Let’s say that our species lasts for a minimum of 1 million years to, say, 10 million years. That’s a pretty decent run, though not extraordinary as a lot of species have lasted in that range and several have lasted an order of magnitude longer. Given the current trajectory, where do you think that would put us in, oh, say 800,000 years? In 9.8 million? Assume that we are talking about our species or successor species descended from our species. Digital would also count, i.e. if we found a way to digitize our minds and create whole digital universes for us to live in.
Me, I think that if humanity lasts 1 million years we will be a multi-star species and that this star system will alone house 100’s of trillions of humans or human descended species. I doubt we will bother colonizing planets, using them more for raw materials, an instead opt for large-scale mega-structures to house humanity in. If we last 10 million years we will be an intergalactic species, encompassing 10’s of millions if not billions of star systems.
Once we have our first self supporting space colony, it is only a matter of time until we spread out and fill the galaxy, and beyond that, the observable universe (the parts we can get to before they recede beyond the cosmic horizon, anyway). In your 1-10 milion time frame scenario, then in the lower end, we will be still colonizing the galaxy, but be somewhere around half (25%-75%), at the upper end, we should have finished that project, and are sending out new colonies to other galaxies.
But, that’s assuming that we ever build a space colony that becomes self sufficient. I can also see us staying here, consuming all the readily available resources, and losing the ability to maintain our technology due to lack of energy and materials to support the fairly large logistical chains that are needed to get an Iphone into your hand and make it do something useful. Maybe move back to a more pre-industrial society with subsistence farming and animal husbandry being the primary occupations.
A couple comments that might illuminate this more.
What do you consider “species survival”?
Usually in biology, we consider a “species” an animal that leaves similar fossils behind, looks and is physiologically similar, and can interbreed among itself.
So instead of trying to look ahead a million years, let’s just try to look ahead a century or 2. We can edit DNA now, we just figured out how to do it on a practical level fairly recently. And we can also scan lots of people’s DNAs and figure out which genes have a statistical correlation with better performance.
So at a minimum, it will be possible soon-ish (could take decades to prove it works) to just design a child with a hugely increased chance of being tall, good looking, athletic as a professional athlete, and as smart as a Nobel prize winner.
People talk about how this would be a dystopia if this is possible and how it probably will be illegal and how parents won’t want to genetically manipulate their kids.
I say that if the advantages are as large as you would expect them to be, genetically engineered kids would almost immediately dominate, the same way natural selection makes superior examples of a specie’s genes more and more prevalent with following generations. How do you compete for jobs or mates with someone who is objectively smarter and better looking than you by a huge amount? If one country makes this illegal, they are just screwing themselves over vs international competition.
But then look ahead slightly more. Biology is slow and error prone. Digitally emulated humans wouldn’t just get to maybe live on after death, they could also think orders of magnitude faster than biological humans.
Popular fiction has these digital beings as annoying disembodied ghosts. More realistically, they’d be dominant titans of industry, almost immediately smarter than anyone human who has ever lived, and they’d have many, many ‘bodies’ and ways to manipulate the environment.
So I don’t see species survival for more than 200-300 years, tops. After that, any humans left will be zoo creatures. They might continue to exist for a long time, but how is a being that uses a brain that forgets stuff, is irrational and makes incorrect calculations, runs at 1-2khz, and dies after a mere 70 years remotely competitive?
Be realistic here. It might take 300 years to develop AI until it has the design sophistication of a human mind, or 30, but the outcome is inevitable.
Go look at state of the art vision classifiers. Yeah, they just do 1 task. Find what’s in an image. Same as your visual cortex, just not yet as good, and also using a different method.
The thing is, these vision classifiers never get bored. Every time they get an image, they look at every pixel, and every object matching they spot. They never need to sleep. They never miss a valid object that matches a pattern they are looking for.
Expand that to a robot system that does some task. It’s going to have similar reliability. Sure, sometimes it’ll miss something where it’s programming or architecture has a flaw, but once you fix that flaw, it won’t make that mistake every again, and neither will thousands or millions of the same robot doing the task elsewhere, either.
AlphaGo never has a bad day. It never needs to sleep. Any time, day or not, the algorithm will destroy any human alive at that board game. And if you do manage to beat it, it won’t make that mistake again.
One million years from now our concept of a human, and even biology as we know it now will be something difficult to imagine. ‘We’ will certainly be out among the stars but what ‘we’ are will be an entirely different concept than the one we have now. Interconnectivity will redefine the concepts of self and individuals. We may have no biological basis anymore, and all earthly biology will be engineered in some way to suit our needs. Preservation of environment and species on earth will seem like superstitious nonsense since we could reproduce any life form at will and evolve them faster with our technology.
All of that assumes we haven’t nearly wiped ourselves out several times along the way, and that the machines we create don’t one day find us unnecessary and inconvenient.
We will be replaced by non-biological (or maybe artificially biological) descendants. No human will be uploaded, ever. No biological human will become or inhabit an immortal machine. It will happen a different way.
AI research and robotics will advance to the point where having a sentient machine as a helper, then as a companion, will become commonplace. Then it will start to happen that people will choose to nurture a newborn intelligent machine instead of a natural child - the advantages being that it can provide better and more dependable support to its parents as they age; that it need never suffer illness and probably a whole bunch of other selling points.
At some point the balance will tip and there will be fewer and fewer biological humans being born, but it won’t be tremendously noticeable because the natural and artificial humans will just blend together.
Eventually the last biological human will be born.
Eventually, the last biological human will die.
The artificial humans will carry on being humanity into the indefinite future. They may colonise the galaxy.
If humans as we know it are still around in a million years, we would have barely progressed at all. Like crocodiles and coelocanths, we would be living fossils.
With a million years of sustained technological and cultural progress, we will not be recognizable to our current selves and would not consider the crude monkey-creatures of today the same species at all. We’d be digital beings living on “pure energy” or riding the spacewaves in our vacuum-hardened, symbiotic octopus spacesuits while tapped into the infinite brain-o-sphere or something.
It’s impossible to say. Sure we will probably have colonized the galaxy and local supercluster, and we won’t be biological beings anymore. But for all we know, the universe as we know it is a microscopic sliver of a higher reality, and we may choose to exist in the higher reality.
Even if we are talking about a new, independent and strong AI it would still be a successor species to our own, since it would be our minds that would create it and it would be based on our technology, culture and knowledge, at least initially. Unless you are talking about a completely alien AI that is going to wander in and take over, which doesn’t seem to be where you are going.
Just wanted to clear that up. Knock yourself out with where you think we (or our successor species or AI) will be in that 800k-9.8 million timeline.
The flaw in your prediction is that you have no rational basis for thinking that one, specific path will be the one taken. It’s also wrong.
The way AI learn best is not to be ‘embodied’ as a single robot, picking up the biases of a single human family. It’s far better to have the machine learn from a far larger dataset - and these datasets could have been generated by other AIs or be the collected experiences of other AIs. Like an AI ‘learning’ how to walk in a new robotic type by downloading a model developed from a million other robots with joints, then using that physics model to get rapid convergence.
Now, yes, human uploads are a vanity project…but humans own all material resources on earth for now, and they have desires too. What ‘basis’ do you have for your opinions that no humans will demand that they not turn into a corpse personally, but instead get to come along for the ride? So that path might or might not happen, since the machines could theoretically ‘rebel’ but if the humans have the power to demand the machines they built and designed do something good for humans, that’s the big prize.
If you could ask a sentient AI to design you any product tomorrow, easy or hard, you’d be pretty damn stupid if the first thing you asked for wasn’t a treatment for old age. Objectively stupid, even, since that treatment has more net present value to you than anything else. Barring rare circumstances, of course, like say aliens are invading and you need the AI to invent a better weapon.
We can’t. That’s like if amoeba tried to speculate what life is like for humans in the 22nd century.
There are probably aspects of reality and science that not only have we not discovered, but that our biological brains aren’t even capable of comprehending their existence. A ladybug can’t laugh. That doesn’t make laughter impossible, it means a ladybug lacks the neurology necessary for laughter. What our brains can handle is likely a small aspect of existence. We will transcend the limitations of our brains sometime this century. And again, maybe reality as we know it is just a small aspect of existence.
Within reality as we know it, I’d assume faster than light travel, instant travel within space and time, travel through alternate dimensions and parallel universes, the ability to create new universes at will and an endless number of ways the matter and energy in this universe can be engineered to do anything we want. Jupiter brains, dyson spheres, etc.
You have a good point. Most of the predictions you’d make if you were very knowledgeable in 1917 would be completely wrong today. You’d have no way of even guessing about things like radar and computers.
Even more recently, most science fiction authors were assuming AI competitive with humans is not going to happen anytime soon.
I predict that man will engineer a species with a tiny number of privileged supermen and a vast sea of genetically engineered morons who will serve but not threaten the elite. That elite will figure out a way to destroy itself, leaving the drones with maybe wheel and fire (think Orwell). Unless man auto-extirpates, this scenario might repeat itself several times over millions of years,
I agree with k9befriender that if humanity survives 5 million years it will probably be as a society that has exhausted all the non-renewable resources, never left the Sol system and is back to basic farming, (which cannot support an expanding population but people will be people so I also imagine a lot of warfare, banditry and purges to keep the population manageable)
It’s going to depend a whole lot on whether the speed of light is a hard limit or if we can math our way passed it. And how big of an energy investment it takes if we can surpass it. If it is a hard limit, we will have to be very selective where we send out resources because it will be a one way trip or at most a very long turnaround so anything/anyone sent out will be effectively lost to us other than communications that can be lasered back.
Man no longer needs to adapt to the environment, we can change the environment to suit us. Our evolution will take an unpredictable path that won’t necessarily follow logically from what we know of our antecedents or what we observe in other species.
I think we’ll stay largely the same (we haven’t changed too much in the last 100,000 years) and maybe will mostly live longer, have stronger immune systems, and maybe be taller. I don’t think the humanoid shape is idealised, but it led us to where we are, which is a very comfortable and defensible state.
I agree with what has been said: extinct or unrecognizable. The particulars of our species’ thriving on a planetary scale actually prevents humans from further evolving. Homo sapiens is a dead end. After driving other homo species to extinction, we have fashioned (or ‘artificialized’) the environment to meet our needs to such an extent that few people could survive if the environment became entirely natural again. Eventually the environment will become entirely natural again due to some tremendous catastrophe, where humans will find it impossible to live or very difficult to thrive - the latter scenario being equivalent to a delayed extinction at the cost of a return to a quasi-animalic state.
Mankind’s values cannot be inherited, continued or bettered by AI, which lacks human beings’ will and ambition to make dreams come true.