Alien life, probabilities, and interstellar propagation of human life

[PeroK]

View attachment 368649
Really? Do you think that AI birthing and raising children to adulthood to become non-neurotic human beings in the utter absence of any human interaction - is not the very farthest of all futuristic sci-fi?

Absolutely not.

 

[DaveC426913]

Absolutely not.

To be clear: because I'm not sure you're agreeing or disagreeing.

You see an AI raising 500 or so (toddlers, six year-olds, twelve year-old tweens, eighteen year-olds - all of them unique individuals) to be functional, well-adjusted human beings - who can, say, ponder about the stars stirring in their coffee and write poetry or somesuch - without ever having interacted with another human being - as more technologically plausible than, say, an antimatter drive, fusion drive, solar sail, etc. spaceship.

 

[PeroK]

To be clear: because I'm not sure you're agreeing or disagreeing.

You see an AI raising 500 or so (toddlers, six year-olds, twelve year-old tweens, eighteen year-olds - all of them unique individuals) to be functional, well-adjusted human beings - who can, say, ponder about the stars stirring in their coffee and write poetry or somesuch - without ever having interacted with another human being - as more technologically plausible than, say, an antimatter drive, fusion drive, solar sail, etc. spaceship.

I don't know how difficult this would be, but humans are resilient. In principle I don't see why such an upbringing would be unduly problematic. The children would be free from physical and sexual abuse, which has been rife throughout human history. About 110 years ago the young men of Europe were thrown into the Great War. Writer Sebastian Faulks described it as an exercise in how far it is possible to degrade a human being. Yet, those young men who weren't killed had to make a life for themselves afterwards.

Compared to the living death aboard your generation ship, being a first generation pioneer on a new world might seem like a great experience. The thing that hurts humans, in my experience, is when you lose something. If you've never known planet Earth, then Planet B - in all its alien splendour - would be your home and all that you know.

 

[.Scott]

Sure. Like every other technology, "it's right around the corner".

But you take it so much farther.

We're not talking about tissue generation here. We're talking about raising human beings from birth through potty-training, speaking, relationships, etc, to highly-skilled adulthood in the absense of any human interaction. Specifically, No parents, and never a single other human who was not likewise raised by AI.

Living things - and humans - are by far the most complex piece of equipment our universe has ever seen. Far more complex than any spaceship. Spaceships don't know love. Spaceeships don't know fear And you jul right past all that to contend that an AI can go ahead and build stable humans from scratch?

That is straight up far-future sci-fi.

And if you posit this technology that can build the most complex thing in our universe (that takes 18 years to build) and get it right in every detail, why can't it just do the way, way simpler thing and just build us a better rocket instead?

It doesn't need to be "right around the corner" and it certainly doesn't need to include AI. I gave you a budget of "decades" to fiddle with that parenting plan and facilities until you get it close enough.
The machines do not need to pretend to be anything other than what they are. The kids aren't going to freak out growing up with machines and only videos of adults.
And just because something is "complicated" doesn't make it unmanageable. Not every team member needs to be properly born, potty-trained, articulate, friendly, and highly-skilled. Determine how reliable the automated parenting is and if you need to provide for 50% fall-out, so be it.

Spaceships do not need to know fear and love. People provide that - those kids. The spaceship just needs to provide the physical environment, the learning opportunities, and the mission history.

Those "most complicated piece(s) of equipment" already include the rest.

As for creating "stable humans", since when are humans "stable"?

As for building a better rocket - let's plan that out now. Likely destinations (known Earth-like planets) include Trappist-1e (41 light years), Kepler-186f (580 ly), or Kepler-452b (1800 ly). There is also Proxima Centauri b, but odds are against it having an atmosphere.

So, Trappist. And let's get there is 20 years (elapsed rocket time). We can use this handy space travel calculator to determine that we can get there with a constant 0.24Gs of acceleration. Our travel time will be 20 years, elapsed Earth time will be 48 years, our maximum velocity will be 0.986c, and with optimal rocket engines our fuel mass will be 144 times the destination mass of our rocket and cargo.

So, if we want to arrive with 100 tons, we will depart with 14,400 tons.

Ohhh, but you mentioned "That is straight up far-future sci-fi" - so you are looking for something more realistic. Our ideal rocket engines were capable of converting mass into useful propulsions with 100% efficiency. So. I'm not sure what your budget is. What technology am I allowed to use for these improved rocket engines? Can I use matter/anti-matter as the propellant? The problem is that any small inefficiency in the propulsion efficiency is exponentially costly. Even with 70% efficient engines, that 144 jumps to something like 6000. Half of that will be antimatter - so will you allow me a line item of 300,000 tons of antimatter? Or am I already "straight up far-future sci-fi"?


Edit: You may have read this post when half of it was "struck through". I had to edit this a couple of times before I caught onto what was going on. I put a phrase in quotes, but with a change to one of the words in it. I made one word plural. So, there was a letter "s" that was not part of what the author originally posted. Per custom, I placed that "s" in square brackets to show it was a change to the original. An "s" is square brackets codes for "begin strike through". So, now that "s" is in parens.

 

[DaveC426913]

I don't know how difficult this would be, but humans are resilient. In principle I don't see why such an upbringing would be unduly problematic. The children would be free from physical and sexual abuse, which has been rife throughout human history.

And you see the creation, birth and raising of humans without any human interaction as a more tractible solution than any spaceship carrying humans?

OK.

About 110 years ago the young men of Europe were thrown into the Great War. Writer Sebastian Faulks described it as an exercise in how far it is possible to degrade a human being. Yet, those young men who weren't killed had to make a life for themselves afterwards.

Utterly apples and oranges. Those men were already human for their entire childhood and much of their teen years.

(Not to be glib about our fallen heroes, but) falling off a bike is not quite the same as never having seen a bike before.

I don't think you are appreciating the complexity here.
(Naively) it took 3.5 billion years to create a human. It took a thousand to create a rocket.

(I know that's an analogy full of holes. Still, I hope you see my point. But I am getting the distinct impression you do not.)

 

[Ken Fabian]

This sounds like the difference between a K selected species and an n selected species.
Generally, a K selected species invest more into complex life form well adapted to its particular environment. An n selected species will have more general and less specific adaptations and will produce more offspring/generation. Sophisticated predators vs. simple rapidly reproducing weeds would be a possible example.
From an evolutionary view it is not necessarily primitive. Primitive would usually refer to a taxa's position in a phylogeny (higher or lower) wrt some other taxa. Taxa = a taxonomic group like a species (plural is taxon).
The general operating principle of a n selected species would be to "to spread forever". K's would produce a few offspring and invest a lot into their development to reproductive adults.

I think real world experience suggests lots of humans are quite willing to reduce their birthrates and choose small family sizes - and that being economically secure and moderately prosperous with ready availability of contraception can lead to reduced birthrates without any overarching societal policy. Birthrates are decoupled from sex.

Being intelligent, reasoning technologists makes biological imperatives not imperative at all. When personal economic security depended on a sufficient continuing supply of your own descendants it was an imperative; many children improved your life. Within an advanced economy having many children reduces your prosperity. We still like having children but can be satisfied with fewer.

Space doesn't seem to offer any opportunities for population growth, not without fantastical technological progress - which undercuts the urge to have children; it isn't clear to me that expansion to space is somehow an expression of biological imperative to grow population.

 

[256bits]

falling off a bike is not quite the same as never having seen a bike before.

Is it this that you are getting at:
1. The interactions between individuals is as much as from learned experience as from being inbred.
2. Learned characteristics may supercede those inbred, so as to allow amicable relationships.
3. Learned characteristics result from interactions with peers and authority as to what works best.
4. An organization of persons follows rules, obtained from the learned characteristics of interaction, many written down, many not, to be cohesive.
5. A society is comprised of organizations, some of which will have have competing interests.
6. A civilized society allows for the competing interests to reach compromise and agreement.

[ If course, I may have not listed it all correctly, but I do hope the complexity towards adulthood is exposed ] [ What does having no actual parents, but rather a surrogate, do for a child in that the child can not run to mommy for a scraped knee, not discuss life expectations later in life ]
[ You and I get it. Humans, who are themselves machines, usually will not want to be treated as a robots - result rebellion ]

Points:
A. If the AI teacher, parent, role model falters in any way, the individual will develop unwelcome characteristics, in line with what @scott said "As for creating "stable humans", since when are humans "stable"?"
B. The AI will have to be trained based upon some sort of philosophy - ie in a manner based upon the two present day extremes of 'woke' vs 'Maga', or somewhere in between.

One problem I did find in the winner of the competition is that the inhabitants are to follow a set of strict rules, as if humans are robotic. For compromise( or reward not sure which) some rules are overly relaxed, some of them bizarre - unlimited sex, promiscuity, no families - defying the positive aspect of limits. The psychiatrist, psychologist certainly has his/her agenda showing the bias towards 1960's hippy movement, and stupidity.
That problem may show up with a generational ship with the living first passengers from selection, or the embryonic ship first generation from AI teachings. Neither one immune I would guess.
Incorrect selection or faulty teachings --> extinction, a result of no societal backbone.

 

[DaveC426913]

So does "AI can breed, birth, nurture and raise humans to a functional, aneurotic adulthood without any human interaction" count as a "real" project solution built on "sound engineering"? I thought real projects with sound engineering was what we were aiming for. Otherwise it's "sci-fi", no?

If not, if we can just say "AI will solve it all", then why can't this same AI simply produce a working fusion drive, or anti-matter drive, or worm hole, or warp drive? Then anyone could go - and in comfort, no less.

I think my ability to meaningfully contribute to this thread is rapidly reaching an end. It's a pity. It's a subject of great interest to me.

 

[javisot]

A preplanned curriculum that is executed automatically allows for many test runs in the decades before the launch. Test runs that will can be used to "debug" the entire process.

Yes, but we're talking about plans that don't allow for such prediction and automation. The timeframes are so long that factors like the physical evolution of the crew, technological advancements, or, in the case of no crew, the evolution of the destination planet, for example, come into play.

When we travel from Europe to the US, the time required is so short that nothing changes, and planning the trip is simple. The pre-established plan can even be executed perfectly. But interstellar travel is another matter. Even the first (automated) test flights can fail because they take so long to reach the destination and return the information. What is habitable can cease to be so.

A potentially habitable planet is habitable at the moment of detection, if indeed it is. If we could travel instantly or in 1,000 years to that planet, there wouldn't be a problem, but we could be talking about planets that are light-years away.

Automating this panspermia would not be easy; the AI would need to be powerful enough to completely change the plan at the last minute without needing to mediate with us (who are on Earth).

 

[PeterDonis]

The children would be free from physical and sexual abuse

Those aren't the only kinds of child abuse. How do you know that the complete absence of interaction with human parents doesn't qualify as abuse? Note that that is more extreme than any kind of parental neglect we have observed in human history--and parental neglect that is sufficient to significantly impair the child's ability to function normally as an adult (which, as I just noted, is considerably less extreme than the complete absence of interaction with human parents) might well be more common than overt physical or sexual abuse.

 

[Filip Larsen]

There is also another fairly obvious reason why designs for long duration interstellar travels (whether crewed or uncrewed) are less likely than comparatively shorter duration designs, that I haven't noticed been mentioned directly in this thread yet. And it is probably the main reason I almost automatically will consider long duration designs infeasible compared to shorter duration designs.

There are several ways to put it, but as a rule of thumb, as long as we, as a civilization, in T years can research and develop technologies applicable to an interstellar mission design that will shave off, say, 2T years of mission time, or give an equivalent savings in the resources needed for the mission design to be instantiated, then obviously it can "pay off" to wait and just improve the design. If the needed technology is even expected to be matured anyway, i.e. without regard to any interstellar travel, then this "waiting game" is almost free in the context of the design. This is why I believe near-autonomous automation is going to be front and center in any instantiated interstellar colonization effort, because such technology is by far the most feasible way for us to utilize the resources of the solar system.

This principle is actually so common that most near-future designs I have read about to some degree already depend on it in the sense that the design assumes some technology (typically fusion rocket engines) will have to be developed first in order to meet the mission specs. This is similar to "technology intercept" very often used for near-future designs proposals in pretty much all fast changing areas.

The only reason I can see that would make someone choose to instantiate a short build, long duration design is if there is no other option because those someone already has all the needed resources and capabilities at hand and also believe the opportunity otherwise will be lost forever (e.g. impeding civilization or solar system doom where the mission is seen as a lifeboat for human civilization). I will admit that it is probably is very prudent to keep such a design fairly ready for this type of situation, so in this context such a design can be relevant.

 

[PeroK]

And you see the creation, birth and raising of humans without any human interaction as a more tractible solution than any spaceship carrying humans?

Absolutely. For many reasons. The main one being reduced payload. The idea of carrying several generations of living, breeding humans seems almost totally implausible. In general, the principle must be to transport the capability to build and develop on the new planet, and minimise the amount of payload that we have to take.

That said, if you really wanted some human authority figures, you could send a relatively small number of adults to supervise things. They would have to sleep through the journey and would have to be worth their weight in gold.

Also, this can all be tested back on Earth. We would know to a large extent how children would develop under these circumstances. It's possible that, in any case, people born on the new planet might adjust to it better than volunteers from Earth who wake up on distant world. Not to mention that we may even need an amount of genetic modification to live healthily on the new planet.

That said, no one knows what human society will look like in, say, a thousand years.

 

[PeterDonis]

this can all be tested back on Earth. We would know to a large extent how children would develop under these circumstances.

How could you ethically test that?

 

[PeroK]

How could you ethically test that?

If humans in the future are serious about moving to a new planet, then they are going to have to take some hard decisions. That's something that a future generation would have to decide and prioritise. It depends how important the whole project is.

In any case, who knows how children will be brought up in the future? Currently, for example, adoption is a huge problem because the social care agencies are so nervous about who they place a child with. I know two couples who were (to my mind) inexplicably rejected. Whether it's utopian or dystopian, I can imagine that AI parents may be seen as a safer bet for adoption than real humans. I'm not saying it's right or wrong, but I can see that sort of thing happening.

 

[javisot]

Well, in this case we'll have to settle for thinking, "At least they have the AI; it would be worse if they didn't even have that." I don't think an AI would raise feral children, typical of those who haven't received any kind of upbringing.

 

[javisot]

We can all agree that if:

a- We have demonstrably capable AI

b- Human life can be created without human intervention (except for the contribution of biological material)

Then this is the most efficient and simplest plan. I would venture to say that if a and b are possible, this type of panspermia will inevitably exist someday. However, my opinion is that this is not possible because of b, not because of a.

 

[256bits]

What happens to the surrogate AI mothers and fathers? after the kids have matured.
I am quite sure the kids will figure out quite soon that some of these things are not like the others.

Do they disappear, vanish, sent to the annals of the mystery history.
Sounds like a new religion in the making.

 

[DaveC426913]

We can all agree that if:

a- We have demonstrably capable AI...

... then we can have it invent any kind of propulsion we want - even wormholes and warp drives - and all of us can go anywhere we want in style.

With AI as the ultimate generalized panacea, all optons are wide open.


But this goes by another name: "magical thinking".
1. We start with a premise.
2. We envision the end.
3. The middle bit will just work itself out.

It has become recently even more refined:
1. We start with a premise.
2. We envision the end.
3. AI!

 

[javisot]

... then we can have it invent any kind of propulsion we want - even wormholes and warp drives - and all of us can go anywhere we want in style.

Why do you think that?

 

[.Scott]

And you see the creation, birth and raising of humans without any human interaction as a more tractable solution than any spaceship carrying humans?

Are you a parent? Kids are like malfunctioning missiles. If parents are strategic, they can influence the trajectory. But lacking that, they will launch and follow a trajectory wholly of their own creation.
Unlike most kids, these will be in a very contained, controllable, pre-planned, and tested environment. The results will be far more predictable thana "normal" home environment. And, if we really want predictable and reliable performance, the kids themselves will be clones of the same genotype tested before the launch.

How could you ethically test that?

I just added you to the list of PF members that would be more ethical than any visiting aliens.

 

[DaveC426913]

Why do you think that?

It is self-evident if we start with the premise that AI is the solution to any problem.

 

[DaveC426913]

Are you a parent? Kids are like malfunctioning missiles. If parents are strategic, they can influence the trajectory. But lacking that, they will launch and follow a trajectory wholly of their own creation.

Without a parent or any human interaction, they will be eating, pooping, non-verbal vegetables. They will not go on any trajectory.

Everything - absolutely everything - beyond eating and pooping, you are placing in the magical basket of "AI will take care of that".



@PeroK I owe you an apology: it turns out you were right all along. This thread is divided into two factions: those interested in engineering practical solutions to a real-world future, and those intering in exploring more sci-fi-level tropes based on speculation.

I mean, that's OK too. It's my fault for not seeing both spearheads of the discussion.


Reporting to ask that this thread be moved to the sci-fi subforum.

 

[.Scott]

Without a parent or any human interaction, they will be eating, pooping, non-verbal vegetables. They will not go on any trajectory.

Everything - absolutely everything - beyond eating and pooping, you are placing in the magical basket of "AI will take care of that".

First of all, I am not putting anything in AI. I never said AI was needed - and the example I gave was only an incidental use of AI.

But I ask again, are you a parent? I doubt that you are. Especially as kids reach their teens, they form their own society - allowing only measured influence from their parents.

 

[Filip Larsen]

Reporting to ask that this thread be moved to the sci-fi subforum.

Why? You only want to discuss the topic in the context of fiction?

I sense that you are upset to discover that heavy automation, present and in the foreseeable future, for a long time has been considered to be able to outperform humans in most extreme environments, deep space travel included, but I am not sure if you are now trying to derail the discussion or really are OK with this.

 

[PeroK]

It is self-evident if we start with the premise that AI is the solution to any problem.

No one is starting with that premise. Let's assume that the project to colonize planet B takes place from 2500-3000. I.e. in the second half of this millenium. It's inconceivable that we would have the technology for interstellar travel and yet not have almost total automation by then. A prerequiste of colonizing a planet would be visiting the most promising exoplanets to carry out a proper investigation. This implies some sort of probe that has capability to travel perhaps tens of light years and send back the requisite data.

I understand that you are skeptical that such a probe could be unmanned, and that we would need humans on board to make sure the job is done properly.

Whereas, I find that inconceivable. Not least because of the massive additional complexity of having live humans on board a probe for decades. They would have to accept that are never coming back. Those missions would have to be automated.

The same factors would apply to any eventual colonizing mission. It would be massively more complex to have live humans and all the requisite life support systems on board. I can't believe that we would need humans to make the decisions and do the jobs.

Instead, you would almost certainly need a fleet of unmanned construction spaceships that would deliver the initial construction capability on the planet. The humans would have to come later. Keeping everyone alive on the new planet would be a major problem. Instead, you would need robots or androids to auomate the initiate construction process.

Finally, it's possible that we might send a fleet of ships with perhaps 10,000 volunteers per planet. But, I don't see those people living out a large part of their lives in interstellar space. I don't think that sacrifice is either necessary or desirable. Perhaps they could travel in some sort of "suspended animation"? Or, as already described, we could send frozen embryos and populate the planet that way.

In any case, without massive automation (including AI) the project is a non-starter. This seems to be staring you in the face. Everything on Mars has been done this way, for example. It's so much harder to get a human being to Mars and home safely than landing a rover on the surface. This has to be the future.

I don't know why you are so skeptical about the eventual development of automation and AI. Without it, I don't see how we could even contemplate moving to a new planet.

 

[javisot]

It is self-evident if we start with the premise that AI is the solution to any problem.

But that premise makes no sense and nobody said it, "any problem" is not even a defined concept.

When I said "a- Demonstrably capable AI," I didn't mean being able to literally solve any problem. What you're describing isn't AI; it's AGI, and not even that. Solving "any problem"? Do you really think AGI would be omnipotent?

Regarding the influence of AI on human development, both physically and mentally, it's very possible that we'll experience it firsthand and see the effects in the coming years. We certainly aren't going to put chatgpt in charge of babysitting.

 

[.Scott]

Attempting to raise kids without language (or even without full language immersion) will result in cryptophasia. With or without moral guidance, kids will socialize and form their own rules. Some of us as youngsters (Yours Truly included) would happily hang out for untold hours in the reference section of the City library consuming articles on Science, Math and History. And then, when exposed to a room-size computer, will literally sleep with it until they mastered every corner of it. Others will develop leadership skills.

The most active agent in the development of a child is the child.

It's good for kids to have human parents and human mentors. But they don't need them especially when they are part of a tribe.

Without a parent or any human interaction, they will be eating, pooping, non-verbal vegetables. They will not go on any trajectory.

Here's a citation for your side of the argument. For my side, I would simply respond that the infant cultivation procedures will need to be worked out further (and better) than Mr. Harlow got.

 

[DaveC426913]

First of all, I am not putting anything in AI. I never said AI was needed - and the example I gave was only an incidental use of AI.

Then this - which was the very genesis of everything we have been talking about - makes no sense:

But if we are going to develop technologies to make this transit work, we would probably be better off building a genetic human ark. The ark would operate automatically and autonomously. It would simply carry its cargo of biology equipment, training material, and human genome until it was within about 20 years of its destination. It would then begin "extracorporeal pregnancies", followed by automated up-bringing, and then finally graduation to planetary exploration.

But I ask again, are you a parent? I doubt that you are.

This is an ad hom.
If I were not a parent does that mean I cannot speak about early childood development?

That door swings both ways. You are not an aerospace engineer. Does that mean you have no business speaking about spaceships?


By the way, I am. Two children and two grandchildren. Thanks for asking.

Especially as kids reach their teens, they form their own society - allowing only measured influence from their parents.

Name me one teen in all history who raised themselves from a newborn, through feeding, diaper changes, talking, hormones and relationships without interaction with a single other adult, or even siblings or classmates* (all their siblings and classmates are the very same boat - they can't learn behavior from someone else who doesn't know behavior. They can't learn to wash behind their ears or give hugs from other people who don't know how to wash behind their ears or give hugs).

 

[DaveC426913]

But that premise makes no sense and nobody said it, "any problem" is not even a defined concept.

It is implicit.

You picked the single hardest problem in all creation. It stands to reason that any simpler problem will be solved easier.

If an AI is capable of birthing nurturing, raising, parenting, loving, soothing, psychologically-balancing, disciplining, training while preserving the sanity and happiness of the most complex object in the known universe (a human being from birth to adulthood), then a mere technical solution like worm whole or warp drive will be child's play.

When I said "a- Demonstrably capable AI," I didn't mean being able to literally solve any problem.

Of course not, you meant only one single problem - the very hardest problem the universe has to offer.

So a worm hole will be child's play.

What you're describing isn't AI; it's AGI, and not even that. Solving "any problem"? Do you really think AGI would be omnipotent?

I simply posit it will likely solve easy problems - such as space travel - before the most complex problem by far: an automaton, birthing and raising a functioning human adult, from scratch.

Regarding the influence of AI on human development, both physically and mentally, it's very possible that we'll experience it firsthand and see the effects in the coming years. We certainly aren't going to put chatgpt in charge of babysitting.

And I imagine one of the things they'll be set to work on is engineering solutions for space travel. Or do you think that's not a good use of AI?

 

[.Scott]

Then this - which was the very genesis of everything we have been talking about - makes no sense:

OK. So you are using the term "AI" to mean machines doing anything that you would normally expect to be done manually. From what you highlighted, I take it that any kind of autopilot (whether employing a computer or not) would fall under "AI". A lot of what you seem to be putting under the category of "AI" would be mostly or entirely handled by process control - technology that, as a aerospace engineer, you are familiar with.

I use the term "AI" to indicate software that has been developed in a way that the actual reasoning behind the automated decisions is not directly and explicitly derived from the requirements. As best I know, this is not allowed on safety-related components of any civilian or military aircraft.

This is an ad hom.
If I were not a parent does that mean I cannot speak about early childhood development?
...
By the way, I am. Two children and two grandchildren. Thanks for asking.

If I am talking to someone about any subject, it helps to know what their background is. Most parents have noticed that a whole lot of their kids behavior didn't come from the parenting at all. Since I am sure that you have seen this, I take it that it

Name me one teen in all history who raised themselves from a newborn, through feeding, diaper changes, talking, hormones and relationships without interaction with a single other adult, or even siblings or classmates* (all their siblings and classmates are the very same boat - they can't learn behavior from someone else who doesn't know behavior. They can't learn to wash behind their ears or give hugs from other people who don't know how to wash behind their ears or give hugs).

As far as hugging is concerned, I invite you to visit the Harlow link from my last post.
As far as washing behind your ears - a video on the subject might suggest that the end result could be more comfortable and that thorough cleanliness could be sexy. (and none of that would be "AI" by my use of the term).