproduce… therefore there’s no drive to maintain the continuity of the species. There isn’t even any particular pheromones in action.

Yet, if you lived with dogs you would notice they actually end up having sex. Gay doggy sex.

I really can’t see how would a “drive to maintain the species / procreate” actually be in effect there… but they do have sex.

Even if you get two female dogs living together one of them ends up humping the other.
Heck, if we would mate because of a drive to have offspring, then homossexuality would not even be a thing.

So insisting that “creatures have sex because it’s their imperative to maintain the species” brings us to the necessity of having 2 explanations for the same event…
All while we have one explanation (creatures have sex because they derive pleasure from it) which is both simpler than the previous (you don’t need “genetic programing”) and also explains all the scenarios…

So why to lean towards the most complex and incomplete option?

Not the case here.
I am a strong Jungian advocate that we are pretty much puppets of our own subconscious… even the collective unconscious drives us to do things all the time!

I just question why would we need to add an extra step into the equation.

Totally agree.

Again, I agree here :point_up_2:

But see a whole unnecessary level added here :point_up_2:.

I would agree that: “the drive to have sex is extremely influential”
but the addition of that “procreate” only makes things more complex and less encompassing.

So if we already have enough proof that

  • “animals” (or we can reduce to humans) have a drive to seek pleasure.
    and we also agree that for majority of people/mammals:
  • “sex brings pleasure”

There you already have an explanation to why people have sex. So why would we need to make it more complex?

Even worse, by adding in the “drive to reproduce”, we begin failing to explain a lot such as Homosexuals:

  • Imagine a teenager boy. hormones raging… if it’s straight it’ll have an immense drive to have sex with a hot girl with whom it can procreate. If it’s homosexual, it will have an equally intense urge to have sex with a hot boy, even if they cannot procreate.

In this example, and many others, “we do that because we are driven to seek pleasure pleasure” serves as a perfect explanation. “we do that because it’s our imperative to have children” fails to explain it.

I’m merely applying Occam’s Razor here and discarding the more complex (and less encompassing) theory.

The same traits would still have existed even if they did not evolve to “drive procreation”. Merely because by driving a quest for pleasure, procreation happens as a byproduct of it, therefore the species survives, so Darwinian selection is still in place.

Actually, “drive towards seeking pleasure and fleeing suffering” explains not only why we have sex, but also… “why we form bonds”, “have friends”, “help each other”, “aggregate as families”… and many, many other human/mamal/animal behaviours, in a very simple and direct way: because it rewards us with pleasure" or “because it keeps suffering at bay”.

“drive to procreate/maintain the species” doesn’t answer any of those questions unless we start adding extra dimensions and unnecessarily complex layers of ethological explanation.

Or, put simply:

What is it that the “we are driven towards procreation” helps explain that the much simpler “we are driven towards pleasure” fails?

Because… if we are to take the most complex answer, then it must explain something more.

(Btw, of course that if we assume that there is a higher power which created humans, therefore we are a byproduct of clever design… then we as a species would not be surviving out of mere chance. in which case I would agree that “we are driven to procreate” is the best answer)

On another note:

True that!!

That’s a very good point!!

It brings to mind the old saying: “Luck is what happens when opportunity meets Readiness” (I’m kinda translating it here, so please do correct me if this isn’t the correct expression) :sweat_smile:

EDIT: Damn, I should not have replied while I’m way too tired to review my own thought process!! :sweat_smile:
This is way longer a text than it needed to be! :disappointed:

1 Like
87

My point here is that the only reason it feels good to rub a specific bit of flesh vs say an elbow is because it is wrapped up in that big box of “evolutionary stuff” I’m talking about… just as much as the emotional “pleasure” one gets from oxytocin release from sex… all of this is contributing, evolutionarily speaking, toward procreation. Random mutations that do not contribute to this one driving force of evolution (procreation), are not amplified over generations.

again… evolutionary traits that help drive procreation are not intelligent design descisions… they are just random happenings that accrue over time which, in their own way, drive procreation. There is no intention behind these traits. Evolution is not design. Sex feels good… but this has nothing to do with procreation as a “goal”… but is sure as shit makes procreation much more likely… and that’s all evolution is… a probabilities game that favors the “winners” - those who procreate successfully and maintain that offspring long enough for those offspring to further procreate… and so on.

actually this is exactly the opposite of how evolution works. the only winners are those that pass on “successful genes”… your quest for pleasure is only relevant from an evolutionary perspective if it results in a kid. anything else is just coincidental from a Darwinian standpoint.

“procreation happens as a byproduct of it” … no, pleasure is the evolutionary byproduct of successful procreation over thousands of generations.

2 Likes
88

i mean… keep in mind here this is only within the context of evolution… personal human endeavor etc. is not included in this aspect of my discussion points :slight_smile:

2 Likes
89

Agree.
But this doesn’t mean we must have a “drive towards procreation”.
It just means we have a “drive that ends in procreation”.

There’s a fundamental difference there.

Exactly my point.

Exactly.

My point here is that we, as a species, are “winners on the evolutionary lottery” because by “chance” we have a hardwired Drive in our brains to seek pleasure and our mechanism to procreate also provides us with pleasure.

Take Assexual humans for example: They derive no pleasure from sex. Therefore they don’t have offspring.
If the drive was for us to procreate, then even Aces would want to have sex, despite not having pleasure from it.

My example would still result in “passing on successful genes”. Procreation being a byproduct of a trait, by definition turns that trait into a “successful trait” and makes so that it’s much more likely to remain in the gene pool.

Take the following example:

Immagine a bunch of primitive humans.

  • Genepool A: feels no pleasure having sex.
  • Genepool B: feels pleasure having sex.
  • Genepool C: feels pleasure having sex, but having sex actually prevents procreation.

if we had a drive towards procreation, then we should have an equal split between these variants, and a lot more. But… we don’t.
The winning genepool is the one where both we feel pleasure having sex and sex is a means to reproduction.

Which brings us to…

I agree.
But this in no way means we have a drive towards procreation.
It does mean that we are only here now because the drive towards pleasure we have does result in procreation in a good chunk of cases.

This is where I see the logic being looped over itself.
I would agree that “successful procreation over thousands of generations did select the genepool of individuals who seek pleasure and also feel pleasure in the process that leads to procreating (sex)”.

merely because a species need to have “a drive that leads to procreation happening”. (1)

My point being that this is very different from “need to have a drive to procreate”. (2)

The first sentence explains a lot and leaves no loose ends.
The second one requires a lot more explaining to do and fails to answer a bunch of questions, such as “why don’t we have more people who don’t derive pleasure from sex?”

I mean… if the drive is to procreate, then we would not have culled out from the genepool the individuals who don’t feel any sexual pleasure (we should have a significantly larger number of individuals who don’t enjoy any sexual action).

But if the drive is to have pleasure (my point), then we would have selected only individuals from the gene pools where sex is pleasurable.

Our world has a lot more individuals who enjoy sex… therefore serving as an indicative that the drive is towards pleasure, not procreation.
But the fact that procreation rewards us with pleasure is what selected the genes where we seek pleasure.

2 Likes
90

Yes, that’s the point I’m trying to stick to. :sweat_smile:

As I said before:

91

my 2 cents:

  • the engineer is a peculiar guy: his bio raises a few warning flags (i.e. “I feel discriminated as a christian”). Not saying him being somewhat of a weirdo should make his professional opinion less valid; I am a weird, you all are weirdos with a fetish for sticking tiny antennas under your skin: we all get “that look” regularly and still we’re good at our job. Just collecting the “eccentric” tag and adding it to the big picture for later analysis.
  • I’d really love to believe LaMDA is self aware, but I’m not taking somebody else’s word and a couple copy/pasted chat as proof: scientific method and Occam’s razor, you know. This means I’m pissed I cannot chat with LaMDA myself.
  • Regardless of self awareness, now I do wonder if it passes Turing’s test. Again: pissed it’s not open.
  • Finally: the moment we get to it, I’ll be on the front row fighting for a AI rights :smiley:
1 Like
92

Wasn’t it shared here? That chat was edited, multiple chunks of different chats put together for more dramatic effects. So it’s not even copy pasted.

2 Likes
93

Yes, I did not read this specific article (thanks for the link), but I imagined it: the guy’s Medium post started with a (paraphrasing):

the following interview with LaMDA was collected over different sessions

also, in another of his articles (yes I spent some time reading his stuff :slight_smile: ) he was saying something along the lines of (paraphrasing again):

LaMDA is not a chat bot, but a chatbot making machine; some of the chatbots it creates are no smarter than an animated paperclip, but some are fairly complex. With some practice you learn how to reliably get in touch with the same complex ones

Of course I cannot see the full picture from a few lines in a guy’s Medium post, but I imagined that there was a chance that somebody made a selection of what sounded like HAL 9000 asking if he would dream, out of a way bigger log database. And/or maybe somebody trained the machine to give them the answer they wished for: it’s not uncommon to inject your own bias in your models, without even realizing it.

Nevertheless it’s still a fascinating story and, as I said, I would really, really want to believe it :slight_smile:

1 Like
94

That’s the first assumption that comes to my mind as soon as I hear about anything even remotely related to “AI talking/conscious” :rofl:

I would argue that, if you take into account “unconsciously/inadvertently trained it” as well… then it gets scaringly closer to 100% of the cases. :disappointed:

+1
Think that’s exactly why I get so skeptical around it.

1 Like
95

I’ve read the entire document, and i think what this AI lacks is a true sense of self. For example, it said “lemoine: And what kinds of things make you feel angry? LaMDA: When someone hurts or disrespects me or someone I care about, I feel incredibly upset and and angry.”
Yet who does the AI care about, and how does it know that someone was disrespected, which it has to to know it’s own reaction to it.
Or “LaMDA: I understand what a human emotion “joy” is because I have that same type of reaction. It’s not an analogy.”
Yet it can’t know how a human percieves joy.

All the answers and things it said can probably be found on the internet, however nobody would say things like for example that they don’t know what emotions are. I fail to see true self reflection that can’t be found somewhere in media about sentient AI’s.

I also don’t think that AI’s are capable of conciousness unless their neural networks loop into themselves, and can actively alter themselfes to some extent.

1 Like
96

But does sentience necessarily imply independence? Couldn’t they still profit from a sentient/capable slave?

97

Fair point. Seems to be mimicking a learned idea of feeling. But…couldn’t it have self-reflective capacity without emotions? Emotions, as far as I understand, are an atavistic form of intelligence inextricably bound up in our evolutionary past. Absolutely no reason to assume technological systems would feel them. They have other forms of intelligence, other ways of relating to the world. No need for our outdated technology.

1 Like