The Ascent of Mankind

[Dado94]

funny thing about quantum physics is that even physicist say they don't really understand it. indeed academics stopped debating about how to understand all the strange random nature of particles in the theory a long time ago. the last time it was debated was in the time einstein was alive and strongly opposed the probability in quantum mechanics with his well none quote "god does not play dice". when he died physicist ceased to bother what their calculation meant as long as the values they computed fitted to the experiments.

since that time theoretical physics developed slowly and there was no major breakthrough. the derived superstring theory turned out to be an epic fail yielding no new results. QFT helped to get more precise results then original QM but for the trade-off extremely complicated calculus computers can't even handle. attempts of quantum gravity are totally broken. so all we have are small improvements to the basic QM but nothing revolutionary. the situation reminds me somewhat on the year 1900 where physics stagnated until relativity and QM came. personally i think they won't progress much further until someone resolves the core problem of randomness in the theory... and this time seems close.

Questions:
1) if the nature is really "randomness", how can the QM be a problem? Isn't it the right way?
2) Superstring theory is now only "under construction": how can you say it will be an epic fail? why?

 

[Killtech]

Questions:
1) if the nature is really "randomness", how can the QM be a problem? Isn't it the right way?
2) Superstring theory is now only "under construction": how can you say it will be an epic fail? why?

1. not the randomness itself. but how the probability is obtained is. or to be more precise: the measurement process. it's cannot be correctly defined how that works but is essential for every interpretation of QM as it is responsible for obtaining the probabilities out of the wave function. partially it is explained with the collapse of the wave function but this concept is inconsistent in general...
the basic problem starts with schrödingers cat gedankenexperiment. since a measurement of the cats defines its state you could just ask what action is counted as a measurement and which not. and since a measurement is nothing more then an interaction in general this is a problem. e.g you could say that every single atom permanently measures the cat because it feels its mass via gravitation interaction. thus the cat must not have an uncertain state ever and the QM breaks.

2. superstring is under construction for quite some time and it has found nothing useful yet. so there is a rising crowd of skeptics doubting this attempt will yield anything at all. up to now the theory is a disappointment. but you are right. it didn't fail... yet.

 

[os79]

See why I love Civ4? It has every branch of academics and beyond under one umbrella. I grew up with a strong science/math academics but got bored with them alone in college career. Right now, I'm aiming at becoming a history professor at Gallaudet University one day. Because i got bored with specific social science subjects alone. Finally I found what I needed: history! History is only subject in academics that can synthesize all other subjects under one roof.

Not related to my future career, but my hope is that historians will inspire future social and natural scientists to work together such that the divisions of these become practically non-existent except in formal documents.

I had to say this because of a fascinating discussion going on here .

 

[Killtech]

Not related to my future career, but my hope is that historians will inspire future social and natural scientists to work together such that the divisions of these become practically non-existent except in formal documents.

the day that the first real AI is created will be the day when social and natural sciences will certainly merge. at the one hand psychology and social studies can found all their theories on the basic model of an intelligence and derive all properties form it's basic structure. at the other hand a source code is enough for nature sciences to build a formal model for further of a mind that they can work with.

 

[iontom]

See why I love Civ4? It has every branch of academics and beyond under one umbrella. I grew up with a strong science/math academics but got bored with them alone in college career. Right now, I'm aiming at becoming a history professor at Gallaudet University one day. Because i got bored with specific social science subjects alone. Finally I found what I needed: history! History is only subject in academics that can synthesize all other subjects under one roof.

Not related to my future career, but my hope is that historians will inspire future social and natural scientists to work together such that the divisions of these become practically non-existent except in formal documents.

I had to say this because of a fascinating discussion going on here .

I don't know how many "academics" play civ. I know plenty of kids my age who play, but nobody with any real scholarly responsibility. I myself have been studying astronomy and physics for the last three years but I'm hardly an academic. Instead of putting enough time into understanding physics I've sank it all into 4 categories: beer, women, rugby and Civ. Fortunately astronomy is easy enough to get away with that, but not physics. I can't speak for everybody though, I'm sure there's plenty of other achievers like you out there that get their civ on as well.

the day that the first real AI is created will be the day when social and natural sciences will certainly merge. at the one hand psychology and social studies can found all their theories on the basic model of an intelligence and derive all properties form it's basic structure. at the other hand a source code is enough for nature sciences to build a formal model for further of a mind that they can work with.

Yes, Hard AI will change everything. If you've ever heard of Ray Kurzweil, he's made some bold predictions about the rate of increase of technology. He's a smart sonn'of'a'birch, and worth looking into with an open mind, but somewhat skeptical eye.

According to him, by 2029 we'll have hard AI that can pass the turing test

And in 2045:
"$1000 buys a computer a billion times more intelligent than every human combined. This means that average and even low-end computers are vastly smarter than even highly intelligent, unenhanced humans."

Imagine playing civ on that

I don't think the future will be men and women riding in spaceships to explore the galaxy; I think machine intelligence will replace human intelligence as far as technological progress goes (probably not forcefully, ie Terminator). Exponential expansion of the humanity's technology means space may be conquered pretty uneventfully and pretty fast by self replicating intelligent robot probes.

As far as extraterrestial intelligence goes there are a few options:
1) It either doesn't exist locally
2) It's tech hasn't exploded exponentially, or they never choose to expand into space
3) Intelligent machines permeate the galaxy, possibly the universe, but remain hidden. They use communication techniques which are superior to radio or any EM signal.


This forum has deviated pretty far off topic from the creator's intention for the mod though. Perhaps it should be moved.

My suggestions from this discussion as related to Ascent of Mankind:

The tech race for AI. Honestly by the time we have AI in our world though, the nation state is going to be a nearly extinct concept; but maybe for civ purposes it isn't. Hard AI can either give you nearly unlimited research and production power...

Or it can backfire and spawn a "barbarian" robot legion in the city that built the AI. Included might be a few dozen fusion nova nukes. The civ AI targets all players' capital cities, and goes through the list 3-5 times, thus wiping out most of your important cities, causing massive upheaval if revolutions are turned on.

As for space travel and colonization, that can occur if you don't have a 'terminator event' or if you destroy the threat and move on.

Again though, should we move this discussion?

 

[Galacticat42]

The tech race for AI. Honestly by the time we have AI in our world though, the nation state is going to be a nearly extinct concept; but maybe for civ purposes it isn't. Hard AI can either give you nearly unlimited research and production power...

Well it seems this thread has already taken on a mind of it's own
Anyway classes are over and I'll be continuing work on the mod again. I love the discussions that have been going on here but I must ask what's FTL stand for? (the context in which it's used argues that it does not mean for the lose)

Or it can backfire and spawn a "barbarian" robot legion in the city that built the AI. Included might be a few dozen fusion nova nukes. The civ AI targets all players' capital cities, and goes through the list 3-5 times, thus wiping out most of your important cities, causing massive upheaval if revolutions are turned on.

I actually had something like this in mind before RoMiS was was born, it just so happened that Hydro suggested a future mod which gave me an excuse to do my idea which happened to get lost in all the mod clutter that ATOM has brought.

And in 2045:
"$1000 buys a computer a billion times more intelligent than every human combined. This means that average and even low-end computers are vastly smarter than even highly intelligent, unenhanced humans."

Did he take into account the inflation that has and will happen? I'd say $200,000 if we get that by 2029 which I doubt.

 

[Killtech]

And in 2045:
"$1000 buys a computer a billion times more intelligent than every human combined. This means that average and even low-end computers are vastly smarter than even highly intelligent, unenhanced humans."

i really doubt that. i'm pretty sure AI will be very dissatisfying when it comes to performance. most probably it won't be any better then we are. the basic problem is that if it's and AI or a natural brain the 'algorithm' says the same. and by experiment on oneself you can see that neural networks aren't so good for calculation. fastest method to calculate something with a brain is to 'know' all the partial results already and merge them together accordingly. that's how the genius mental arithmetics do it. so it's no calculation at all.

the only big advantage an AI will have compared to us is that it will not need a human-machine interface to have access to a computer. and probably that's the way it'll go for humans too: implanted computer hardware to compensate all the fields a neuronal network doesn't excel. i guess in the future you'll have an arithmetic calculus unit that you control via your thoughts - just like you can control you arm - and then you will make the same funny exact machine like statements like data... just because you can!

And i doubt AI will be like in films. it will be much more human... it will forget things, need 50 attempts before it learns to do something right, it will be able of frustration and idleness and it certainly won't be so machine like precise (except it uses a AI - computer interface so it doesn't need to calculate anything by itself).

Well it seems this thread has already taken on a mind of it's own
Anyway classes are over and I'll be continuing work on the mod again. I love the discussions that have been going on here but I must ask what's FTL stand for? (the context in which it's used argues that it does not mean for the lose)

Faster then Light travel. that's what i assumed.

 

[Randomness]

And i doubt AI will be like in films. it will be much more human... it will forget things, need 50 attempts before it learns to do something right, it will be able of frustration and idleness and it certainly won't be so machine like precise (except it uses a AI - computer interface so it doesn't need to calculate anything by itself).

The thought of computers needing to go to therapy because they tried to calculate pi comes to mind for some strange reasone ...

 

[Galacticat42]

In that case...

Then again... This mod also assumes FTL, which breaks a lot of rules too.

No this mod does not assume FTL travel, it merely advertises ways to travel circumventing the FTL problem such as using wormholes and/or folding space(FS is gunna be an easter egg type of tech since it's more sci-fi but also a great way to reduce maintenance).

And for the AI discussion goin' on here: AI is virtually impossible to achieve. You may get computers that mimic organisms, but not anything that is the real deal. Look at a prokaryotic cell for example; It has no brain it has no neural networks, it just moves with no means of being able to through man-made algorithms. All it has is DNA, cytoplasm and a membrane. The day man cracks the algorithm for a prokaryotic cell is the day man will be able to make true AI.

 

[Killtech]

And for the AI discussion goin' on here: AI is virtually impossible to achieve. You may get computers that mimic organisms, but not anything that is the real deal. Look at a prokaryotic cell for example; It has no brain it has no neural networks, it just moves with no means of being able to through man-made algorithms. All it has is DNA, cytoplasm and a membrane. The day man cracks the algorithm for a prokaryotic cell is the day man will be able to make true AI.

ouh... i doubt that. AI is just a completely other concept of an algorithm. it can't be compared to any linear program or something which is our problem with developing one: we have absolutely no experience with this so we have a hard time figuring out how it could work. and since we associate something complete different with the term AI then an AI really is this make programming one even harder. all we could do so far is reproduce our memories functionality which is what neuronal networks are. but as far as i know the development in this field nothing had been done to create a kind of controlling core which determines a kind of direction the AI should go, a kind of will.

i suppose first AI will be developed during our lifetime... but as i said, it won't be any similar to the science fiction visions of AIs.

 

[iontom]

Maybe I'm incorrect, but what I've gathered is that one of the principles of AI is going to be parallel processing. The reason a person can catch a baseball or shoot a basketball is because part of the brain's motor control is matching velocities and positions solving for the best placement. For a computer as we know it to do this it would require solving several differential equations. Likely with multiple solutions for which maximization can be had.

i really doubt that. i'm pretty sure AI will be very dissatisfying when it comes to performance. most probably it won't be any better then we are. the basic problem is that if it's and AI or a natural brain the 'algorithm' says the same. and by experiment on oneself you can see that neural networks aren't so good for calculation. fastest method to calculate something with a brain is to 'know' all the partial results already and merge them together accordingly. that's how the genius mental arithmetics do it. so it's no calculation at all.

Even if first gen hard AI lacks the sophistication that people have, and it probably will, I think the benefits lie in something else. Scalability, duplication, faster processing (nerve impulses move only around 100m/s). Then of course miniaturization and networking...

 

[wilson6907]

How much longer for this to be out guys?

 

[Hydromancerx]

Awhile.

 

[JosEPh_II]

Is it Dead, Jim?

What's your opinion Bones?

 

[os79]

Is it Dead, Jim?

What's your opinion Bones?

Captain and Doctor, please refrain from hasty assumptions. There may had been several possible logical conclusions other than a dead modmod. These are...

 

[Arakhor]

It's alive, Jim, but not as we know it.

 

[Galacticat42]

It's alive, Jim, but not as we know it.

Hate to necro this thread BUT: the above statement is very true, although it will not be out for CIV simply because my original estimates for timing and knowing myself was correct and CiV would be already out for a while by the time it would have been finished. On a related note, I must see how Afforess and/or Zappara carry RoM over to CiV and see how far they bite into the future techs before I can safely carry on my plans for a multi-planetary mod. Also, my first two objectives for CiV will be re-making ANM and supplying Afforess/Zappara for any necessary art/models/etc.

 

[NBAfan]

Hate to necro this thread BUT: the above statement is very true, although it will not be out for CIV simply because my original estimates for timing and knowing myself was correct and CiV would be already out for a while by the time it would have been finished. On a related note, I must see how Afforess and/or Zappara carry RoM over to CiV and see how far they bite into the future techs before I can safely carry on my plans for a multi-planetary mod. Also, my first two objectives for CiV will be re-making ANM and supplying Afforess/Zappara for any necessary art/models/etc.

Glad to hear that. One thing I would miss from Civ 5 is ANM.

 

[Hydromancerx]

Yes it will be much easier to make this if its ideas are included in RoM from the beginning. Because really space is one part that was majorly lacking in RoM/AND.

 

[Galacticat42]

Yes it will be much easier to make this if its ideas are included in RoM from the beginning. Because really space is one part that was majorly lacking in RoM/AND.

Yes, that and I need the technology tree to be premade so I can align all the necessary techs for this mod that vanilla ciV doesn't have.