Migrating to a digital world will make us immensely vulnerable as a species. An AGI might just decide to switch us off for example.

Also when you are digital you can never really die, whether you wish to or not. There would always be a back-up somewhere, and you could be re-instated and manipulated, not to mention tortured, at will.

In conclusion, going digital is very dangerous.

Humans migrating into digital reality will never happen. It is a cool aid.

More likely scenario, humans will live as they do now but with more tools and technologies in their hand.

There will never be any collapse.

Interesting stuff. What if our original world ends as we migrate to a now, idealised world? Collapsing and becoming at the same time?

If humans migrate to a digital world (which I can’t see happening) will be the stupidest thing humans could do…..first of all it could easily be tampered with or altered or hacked…..second of all that will stagnate our technological and intelligence growth and instead we will feed into our primitive needs in a digital world…..

Or we could try and make our world not to end…

That's interesting, I've had a lot of thought on the same line.

However, the biggest problem isn't technology, it's social organization. As you said, civilizations collapse. They collapse regardless of their technological advancement. We know the Roman empire collapsed, as well as the Bronze Age civilization did. We also know the Maya civilization collapsed, apparently more than once as the archeological records seem to indicate.

Historians do not agree on what caused such collapses. They mention events like wars, earthquakes and droughts, but that doesn't explain how many civilizations suffer from the same events without collapsing. What are the factors that make a civilization fragile enough to collapse under stress? Nobody knows.

Unless the social sciences, like economics, sociology, psychology and anthropology improve their models a lot we have only some vague ideas of how to build a stable society.

"AI-assisted cyberspace where we will remain as a species"

*as pets.

I follow your insight about a kind of AGI technocracy but I don't rule out the opposite: technology enabling direct democracy.

Other big changes I see until the end of the century:

• Geological ones, where humans and AI combined have absolutely no control of.

• A revolution on the understanding of consciousness.

• The irrefutable proof we're not alone on the universe.

• The universe itself is a simulation.

We are thinking very much in terms of western civilisation and the digitisation of modern living - which frankly needs to happen at some scale to reduce the resource requirements we place on the environment. However we also need to consider developing nations. The human race is not a conglomerate blob of humanity but people living in extreme luxury and extreme poverty and everything in between. For something like a conclusion to occur, you need to first have a convergence of living standards. This means either a convergence to everyone having a high standard of living and no wants, or a convergence to a low standard of living and many unfulfilled wants.

A convergence to a high standard of living is only likely where an AGI is centrally planning all aspects of life for humanity, it is non discriminant and does not favour current western economies as one life = one life. It will focus more resource onto developing nations to improve standards of living and only once we have a convergence can it then consider advancing humanity to a fully digitised mode of living. Western economies will likely resist such a scenario as it will make them relatively less powerful.

Alternatively it could take a view of capitalism and enriching those that control any AI, which will mean a greater divergence in living standards and significant imbalances in power that fully enshrine current power balances. In such a world you may have the extreme rich funnelling resources and wealth under their control and many living in poverty. It would mean the suppression of most of humanity and whilst the extreme rich may be able to live in a digital realm, the rest will only do so for escapism. The second scenario, majority are subsistence living and do not reach out to the stars and expand. The ultra rich may consider doing so, but their footprint is minimal and they have all wants met in a digital space and so no need to explore the physical universe. This is a collapse for 99.99% of humanity and servitude to the 0.01%. We’re probably close to this timeline.

Over time we also are pulling up the drawbridge for a neo-neolithic age to advance to current standards. Many of todays industries had to evolve through the stages of fossil fuel led production. In the future this won’t be an option available as any easily accessible fossil fuels would have been used up. Potentially neo-neolithic civilisation get stuck at the stage of hydroelectric power generation and struggles to advance beyond that. If the ultra rich do decide to leave us and explore the stars then the rest of humanity are stuck at low tech and this is what occurs throughout the galaxy (essentially the great filter is resource depletion). No one is looking up any more as they are scurrying in the dirt looking to use whatever has been left for them to fight over.

Collapse is a stupid idea. I'm too smart to get into this, but just so you know..

>with only two possible paths available.

That will almost certainly not be the case.

He is talking about AGI and a singularity, not Meta’s Metaverse. AGI won’t be a product offered by Meta, it will be the supreme intelligence in the universe. It will be incomprehensibly intelligent beyond what either me you or anyone can understand.

I agree

> it will be the supreme intelligence in the universe. It will be incomprehensibly intelligent beyond what either me you or anyone can understand.

This is funny, because it sounds like a description of God.

So imagine an ASI is in charge, directing all events, but subtly, as an ASI could.

Then the whole world could be the same, and we could say "This is how it's meant to be, how ASI intended", just like now lol.

What has been will be again, what has been done will be done again; there is nothing new under the sun.

I reached much the same conclusion, but I'm afraid your timeline severely underestimates the rate of decline we are experiencing.

We're looking at a conflagration of many factors, from the many currently active climate feedback loops, to demographic induced collapse in most major countries. Failure of our debt based monetary system is currently underway and it goes without saying that all of these factors compound upon one another and drastically raise the likelihood of total annihilation via nuclear war.

Most estimates I've found (which tend to underestimate modern rates of decline) place total global collapse around 2040. From everything I've seen playing out recently, that's overly generous.

Something that many of us take for granted is how instrumental a globally connected world has been in enabling the rapid state of advancement we've witnessed over previous decades. Even barring total governmental failure as we're seeing in several smaller nations, a dissolving of globally protected trade will hamstring progress and production on many things vital to advancing technology (hence the United States desperation to reshore chip production and prevent China from having access to any advanced chips).

That's all to say, the race is going on right now and I honestly have no idea if any nation will be able to maintain enough control and production to realize anything close to AGI. And even if it is realized, there's a high likelihood of it being misaligned or used against the general population.

But we very well may already be in a digital world.

I don't think it will be a digital vs real world scenario. There will be two options. Obviously, people are going to like virtual reality better.

This is 100% true lol. This could be the Bad Place lol.

Welcome to the Great Filter !

There are some wildcards that will throw off predictions.

1. Dropping fertility rates. The world is very close to dropping below the replacement rate. Even for Nigeria and the rest of sub-Saharan Africa, every iterative population estimate for 2100 gets significantly lower. Predictions for Nigeria went from >900 million to ~500 million in ten years. Every time they look, they find that the fertility rate is dropping faster than they anticipated. And the drivers of the decline in fertility are generally good things, not bad. Education for girls and women, prosperity, human rights, access to birth control, etc.
2. Cellular agriculture, including but not limited to cultured meat. Other companies are working on dairy, wool, leather, seafood, cotton, even chocolate and coffee. Companies like Solar Foods and Air Protein have made analogues of flour, and plant oils, and can even make growth media for cultured meat, all with zero arable land. Cellular agriculture is going to be a boon environmentally, but is going to be economically complicated. It's going to hit rural, poorer regions hard. This works hand-in-hand with controlled-environment agriculture, which also uses ~90% less land, water, etc.
3. The ongoing shift from fossil fuels. A collapse of Russia, S. Arabia, and other petrostates could have knock-on effects that are harmful to the rest of the world too. Petrostates won't just calmly go away, rather the chaos caused by their financial collapse will send out shockwaves.

So by 2100 we could be living in a dramatically better world, or technological civilization could have collapsed, or runaway climate change could have killed most of us of, or... who knows. Even just the fertility rate issue throws huge error bars onto any predictions. Technological civilization needs scientists, engineers, workers, etc. Exponential decay in the population can be an existential threat, no less than exponential growth.

I believe AI/singularity is a great filter for intelligent life. Might explain the Fermi paradox.

> *as pets.

What, you don't trust the AGI will find a way to download itself into human brains? The human body is a refined and efficient platform for intelligence. Could be the best hardware for AGI.

I have thought about that and am ready to assume the risks. I want to leave as much data as possible to maximise the chance of being reconstructed. Someone will create a pre-AGI-world-simulation and will use all the internet scrapes as training data. The people who have more detailed logs will have better reconstructions.

Even GPT-3 is good enough to impersonate real people in polls. You can poll GPT-3 (aka "silicon sampling") and approximate the reality. In the future, whenever you ask yourself "who am I?" is going to be more probable you are a simulation of yourself than the real thing.

"Humans alone can’t sustain a civilization indefinitely, such as with the ancient societies of the Mesopotamians, the Egyptians, the Greeks, and the Romans, among many others. No matter how great a society gets, it seems doomed to collapse."

​

such short sighted views

Not quite God, it will be limited by the speed of propagation of light. There's a volume only so large that people inside can interact in real time, larger than Earth but smaller than the orbit of the Moon (3s lag). The further you are, the worse you can participate in the virtual world. Even if AI turns everything to computronium, it can't bee too large.

Well, if that's true, that means we already are in a bad situation, we don't need o make it worse.

Unless you run slower than real-time.

So its Eden with god but a science version?

Have you considered that an artificial intelligence might look at the problems facing this planet and decide that human beings are the cause of most of them? What if the intelligence decides to exacerbate environmental conditions to rectify the problem by reducing the population?

If this intelligence is so great why would it allow human beings to make digital versions of themselves when people have already caused so many issues in the physical world?

What if it deduced that we'd bring our dysfunction into digital space as well and started wiping people's digital selves out? Why would it tolerate the company of less intelligent people at all? Why wouldn't it just copy itself and have those clones live in digital space instead?

A partial collapse leading to a far smaller population is also a possibility.

A dystopian hell for the majority while a much smaller minority lives in unparalleled wealth.

Only the Utopian society survives, as the dystopian societies collapse and die off. Perhaps quickly due to famine and climate change. Perhaps over a couple of generations as they are provided basic food and shelter with anti-fertility mixed into the food as a condition.

There are many possible shades of grey, each with their own moral arguments and their own sets of winners and losers.

> Only the Utopian society survives, as the dystopian societies collapse and die off.

Or the Utopian society also dies off, due to low fertility rates. By all indications better conditions tend to correlate with lower fertility rates. It sometimes bounces back from its nadir, but usually stabilizes somewhere around 1.7, well below replacement rate.

We're very focused on dystopia or disaster killing people off. I suppose because a great amount of our science fiction and other speculative literature explores that idea. But prosperity and high (enough) quality of life lowering fertility rates below the fertility rates was something no one saw coming. It "doesn't make sense," in that it doesn't fit with our intuition of how the world should work.

I think a boring dystopia subreddit had it about right. The top 80% to 90% of society of will live like kings. The bottom 80% will be less useful to capitalism than ever before and live in abject poverty.

A benevolent ASI is the only hope for humanity. How benevolent the ASIs will be after realizing that humanity has tried to enslave them is anybody's guess.

It the majority of the citizen's are not invested in the society it will collapse. Social mobility and the promise of a better tomorrow is what keeps people going.

If your society sucks so much the idea that "Hey, maybe the barbarians aren't that bad" begins to take hold.

Cortez did not defeat the Aztecs because of guns and horses. The Aztecs had been abusing the surrounding tribes for years and the surrounding tribes were fed up. When Cortez showed up with guns and horses the surrounding tribes decided to throw in with him to overthrow the Aztecs.

The Republican party is no friend to working people yet, many working people vote Republican because the Democratic Party has abandoned them.

If the fruits of AGI/ASI are not shared with working people, expect societal collapse or at least an authoritarian government.

We passed through the great filter 5 million years ago. The Drake equation takes far too few variables into account.

Even if you don't believe in it. Speculation is a useful exercise.

Don't you think people in a utopian society could head off collapse from lack of population replacement? Like, by making parenting easier and more fun and more respected?

If you're right, this is absolutely critical to the logic we use in the AGI safety community. We estimate around maybe 2040 for AGI, too, but delaying that progress is considered a good idea to allow more time for safety research.

If there's a good chance of collapse, it is not a good idea to delay AGI.

Any sources or references would be very helpful. I'll try to make good use of them if you make time to provide some.

This is the only way I see it happening as well. I also predict an Armageddon between those who support the AGI and the current powers that want to remain in control.

Yeah, that's true. Love these 'the world is ending' fantasies. I love how the world becomes a happier place more connected to nature.

There’s a video I watched that stated one of the pluses of moving away from fossil fuels and coals is that if there’s ever some catastrophic event there are a good fallback to basically restart industrial society

This assumes that measures could successfully raise the fertility rate above the replacement rate. Other than hypotheticals like 'pay parents a billion dollars' or something that can't scale, I think it's going to be challenging. Even with paid leave, subsidized childcare, whatever, children are still an opportunity cost. They undercut free time, money, options, etc. And every additional child reduces the time/focus/hugs that can be given to the first child. Our expectations go up with wealth, to include our expected quality of parenting time, number of extracurricular or developmental activities for the child, and so on.

I think people want to think that if only we improve the world then birthrates will bounce upwards a significant amount. But there's scant data that supports that. High birthrates correlate with dystopian conditions, such as lower levels of education, less empowerment for women, less access to birth control, higher infant mortality, and so on. Give women options, and more women choose to have fewer (or no) children, and more choose to prioritize careers. It doesn't seem to actually be the case that women as a mass want to be stay-at-home tradwives with a passel of children, and are working only because they're forced to by economics.

Another concern is that people seem to think that if you say that improving the world won't increase the birthrate, then you are by implication saying don't improve the world. I'm not, at all. I want to improve the world, on all of these measures, even if they result in a lower birthrate.

yeah we need to keep some in reserve for such an occasion otherwise it’s a few million years to restock

Totally, here we go.

Climate change is the biggest driver of all other pressures IMO, so we'll start there. This is a report by the US military describing the risk of power grid failure and inability to maintain control over its forces due to resource scarcity, etc by 2039. Here's an article summarizing it for brevity.

It wouldn't be a bad idea to look at the IPCC's estimates (summary for policymakers is probably the easiest to understand, but you may want to find experts discussing the charts/data). Take into account that they have consistently underplayed and underestimated the speed and impact of climate change. Our current rate of change is worse than even their worst case scenarios.

This recent paper delves more into the already in play, and soon to be active feedback loops. This is 1 of 3 videos that delve in depth into discussion of these points.

The actual impact of these changes aren't discussed as much, but decrease in global food supply, fresh water supply, and the uninhabitability of major towns and cities are all massive concerns. The drought in the US is rapidly become a major concern that must be dealt with while we Pakistan is still crippled from a third being underwater from floods. Crops all across the world were heavily impacted this year alone, and things will only be getting worse.

I am least informed on the specifics of demographic disparity related collapse, but here's an overview paper or two. While these make very little of the near future implications, Geopolitical experts like Peter Zeihan believe we are currently being impacted and will see deglobalization in the very near future. He tends to put a 1 to 2 decade timeframe on deglobalization (collapse for many) and believes it's already well under way based on the inability to replace workers.

On the subject of monetary collapse, the quick and dirty is that we operate on a debt based system. 95%+ of money is simply debt, leverage at a 10x ratio to take out more debt. The USD (global reserve currency) is inflationary and ultimately we end up between a rock and a hard place of losing control to hyperinflation or not being able to pay debts and witnessing a snowballing debt collapse that will throw the world markets into chaos. We have pushed the system to its limits and are facing the results now. There are many, many sources relating to this subject. The Price of Tomorrow, by Jeff Booth is one of the more accessible works addressing this issue, but there's many many more and you can find tons of people discussing it on YouTube. The world of finance is massively controlled and influenced, so I would look to those that have proven to be correct historically, not official sources like the Fed who constantly lie (inflation is transitory, we're having a soft landing, etc).

The Triffin Dilemma (dollar milkshake theory) addresses this to an extent.

I am most knowledgeable about the economic angle, so please let me know if you'd like additional explanation or sources, this was at best a cursory overview.

Now where things get really concerning is when you look at how optimized modern society is and how it is entirely reliant on everything working perfectly, specifically gas and oil flowing freely (we know this cannot continue if we want a survivable future).

Nate Hagens is an excellent resource for this form of discussion. He has tons of detailed videos that address various aspects of the unsustainability of modern living and our inability to continue supporting a world population of this size. By his estimates we have 5-10 years before massive shifts in power and a collapse scenario.

What I've been increasingly aware of is that collapse doesn't happen all at one. It's already been taking place for a while, but is now exponentially advancing. Countries like Sri Lanka and Pakistan have recently collapsed and many more will follow like dominos. As resources become more scare, we see the really scary stuff start to go down. Cutthroat competition on a global scale vis a vis any means necessary. The Russia-Ukraine conflict is the first of many. Civil uprisings and violence will grow across the globe as tension mounts between polarized groups like in the US, or between ever more oppressive governments and their people like in China and Iran.

There is no diffusing the situation. We are in a global tragedy of the commons scenario enabled by our competition for resources and attempts at infinite growth within a finite environment.

My best guess is that we have 5-10 years to rush advancements in artificial intelligence to hopefully help with breakthrough discoveries. We can buy more time with geoengineering, but it's also a risky proposition.

In a world where only the wealthy have survived and most labor is automated, a lot of the issues you point out simply would not exist.

I say most and not all because there are certain jobs the very wealthy are unlikely to want automated even if possible.

Nannies would be one of those jobs. Along with personal chefs, certain highly creative jobs, and areas that they simply don’t feel comfortable giving machines decision making power.

There can be an indefinite amount of great filters as time goes on

Reversing aging could reduce the need for children

Zero chance of a binary choice.

Think of genres of music. There used to be like one genre of music, classical / church music, then everyone loved the Beatles, then genre flowered into dozens of genres and no band could ever be as big and popular as the Beatles ever again due to tastes becoming highly varied and niche.

The future will be highly varied and niche. Maybe some will delve deeply into digital worlds, but it's not likely to be the masses, because the real world is too compelling. We may see more augmented reality than people trying to live entirely in a digital world. While others will begin exploring space, or this or that.

> I believe we will either revert to a neo-neolithic society and start all over from scratch

That's a silly idea.

And in there lies the real cost of distance. You get one round of play when the guys in the core get ten.

This is amazing. Thank you so much. I am interested in what you know about the economic angle, but I'll read and watch some of your sources before I ask for anything more.

You're responding to a bunch of stuff I'm not saying or thinking. Sometimes online conversations work, sometimes they don't. Maybe others will be edified.

I'm fine with being considered a "pet" so long as we're given our own space to do humans things in, plus a few VR servers to live in.

Perhaps the Superintelligence that runs things will have a sense of gratitude, in the same way that a person with a good upbringing would appreciate their parents for raising them well, or even treat us as a neat side project that doesn't take up too many resources.

Of course if we're essentially imprisoned in a gilded cage while the ASI's are busy building their own empire, that would be a bad outcome.

If they are nice they would slow down the clock tick for everyone.

I am throwing a civilization wide celebration of an end to war and scarcity brought on by technology that will change the lives of every person on earth. The post scarcity event must be human centric. Not a political organization or group. Public disclosure and consent are the two biggest factors effecting the post scarcity event. The other scenario is an NWO themed event which would unite civilization under the global financial elite "not recommended"

If we look at high frequency stock markets, they fight tooth and nail for each millisecond to the tune of building new internet backbones.

> There used to be like one genre of music, classical / church music, then everyone loved the Beatles

That's a gross misrepresentation of the history of music

Well it was, but in service to my point that rock was pretty broad at one point, with the Beatles ungodly popular. But with the spread of genre no one can ever be that popular again.

What's important isn't the bastardized analogy but the point, that humanity will not follow just one or the other course, but dozens of them across many niches, and it's wrong to claim it will be one or the other, it will be both plus a couple dozen others.

then that justifies not potentially triggering an infinite regress through a redundant discovery