Possibility 1: Civilization permanently collapses due to climate change, resource exhaustion and war. Over the ensuing millions of years the now-isolated human populations evolve into several new hominid species. Some are hunter gatherers, some have regressed to an animal existence. This is the most likely outcome I think.
Possibility 2: Humans have become a society engineered for maximum conformity and obedience. Every last person is a religious & political fanatic, both genetically engineered to lean that way and with brain implants designed to make unapproved thoughts impossible. All of humanity has only one opinion, on everything. Once established this society is impossible to overthrow, and lasts in perfect stasis until some natural disaster it can’t handle destroys it. After which any surviving humans quickly go extinct, as they are too mentally crippled to adapt.
Possibility 3: Once it becomes possibly to automate the economy and the military, the top fraction of a percent of humanity orders the machines to kill the other 99+%, except for a small breeding population kept as slaves. After an indefinite time the resulting tiny population dies out, leaving a world of mindless machines that keep the system running for their dead masters until the dying Sun burns Earth clean.
An argument which ignores that the Second Law only applies to a closed system, which Earth isn’t.
I’m reminded of a thread on another forum about a creationist who made that argument, who pointed out that it would be only possible for the complexity of life to increase if Earth’s biosphere had a constant, large input of energy. The title of the thread was “Creationist Almost Discovers the Sun”.
A look at science fiction since the very concept of humanity changing over time was introduced is both amusing and instructive. Basically views of the future are usually very short-sighted and based on taking then-current trends and extrapolating them, to often ridiculous degrees.
Take H.G. Welles’s The Time Machine for example. He postulates the class division between capitalists and workers lasting long enough for them to speciate into two different lineages.
People anticipated vast improvements in mechanical power and the technological and social changes that might result. But the idea that much or most mental work might someday be automated to eluded most futurists. In much classic pulp SF people are flying to Arcturus in spaceships kept running by engineers with wrenches.
Of course ANY prediction about the human race’s biological future is totally moot the day that genetic engineering of humans becomes practical. And how many stories about the future presume that people will always grow old and die of old age? It’s plausible that sometime in the next century we will finally work out just how senescence is programmed into organisms, and edit it out.
Finally, there’s the unknown unknowns about the laws of physics and what knowledge of them might ultimately make possible. Two hundred years from now we might have an entirely revolutionized understanding of the nature of the universe.
Agreed it’s not impossible. But being an engineer I’m trained to think in terms of probabilities. I’m usually not relying on things to happen that are out more than 2 standard deviations. Based on that poll of “experts” back in 2008, they were estimating about a 19% probability of human extinction by 2100. That’s starting to get my attention. Pure speculation, this is true, but at least from people that spend their time thinking about Global Catastrophic Risks for a living. (Although some might argue that experts from pure academia like University of Oxford are not making an honest living).
A thousand years ago we couldn’t have a real time conversation with someone a hundred miles away, let alone someone on the other side of the planet because “physics”. It is nearly the ultimate hubris to presume we have done much more than scratch the surface of math and physics.
While I believe we’re probably well passed the point where one person can devise the next calculus or trigonometry, I don’t doubt that there are many more fields to be discovered/developed with expanded collusion between researchers and access to greater computing power.
Will pertains to living systems. I don’t think things can be programmed to want something. I haven’t heard of anyone achieving this and I don’t think it is possible. (Believing this involves a huge amount of faith, in my opinion.)
This is the IMHO forum, not GD - my post was idle speculation, and I do not care to engage with you to discuss it when you’re using this sort of sanctimonious discussion style.
Hmm. Not to change the subject — well, not to change it much, and I figure the tangent is relevant — but: what about a machine beating the Turing Test?
Because, see, I don’t think I’ve heard of that being achieved; but I don’t think I’ve heard experts rule it out as impossible, or laugh it off as wildly improbable.
And yet, to do that, wouldn’t it have to act like it wanted stuff? Presumably the only reason I do a passable job of responding the way a human would — over the course of a conversation — is because I want stuff. Heck, anyone who’s in the habit of reading my posts hereabouts can presumably mention (a) not only that I want stuff, but (b) exactly what stuff I want.
And I figure the only way a machine could do a passable job of responding how a human would is — well, by behaving as if it wanted stuff; and, of course, the whole point is that it’d have to be so good at that as to be indistinguishable from someone who actually does want stuff. Right? It couldn’t just blandly type answers to trivia questions at human speed, while maybe tossing in a typo or two; if it can’t show what you say it can’t have, it’d presumably fail on the spot.
Just to be clear, “Neanderthals” did not exist 500K years ago. What existed close to 600K years ago was the common ancestor of Modern Humans and Neanderthals. But more to the point, genetic engineering could drastically change our appearance in the blink of an eye, with respect to evolutionary timeframes. And I suspect it will for some of us. We may be dozens or hundreds of different populations that are unable to interbreed without some laboratory help.
It might work for a while, the scenario you describe. I see some assumptions in there though that inject a big chance for failure, making it not as simple or easy as you describe. What you are talking about requires a lot of specialized knowledge that the vast majority of us don’t have. I know that when I push this button on my computer it does that, and I can even vaguely tell you why. I couldn’t tell you for the life of me where to find let alone process the raw materials to make a computer, or a saw, or a bolt or any of the stuff that makes present day civilization possible.(hell a standard screw was pretty expensive{difficult} high tech not all that long ago relatively speaking). The saving grace for your scenario is the ability to salvage from the remains of what was until that specialized knowledge is regain in a more general way amongst those who remain. Given the nature of information storage and retrieval today, given the somewhat disposable nature of manufactured items today, I’m not certain that salvaging from the remains of today’s civilization is a very viable strategy. I could be wrong though.
Yeah, thats my view. Part of me wonders what human life would’ve been like without fossil fuels. I am sure tech progress would’ve been much slower but there would’ve been far more research into wind, tidal and solar energy (with solar, you don’t need photovoltaics, you can just use solar concentrators to boil water to power generators), and eventually we would’ve discovered nuclear power and that would form the backbone of technological society.
And yes, I agree civilization can’t really crash after a certain point. If it is spread to endless solar systems and exists in multiple formats including ones we can’t fathom, I don’t see how a virus or meteor can destroy it. Supposedly when a society reaches level 2 of the Kardashev scale, it becomes immortal. Perhaps by that level you learn about time travel, creating new universes or traveling to other universes within the multiverse to avoid the heat death of this universe.
I look at all the stuff Russia went through in the first half of the 20th century. Massive civil war, WW1, flu pandemic, Stalinist purges, great depression, Stalinist famines, WW2. It caused tons of death but Russian civilization survived. Humanity is more resilient than people give it credit for.
Software that shows conversational intelligence is nothing but software that shows conversational intelligence.
When I said, “Mankind’s values cannot be inherited, continued or bettered by AI, which lacks human beings’ will and ambition to make dreams come true.”, I answered the OP request that we should speculate. This is my two cents, take it or leave it.
I’m not planning to debate this issue. To satisfy people’s curiosity, I will only mention that my standpoint is somewhat similar to that of Ian Bogost, who says that “in most cases, the systems making claims to artificial intelligence aren’t sentient, self-aware, volitional, or even surprising. They’re just software.”
Successful speculation involves extrapolation; unsuccessful one falls prey to overestimation.
Maybe, but even so, that doesn’t mean I’ll be able to use that knowledge. Much of the technology we have today depends on other technology that depends on other technology and all of it depends on processes for manufacture that I may or may not be able to reproduce. OTOH I may get lucky and fall in with a group that includes a miner, a metallurgist, a machinist, etc. (I don’t think that’s likely, but not out of the realm of possible). I guess it depends on the level of collapse we are talking about between now and 1 to 100 million years from now.
As far as the OP, 1 million years from now, I think we will be much the same as we are now physically. Maybe a little different in stature, maybe a few less teeth, possibly a slightly elongated skull, but not really all that different. 100 million years from now, all bets are off. Would we even be human as we understand the term today? My guess is we would be Homo something something something sapiens
I agree that is the state of AI now, and for longer into the future than many people consider. Not only that but I don’t think AIs will ever be ‘human’ in the sense we consider it because they’ll be machines, not subject to the inherent failings of our biological brains. But that doesn’t mean that machines cannot one day be as inherently worthwhile as humans, likely to be a mix of our original biological intelligence enhanced by technology, and possibly exceeding our own valuable humanity. Or a horror of unspeakable proportions, hard to say which way it will it go.
Well, if there’s 10,000 remaining salvageable tablet computers in the world, and 1 of them has a copy of wikipedia on it, you could transfer that single digital file to all the tablets if you were a ‘priest of Android’ so to speak.
The tribe that does this could lock all but a few tablets for use in a vault, powering them ultimately by DC-DC converters directly (removing the internal battery and powering it directly from a solar panel)
The tribes doing this would after a few centuries have a huge advantage if all the competing tribes descended back to the stone age.
And that’s just one way. Obviously lots of other equipment would survive the disaster, and some of it would include the “gigs of storage, super portable, tiny” advantages of a tablet.
Even just knowing the scientific method lets you work out from whatever knowledge level you end up at how to proceed from there.
And what if the big disaster happened 20 years from now? With the tablets of 2037 all be waterproof, solar powered, and have batteries that don’t degrade? Or be so cheap and common we use them instead of paper in almost all situations? I could see that.
See, my two cents is that you’re glossing over entirely too much, there.
I have values. I have human values. I’m a human being who, brimming with will and ambition, routinely displays — during conversations — that he wants stuff.
You handwave, in general, the significance of showing conversational intelligence. But I’m saying that, in particular, showing wants and values is a subset of showing conversational intelligence. I’m saying that passing the Turing Test means responding the way a human with wants and values would. I’m saying its behavior would have to be indistinguishable from that of a wills-stuff-and-has-ambitions human, or else you’d spot the phony when it failed to converse like one.
But that’s just it: for purposes of this discussion, I don’t care whether it actually is sentient or self-aware or volitional or whatever. If it responds the way a human with sentience and and self-awareness and volition would, then it doesn’t matter whether it has those; what matters is that it acts like it does — and, again, if it doesn’t so act, it’d fail the test. But an AI that behaves the way a human one would, as if motivated by human values, would (a) AFAICT pass the conversational test regardless of what was or wasn’t going on behind the mask, and (b) instantiate those values.
Except they probably rapidly end up illiterate, and can’t actually read what’s on the tablet. So it just becomes an object of ritual veneration until it finally breaks.
I don’t know if I agree fully with Der Trihs either. I think though if you’re talking a civilization collapse of that scale, you’re going to lose a lot of current tech. One of the ways is simply through lack of people, both with the knowledge of how to maintain the current level of tech, and just from simply not having enough hands to do the work needed. A lot of tech will remain viable for various amounts of time depending on what the tech is. The lower level the technology, the more viable it is for a longer time. It’s pretty easy to make plow shares and spear points out of whatever already refined metal you have around you. Its also pretty simple to rig up a water wheel or wind mill to provide some sort of useful energy, but building and maintaining an electrical generator of useful size and capacity is a different thing altogether. There is going to be some loss of technology. I do agree though that it should be possible to limit that loss with what we have around us already and to regain that lost technology pretty rapidly, maybe within 3 to 5 generations for the loss cycle to finish and recovery to at least begin. I don’t see us reverting completely back to pre-industrial levels of technology even though I think it would be a close thing.
ETA your essentially indestructible tablets of 2037 will be mostly useless as soon as the internet collapses from lack of maintenance of the infrastructure that supports it since cloud storage seems to be the current trend.