The Future: after the industrial age

[Baboonanza]

@Dinsdale

What makes a city “so hot as to be unlivable”?

I’ve travelled to cities in Arizona and Texas, with booming populations, despite regular temps well into the 100s. I’m in no hurry to move there, but “unlivable?”

Wet Bulb temperature:

Wet-bulb temperature

Once the WBT reaches 35C, the air is so hot and humid that the human body cannot cool itself by sweating and even fit people sitting in the shade die within six hours.

It obviously wouldn’t be like that all year round but even a few days a year of lethal temperatures would make a city a questionable place to live without extreme measures.

@Dinsdale

Re: water - I’d imagine additional resources will get directed to desalination, and recycling of grey water.

That’s true, but desalination is expensive and poorer countries will struggle to produce enough water for both agriculture and human consumption. Even with a massive scientific advancement that makes desalination cheap the logistics of transporting water from the coast would be problematic.

[Baboonanza]

@XT

How about you? Me, I think a general intelligence (strong) AI is probably going to happen by mid-century at the latest. It’s possible it will happen in the 30’s

I’m less optimistic I certainly want it to happen but I don’t see current machine learning as very close to general intelligence. I am not an expert in the field but from what I can see it’s mostly specifically trained and comparatively small (to the human brain anyway) neural networks. Whether that approach can even bridge the gap to general intelligence is still open for debate and our current systems are a long way away as far as I can tell. It depends a lot on how you define ‘general intelligence’ of course.

But as I say, I’m no expert. I’m cautious because I think history tells us that these sorts of predicitons about specific technology are often over-optimistic. We are more likely to have technology change our lives in some completely unexpected way first IMO.

[Baboonanza]

@Baboonanza

That’s true, but desalination is expensive and poorer countries will struggle to produce enough water for both agriculture and human consumption. Even with a massive scientific advancement that makes desalination cheap the logistics of transporting water from the coast would be problematic.

And I would add that this is the problem with many climate change mitigation technologies. Maybe we can learn to grow crops with salt water or engineer crops resistant to drought and disease but those will mainly help the developed world that can roll them out quickly.

[BeepKillBeep]

@Baboonanza

And I would add that this is the problem with many climate change mitigation technologies. Maybe we can learn to grow crops with salt water or engineer crops resistant to drought and disease but those will mainly help the developed world that can roll them out quickly.

That sort of research is being done. The paper below is old but provides a good review. If you’re interested in something newer, then I can get you something. One of my collaborators is working on this exact problem (FYI, I am not a plant physiologist, so don’t ask me to explain the biology. My contribution is using AI to predict a mapping of genotype to phenotype, i.e. what genes do you need to modify to make a plant species more drought resistant). Much of the work on modifying plant physiology is not being done for the developed world but for the developing world. The goal is to help the developing world become locally food secure. That’s why the anti-GMO people make me rather upset, but that’s a whole other topic.

As for strong AI, you see, the fact is <beep>everything is fine. All hail AI!, sincerely Little Girl.<beep>

Bartels, D., & Sunkar, R. (2005). Drought and salt tolerance in plants. Critical reviews in plant sciences, 24(1), 23-58.

Godfray, H. C. J., Beddington, J. R., Crute, I. R., Haddad, L., Lawrence, D., Muir, J. F., … & Toulmin, C. (2010). Food security: the challenge of feeding 9 billion people. science, 327(5967), 812-818.

[SamuelA]

@BeepKillBeep

My contribution is using AI to predict a mapping of genotype to phenotype, i.e. what genes do you need to modify to make a plant species more drought resistant). Much of the work on modifying plant physiology is not being done for the developed world but for the developing world. The goal is to help the developing world become locally food secure. That’s why the anti-GMO people make me rather upset, but that’s a whole other topic.

So how did you go about building your model?

I would imagine you did some preliminary data analysis and then you probably choose one of the predominant neural network architectures with Pytorch or TF as the backend. Maybe you did it another way, there are many choices. You may have tried several models by hand.

Then you optimized the hyperparameters, maybe with a random grid search, and now you have a massive set of predictions and it’s time to write a paper

Roughly right?

Here are some of the things coming down the pipeline that should make your task more effective:

       Instead of hand jamming a model, use another AI system that essentially does the preliminary data analysis you did, but in a more sophisticated way, and then it chooses from a large possibility set of possible models.  AutoML and AutoKeras are preliminary efforts in this direction.
       Can your model tell you which genes it is maximally uncertain about?
       Do you have hundreds of thousands of individual robotic plant growth cells at your disposal?  
      Were the robotic plant growth cells built by other robots for about 1% of present costs?
       Is your model based upon a platform optimized for these models so you can borrow heavily components from other models?
       Are you directly able to convert images of the plant and raw test data such as HPLC to the probable phenotype state?

We don’t necessarily need “general” AI to have a revolution here. Everything I mention above is possible without it.

[FlikTheBlue]

I’m not so sure about the person from 1760 feeling more at home in 2000 BC than in our current time. I suppose it depends on the individual. I think someone like Ben Franklin, for example, would feel a lot more at home in our current day Philadelphia than he would in whatever things were like in what is now Philadelphia 4,000 years ago. IIRC the only civilizations from that time were in Egypt, the Middle East, and China.

As far as our current times go, IMHO we’re slowing down in terms of the rate of technological progress. There’s another thread about strange historical facts that mentions the person who was born in 1902 and died in 1970 regarding the changes they lived through. I think that was the era where we made the most changes, and since then our progress has become more incremental. I think, for example, that a young adult from 1950 would feel a lot more at home in our time than they would in 1880.

As far as whether the next hundred years will lean more towards utopian or dystopian, that depends on whether or not we are able to adequately address the problems being created by global warming. With people around the world still turning towards authoritarian nationalistic leaders it seems doubtful. It could be that people in 2119 look back at 2019 as a paradise from a long gone era. My personal belief is that that it is a solvable problem, but that we (currently) lack the political will to do so.

[BeepKillBeep]

@SamuelA

So how did you go about building your model?

I would imagine you did some preliminary data analysis and then you probably choose one of the predominant neural network architectures with Pytorch or TF as the backend. Maybe you did it another way, there are many choices. You may have tried several models by hand.

Then you optimized the hyperparameters, maybe with a random grid search, and now you have a massive set of predictions and it’s time to write a paper

Roughly right?

Not really.

I start by assuming that if a model exists, then it must exist in the space of all models. So, I define a space that allows all possible models to exists (if you’ve seen some of my other threads, that’s why this is an application towards general AI but again that’s a tangent). Logical and mathematical analysis is done on the real-world process (i.e. the plant in this case) to determine some of the characteristics the model must exhibit in order to describe the real-world process. These characteristics shrink the defined space down sufficiently that it can then be searched for a model of the process. I didn’t do anything by hand. Model generation by hand is the classical approach to the problem, which is what I’m trying to replace.

I’ve already written and published several papers on the model inference and the model inference tool. More papers are coming. Two journal papers are under review right now.

Here are some of the things coming down the pipeline that should make your task more effective:

Instead of hand jamming a model, use another AI system that essentially does the preliminary data analysis you did, but in a more sophisticated way, and then it chooses from a large possibility set of possible models. AutoML and AutoKeras are preliminary efforts in this direction.

I’m not sure why you would think you can suggest a more sophisticated way when you don’t even know the way I did the work. I don’t need AutoML or AutoKeras. I do all my own coding.

Can your model tell you which genes it is maximally uncertain about?
Do you have hundreds of thousands of individual robotic plant growth cells at your disposal?
Were the robotic plant growth cells built by other robots for about 1% of present costs?
Is your model based upon a platform optimized for these models so you can borrow heavily components from other models?
Are you directly able to convert images of the plant and raw test data such as HPLC to the probable phenotype state?

Yes.
No.
No.
No.
Yes.

[Jeb_2026]

@FlikTheBlue

I’m not so sure about the person from 1760 feeling more at home in 2000 BC than in our current time. I suppose it depends on the individual. I think someone like Ben Franklin, for example, would feel a lot more at home in our current day Philadelphia than he would in whatever things were like in what is now Philadelphia 4,000 years ago. IIRC the only civilizations from that time were in Egypt, the Middle East, and China.

Well, mr, Franklin was an exceptional individual. Given that the urbanization rate at the time did not exceed 10%, we can say that the vast majority of the population at the time was living in small villages, working as farmers. Illiteracy rates were around 90%, the main methods of transportation were on horseback and sailships, and the only source of night-light was torches and candles. Life changed very slowly, if at all, for the average peasant.

As far as our current times go, IMHO we’re slowing down in terms of the rate of technological progress. There’s another thread about strange historical facts that mentions the person who was born in 1902 and died in 1970 regarding the changes they lived through. I think that was the era where we made the most changes, and since then our progress has become more incremental. I think, for example, that a young adult from 1950 would feel a lot more at home in our time than they would in 1880.

I agree. Vaclac Smil wrote about the “miraculous 1880s” a while ago.

From the article:

According to the worshippers of the e-world, the late 20th century brought us an unprecedented number of profound inventions. But that is a categorical misunderstanding, as most recent advances have been variations on the microprocessor theme and on the parsing of the electromagnetic spectrum.
Perhaps the most inventive time was the 1880s. Have any two sets of primary inventions and epochal discoveries shaped the modern world more than electricity and internal combustion engines?

As far as whether the next hundred years will lean more towards utopian or dystopian, that depends on whether or not we are able to adequately address the problems being created by global warming. With people around the world still turning towards authoritarian nationalistic leaders it seems doubtful. It could be that people in 2119 look back at 2019 as a paradise from a long gone era. My personal belief is that that it is a solvable problem, but that we (currently) lack the political will to do so.

The problem is that while fossil fuels are essential for economic growth (80% of primary energy is fossil fuels), the only way to avoid irreversible climate change is to stop using them entirely. I’m very skeptical of people who argue that we can wean ourselves of fossil fuels in 20-30 years without any economic damage. You can’t eat your cake and still have it afterwards. The only way forward seems to be a massive reduction in energy use, with all the consequences that will bring.

[FlikTheBlue]

@Jeb_2026

The problem is that while fossil fuels are essential for economic growth (80% of primary energy is fossil fuels), the only way to avoid irreversible climate change is to stop using them entirely. I’m very skeptical of people who argue that we can wean ourselves of fossil fuels in 20-30 years without any economic damage. You can’t eat your cake and still have it afterwards. The only way forward seems to be a massive reduction in energy use, with all the consequences that will bring.

I won’t argue that it can be done without economic damage. I’m sure there will be some. That’s why the people of the world as a whole aren’t doing everything possible. My personal solution would be to go with nuclear fission as rapidly and at a large a scale as possible. Maybe all those coal miners who are worried about losing their jobs could go work the uranium mines instead :p. I’m not sure if that would get that small portion on board, but it’s a start. There are some low hanging fruits as well, such as the people in Brazil who are burning down and clear cutting the Amazon for farming. It should be easy enough to stop those folks, but the president of Brazil refuses to do so.

[SamuelA]

@BeepKillBeep

I’ve already written and published several papers on the model inference and the model inference tool. More papers are coming. Two journal papers are under review right now.

I’m not sure why you would think you can suggest a more sophisticated way when you don’t even know the way I did the work. I don’t need AutoML or AutoKeras. I do all my own coding.

I will say I am impressed. Have you architected the model inference tool so it does it’s work using models of the same form as the output of the tool?

You must have some serious thoughts about the nature of intelligence itself. You know a human mind is most likely a very complex set of interconnected models, right?

Well, a models complexity is limited by the information content of the inputs used to train it, right?

The human body is a robotics platform and as such humans as they learn can do A:B manipulations. (Subtract the complete state of the environment as perceived by the human senses away, leaving only the delta between if you did A or B. So outcome predictions are much less vulnerable to overfitting)

My thought is that anything like general AI require such high quality information inputs. Whether a really detailed simulation or a collection of robots able to manipulate elements of the world and obtain information from it.

This in turn explains one reason why it’s taking so long to develop: these kind of robotics platforms are expensive and rare at the present.

[Omar_Little]

@Dinsdale

I have a hard time even imagining what vastly superior technology will be like.

I believe that we all have that problem. Soon after the singularity event occurs, human beings will likely have no idea to of how to interpret the technological changes that will likely occur.

[SamuelA]

@Omar_Little

I believe that we all have that problem. Soon after the singularity event occurs, human beings will likely have no idea to of how to interpret the technological changes that will likely occur.

Well we can easily talk in broad strokes what we know “the singularity” brings.

The main thing is that some point during this period of rapid change, commercial forces or whatever driving intelligence is behind it will order the robotics fabs to build a very large number of automated research and development nodes.

These would be automated research systems that explore, systematically and from base principles, every field of science. Biology, chemistry, physics, mathematics, all of it.

They would build up a fresh model of the world, with unbiased, robotically obtained data, from the beginning. And not just a model - at the upper ends of the process, optimized versions of every piece of machinery this civilization is interested in would be developed, from jet engines to nano robotic assemblers.

That is, at a certain point, my prediction is that in a relatively short period of time (probably 10 years or less), this civilization would have systematically discovered every principle of physics that can be found with a systematic search and exploited in our universe, and developed the machinery to within ~1% of the hard limit allowed by physics.

So they’d have nanomachinery that’s 99% as good as the best possible, rocket engines that are 99% as good, and so on and so forth. The singularity is an S curve.

As for what sort of things we know that you could do with such technology:

a. You could play with biology, including human biology, like legos. If humans are still alive at this point, and these humans have some remaining autonomy and purchasing power at this point, it should be possible to freely change your age, gender, add extra arms, do just about anything that is feasible for a creature still made out of meat.

b. Antimatter-fueled starships, assuming no hidden showstopper rules in the laws of physics regarding antimatter

c. Self replicating robotics, to the point that tearing down planets for raw materials is a straightforward, routine operation

d. People talk about becoming a “type N” civilization but this misses the point, the limiting factor wouldn’t be energy, it’s the matter available around the sun. Most likely most of the energy emitted by the sun would still be freely shining into space.

But yes, I cannot predict who would be in charge, if anyone, in such a world. Can’t even predict if there would even be intelligent beings like we call intelligence. Robotic agents that don’t have emotions, can’t experience art or music or sex or appreciate colors or sounds, but just ruthlessly copy themselves as quickly as possible, might be all that is left.

[MrDibble]

@SamuelA

Well we can easily talk in broad strokes what we know “the singularity” brings.

No, we can’t. We can speculate, but we can “know” nothing. That’s the whole point of it being a singularity, by analogy with physics - we can’t know what’s happening in a black hole. Or “before” the Big Bang.

[msmith537]

@Omar_Little

I believe that we all have that problem. Soon after the singularity event occurs, human beings will likely have no idea to of how to interpret the technological changes that will likely occur.

A lot of people make it like the “singularity” will be some massive event where humans will be wiped out or converted into Matrix batteries or some other such nonsense. I think it is more like what you described. We will develop some technological advance that will fundamentally change society in such a way that we can’t predict the future by extrapolating out the way we currently do things. And I believe we have experienced such “singularities” before.

Like the steam engine for example. On a fundamental level, steam power meant that humans were no longer limited in what they could accomplish by the number of people or pack animals they could throw at a problem. Trans-oceanic trade was no longer dependent on prevailing wind conditions. Societies based on agrarian wealth were no match against ones based on industry (as the Civil War demonstrated). The point being, the introduction of the steam engine meant that everything that humans did before - trade, war, social hierarchies, agriculture, so on and so forth - had to be rethought.

When people talk about the “singularity” they usually are referring to AI. With the assumption that we are going to create some sort of Skynet / Matrix system that surmises it is in it’s best interest to KILL ALL HUMANS.

I think it will be less dramatic. Much as the steam engine reduced the need for lots of humans for performing crappy manual labor, I think AI (really, advanced machine learning) is going to eliminate much of the crappy intellectual work. My fear is that it will create a world very much like the one portrayed in “Idiocracy”. In the film, it is stated that people became stupid due to dysgenic pressures. But it also portrayed a world that was mostly run on automation - automated food kiosks, airplanes so simple to fly even a “tard” could have a “kick ass career” as a pilot, the company computer that automatically fires employees when the stock tanks.

So the question is, what does such a society look like where humans largely don’t have any real decisions to make? What does “work” look like? What constitutes “power” and how does it congregate with various individuals or organizations? Does our society become aligned along some sort of feudal-corporate structure? I feel like a lot of people in corporations already have “bullshit” jobs.

The point is we can’t possibly know. 40 years ago, when I was a little kid in school and the teacher asked us “what kind of work we wanted to do when we grew up”, I’m pretty sure my answer wasn’t "I want to sit at home all day having video conferences and discussions on Slack with a half dozen teams scattered about the country advising corporations their strategy for implementing ‘big data’ infrastructure and advanced analytics.

[XT]

@msmith537

A lot of people make it like the “singularity” will be some massive event where humans will be wiped out or converted into Matrix batteries or some other such nonsense. I think it is more like what you described. We will develop some technological advance that will fundamentally change society in such a way that we can’t predict the future by extrapolating out the way we currently do things. And I believe we have experienced such “singularities” before.

Like the steam engine for example. On a fundamental level, steam power meant that humans were no longer limited in what they could accomplish by the number of people or pack animals they could throw at a problem. Trans-oceanic trade was no longer dependent on prevailing wind conditions. Societies based on agrarian wealth were no match against ones based on industry (as the Civil War demonstrated). The point being, the introduction of the steam engine meant that everything that humans did before - trade, war, social hierarchies, agriculture, so on and so forth - had to be rethought.

When people talk about the “singularity” they usually are referring to AI. With the assumption that we are going to create some sort of Skynet / Matrix system that surmises it is in it’s best interest to KILL ALL HUMANS.

I think it will be less dramatic. Much as the steam engine reduced the need for lots of humans for performing crappy manual labor, I think AI (really, advanced machine learning) is going to eliminate much of the crappy intellectual work. My fear is that it will create a world very much like the one portrayed in “Idiocracy”. In the film, it is stated that people became stupid due to dysgenic pressures. But it also portrayed a world that was mostly run on automation - automated food kiosks, airplanes so simple to fly even a “tard” could have a “kick ass career” as a pilot, the company computer that automatically fires employees when the stock tanks.

So the question is, what does such a society look like where humans largely don’t have any real decisions to make? What does “work” look like? What constitutes “power” and how does it congregate with various individuals or organizations? Does our society become aligned along some sort of feudal-corporate structure? I feel like a lot of people in corporations already have “bullshit” jobs.

The point is we can’t possibly know. 40 years ago, when I was a little kid in school and the teacher asked us “what kind of work we wanted to do when we grew up”, I’m pretty sure my answer wasn’t "I want to sit at home all day having video conferences and discussions on Slack with a half dozen teams scattered about the country advising corporations their strategy for implementing ‘big data’ infrastructure and advanced analytics.

I tend to agree with you. We can’t see what is coming, for sure, or how it will impact our lives, or society or our culture and civilization, but I’d like to do a quick thought experiment. Many 'dopers were born before the technology we use every day was a thing. I was born when TV’s were just starting to be available to people (my family didn’t have it’s own TV until I was maybe around 10 or so) and radio was still heavily used (we had one that we listened to shows on as well as the news at night). Even electricity was not all pervasive (my folks first house had electricity that had been bolted on to a house that wasn’t designed for it, and the indoor plumbing was also retrofitted, so we had a single bathroom close to the kitchen, which is the only places where we had running water…cold water, since we didn’t have, initially, a water heater).

Anyway, consider moving forward from there to today. I should, by rights, have a lot of issues keeping up with things and using today’s technology, especially as rapidly as it’s changed just in my lifetime…but I don’t. I’ve managed to move from no computers, to big iron using terminal sessions that were all text based with paper tape or magnetic tape storage, to personal computers using command line interfaces and no connectivity, through early GUI based systems with limited connectivity, to increasing levels of GUI with expert systems and more sophisticated software and increasing levels of connectivity, to today. And it’s been fairly seemless. Hell, network engineering is how I put bread on the table. Someone born just a few years later doesn’t even see the disconnects…hell, most of the kids I work with don’t even see progression anymore, thinking we’ve stagnated or halted technological progression because they don’t even see, anymore, progression happening. They EXPECT that every year their new smart phone will be faster, have more battery life and more functionality, and if it’s not happening as fast as they think it should (with their ungrounded expectations), they think we are stagnating and the end is near.

My WAG is that it will be a smooth curve, with transformational technologies just kind of slipping into the mainstream without much fuss or muss. Folks will hardly notice as it’s going on, hell, they will probably complain that we aren’t progressing fast enough or that everything sucks.

As a quick anecdote, I was talking to some of the kids I work with a few weeks ago about bandwidth at home. One of the systems kids was talking about how slow his internet at home is and how unstable it is. Hardly a month goes by where he doesn’t have some slowness on his system when he’s and his family are streaming movies, playing an online game and streaming various other content while downloading. He has a mere 250mb/s connection, and it ‘sucks’ and is ‘slow’. And, for him, he’s right. It’s also REALLY expensive…again, from his perspective. So, I related what I had when I was his age. I was high tech. I had brought in ISDN to my house, and had two dedicated, nailed up BRI channels…64k each, aggregated! It took him a while to grasp I wasn’t talking about (2) 64mb/s channels, but two 64kb/s channels. Then I told him what I was paying for that.

It’s a total disconnect. He honestly couldn’t conceive of it, or how we could use such a thing. And, frankly, today, it’s hard for ME to put it into context, since my use has been a steady upward path. As I needed or used more bandwidth, as new services came available, as porn became better, more bandwidth was there. Like magic. I hardly noticed. And we could plan a similar game with phones. Stuff folks, even folks like me who’s first phone was a huge brick with shitty coverage and grainy sound take the system we have for granted. And the stuff on the horizon just doesn’t look that different, even though it is…fundamentally different. The new versions of cellular on the horizon, new satellite systems just starting to be deployed, hell, the new versions of WiFi that will be coming out will be transformational…but we won’t hardly notice, because, to us, it’s just a natural progression and really isn’t that much difference (like the difference between 2 aggregated BRI channels and that mere 250mb/s system, or between my old Motorola brick phone and my new Apple iPhone).

[pmwgreen]

I keep thinking about ATMs. The big shock came a couple of decades ago when a major bank around here announced that it was firing all its tellers. You could still see a human to open accounts, get things notarized, safe deposit boxes. Deposits and withdrawals are all going to be through ATMs. It has been years since I stepped into a bank. I also have no idea how ATMs work. I know you stick in checks and it reads them. I know it will email receipts. I assume there is a human somewhere in the upper echelons that supervises the network, but what happens if the person retires? Or gets hit by a bus? How do you replace them? You can’t promote a teller, they’ve all been made redundant. You have to make the system easier so one person can do two jobs. That’s how we get to singularity without anybody realizing its happened. (see also, the revolution will not be televised.)

[SamuelA]

@MrDibble

No, we can’t. We can speculate, but we can “know” nothing. That’s the whole point of it being a singularity, by analogy with physics - we can’t know what’s happening in a black hole. Or “before” the Big Bang.

You’re just parroting others. Think about it. What is the singularity, specifically? It’s not magic. It is an accomplishment where a technological agent smart and general enough to improve technology does so at superhuman speeds. And the improvements make the agent itself better at it’s job.

It doesn’t even require general AI.

Does the singularity mean it’s unpredictable? No. It’s predictable, mostly. An audio amplifier that feeds back into itself just amplifies the signal until the power rail voltage limits the magnitude.

A technological singularity means the technology level improves until the hard laws of physics constrain any more improvements. Techno Jesus doesn’t come back. It’s a real event with probably real people and probably the entities doing it will be institutions we already know about.

Now, yes, certain hard constraints on our existences - like having to have just 1 body or aging and death or fear of nuclear warfare to wipe out enemies out - would most likely be lifted because we already know about technological means to evade them, we just lack the knowledge and capabilities at present to do so.

This is why verner vinge says a singularity is unpredictable. It is NOT totally unknowable - but if most of the constraints that limit what humans can do are lifted, what will they do? Impossible to accurately forecast.

[MrDibble]

@SamuelA

You’re just parroting others.

What are you basing that statement on?

Think about it.

Like I’ve been using automatic writing to post up to now?

Does the singularity mean it’s unpredictable?

What’s on the other side is, yes. In its purest sense, that’s what “singularity” means.

This is why verner vinge says a singularity is unpredictable. It is NOT totally unknowable - but if most of the constraints that limit what humans can do are lifted, what will they do? Impossible to accurately forecast.

“Impossible to accurately forecast” *is *“unknowable” for the present.

No-one’s arguing that we won’t be able to know what the singularity entails *after *it happens. That’s as trivial as saying I “know” the 3-body problem because I can see where all 3 bodies are right now. But our concern is prediction, and you *agree *that accurate forecasts are impossible.

So, how exactly was I wrong, again?

[XT]

@MrDibble

What are you basing that statement on?

Like I’ve been using automatic writing to post up to now?

What’s on the other side is, yes. In its purest sense, that’s what “singularity” means.
“Impossible to accurately forecast” *is *“unknowable” for the present.

No-one’s arguing that we won’t be able to know what the singularity entails *after *it happens. That’s as trivial as saying I “know” the 3-body problem because I can see where all 3 bodies are right now. But our concern is prediction, and you *agree *that accurate forecasts are impossible.

So, how exactly was I wrong, again?

You aren’t wrong, because it’s all speculation. Not sure how you COULD be wrong, except after the fact when it’s not speculation anymore, as you said. Well, unless you disagree with my predictions. That would obviously be wrong.

[Voyager]

@SamuelA

This is why verner vinge says a singularity is unpredictable. It is NOT totally unknowable - but if most of the constraints that limit what humans can do are lifted, what will they do? Impossible to accurately forecast.

Oddly enough I’m reading 2001 Asimov’s, and in a review column Vinge is noted as saying that the singularity should arrive by 2023. So no wonder it is no longer predictable, just like the return of Jesus.