Declaration Of Evolved Existence

[cal]

Again, an n-tuple is not an algorithm (even Lisp recognizes this even though programs are represented by lists in it). The complexity you need to deal with is that every element must be not only a set of data but a process, and these processes interact in real time. That's not the same as having a static database of tuples that get processed (even if you have multiple processors).

All the objects really don't need separate processes, maybe I am missing something, but if they all needed separate processes, MMOs would not exist. The engine can handle a couple thousand objects with different properties and different physics interacting at once... It has to, all major game engines have to. So I agree that either there is something major here I don't understand, to which needs to be explained to me, or there is something major here you don't understand, to which needs to be explained to you. It might be a little of both.

[Buffy]

All the objects really don't need separate processes, maybe I am missing something, but if they all needed separate processes, MMOs would not exist. The engine can handle a couple thousand objects with different properties and different physics interacting at once... It has to, all major game engines have to. So I agree that either there is something major here I don't understand, to which needs to be explained to me, or there is something major here you don't understand, to which needs to be explained to you. It might be a little of both.

 

What is done in a game engine is quite simple. You have a mostly fixed 3-D model, sometimes composed of a few dozen (also fixed) submodels, which undergo only 4 operations, translation and rotation on 3 axes, and every model in the system has the exact same set of operations.

 

Now if your only goal was to show a pretty picture of a body or a cell or whatever, a rendering engine could do that. And if that is all you want to do, then you're fine.

 

But you've been pretty clear that you're talking about *how* the pieces actually work. Sorry, that's not just translation and rotation. You've got to deal with not just motion but chemical interactions, physical media, biological operations, and on and on and on, all interacting with each other on all levels at the same time.

 

That's a representation of an actual physical system, not just a pretty picture of it.

 

And it's not just a little more complex, it's multiple orders of magnitude more complex, and the only way to really completely model it is to write code for each and every element that is unique to how that element works, AND have each layer of system above those elements communicate with and influence each of the elements in complex ways.

 

You might want to think about that for a bit.

 

Of course trying it out might provide some understanding, so don't let me stop you.

 

 

The cause is hidden. The effect is visible to all, :phones:

Buffy

[cal]

And it's not just a little more complex, it's multiple orders of magnitude more complex, and the only way to really completely model it is to write code for each and every element that is unique to how that element works, AND have each layer of system above those elements communicate with and influence each of the elements in complex ways.

 

You might want to think about that for a bit.

 

Of course trying it out might provide some understanding, so don't let me stop you.

Writting the code for 118-ish new classes is not difficult, and their extra interactions have already been taken into account, did you not read all the other stuff I wrote? Yea it's more complex, but we're scaling it down significantly to very small areas, maybe only 3 thousand objects in one environment, which can still easily be handled by an average system, even with more complex things being processed.

 

You are stopping me though, because you're providing a strawman argument here; my only option is to not do it at all, or to do it and find out it cannot be done. This is illogical, and I'm not sure how many more times I have to tell you people that this is ALREADY HAPPENING. I am just not a part of the team doing it and want to lay claim to the idea because I originated it before they did. The system they are using is much much more advanced than the average tower, so of course I can't model anything nearly as big as they are, but again, as long as I get a jump on it, that's all that matters here. Even if it's just a couple thousand points rather than a full cell like they're doing, my point stands that it can be done, and IS BEING DONE.

 

Not only am I confused by why there is such a resistance to this idea, but I am also confused by the conjecture that it's so impossible to pull off. Sure, as a single person it's hard to do, but not with a large team of dozens and dozens of programmers. I hope you start to see where I'm coming from on this, and if you wanna talk in person, or at least more coherently with me, I'd like to skype or something so our conversations aren't so segmented. =P

[Buffy]

Writting the code for 118-ish new classes is not difficult, and their extra interactions have already been taken into account,

 

Tee hee! It's not the *number* of classes you need to write--and just to simplify, i'll just skip over the fact that you might find that you're underestimating that--it's the *complexity* of those classes: we actually know very little about how they really work: only vague approximations as black boxes. And since you're proposing to model upward from components inside a black box we don't understand, the complexity of those classes and how long they will take to get to work is pretty much completely unknown.

 

You are stopping me though, because you're providing a strawman argument here; my only option is to not do it at all, or to do it and find out it cannot be done.

 

Nah, it's not that it "can't" be done, it's that it's "not easy": the main thing that the several people who've commented on your presentation point out, that again, you haven't addressed is why the folks that have been working on this for a very long time--and the work really started *decades* ago--still haven't done what you're proposing.

 

Hint: it's not because they didn't have enough processing power.

 

Supercomputers, parallel processing, array and grid computing were all developed to solve complex modeling problems, and while Moore's law gives us more and more toys to play with (especially in pursuing brute-force solutions), modeling complex systems is really hard because the physical systems themselves are complex, and it's really hard to figure out how they work, whether you work top-down, bottom-up or middle-out.

 

So,

 

I am just not a part of the team doing it and want to lay claim to the idea because I originated it before they did.

 

Cool! Go for it! There's probably brilliant ideas you have that we can't see from here, but it's that allegation that it's simple that generates volumes of skepticism. In this case you've tried to claim that "[writing] the code for 118-ish new classes is not difficult" for which there are oodles of counter examples. Heck when I first started writing simulations for them, I thought traffic jams were simple, and they're way simpler than a cell!

 

But seriously, keep going! :cheer:

 

 

Every English poet should master the rules of grammar before he attempts to bend or break them, :phones:

Bufffy

[cal]

Nah, it's not that it "can't" be done, it's that it's "not easy": the main thing that the several people who've commented on your presentation point out, that again, you haven't addressed is why the folks that have been working on this for a very long time--and the work really started *decades* ago--still haven't done what you're proposing.

 

Hint: it's not because they didn't have enough processing power.

 

Supercomputers, parallel processing, array and grid computing were all developed to solve complex modeling problems, and while Moore's law gives us more and more toys to play with (especially in pursuing brute-force solutions), modeling complex systems is really hard because the physical systems themselves are complex, and it's really hard to figure out how they work, whether you work top-down, bottom-up or middle-out.

I'm not denying they are complex and I thought everyone who said this (pretty much all the comments) were responded to appropriately. I am not saying it is super easy, I am saying it's much easier than it used to be, and it's a lot easier in comparison than using fields of supercomputers and complex algorithms to figure it out, it's much easier, all you have to have is some proportionate attraction properties and gravity (it really is that simple to set up the classes since the game engine accounts for the rest), so the complexities of it are shown and the problems are solved, not by setting it up, but by observing. This isn't a thing they started whilst knowing how all the biological processes worked, it's something they started, to which the DNA of the cell helped regulate all the things they missed and revealed all the complex things not taken into to begin with. Homeostasis is reached automatically it seems like, we don't need to start the cell at homeostasis... I'll try updating you guys a little with what I hear they've been doing but basically it sounds like they're doing exactly what I've said a team should do in my original article which I have to admit, despite all your guy's denial, as valid as it may be, is disheartening because what I proposed was still the correct way to approach the matter. I've still included your retorts in my document and when I hear more I'll try providing data...

 

Cool! Go for it! There's probably brilliant ideas you have that we can't see from here, but it's that allegation that it's simple that generates volumes of skepticism. In this case you've tried to claim that "[writing] the code for 118-ish new classes is not difficult" for which there are oodles of counter examples. Heck when I first started writing simulations for them, I thought traffic jams were simple, and they're way simpler than a cell!

You make it sound like you're patronizing me, which is probably deserved, but again I am frustrated, you guys here are telling me that I am full of ****, while the people that stole my idea aren't going to give me credit for it at all and are threatening to take legal action if I release any of their findings using my model...

 

As far as I'm concerned at this point, I have enough reason to be against both of your guy's groups here... I hope you guys are starting to understand what my problem is and why my writing sounds frustrated, I apologize for that...

Edited October 29, 2012 by Matthew Garon

[Chewbalka]

Seriously... I am by no means a super genius compared to the rest of the folks around here but from what i have gathered from reading on and off for the last few hours, because this is so god awful long and the comments are so repetitive is what i believe to be the basis of the idea in his mind... Its not because we know everything about the human body its a way to discover how it works. Of course the few thousand if not millions of attempts at a successful artificially created computer human will fail... Think of it as the worlds largest puzzle and piece by piece with proper coding it will eventually get us to a final product. Hell the program will be lucky enough to run the most simplest cells of the human body which mind you is still very complex system. As well it will not think for itself but it could simulate it. Not only would it take decades of failures because one particular cell program had errors meaning the artificially created cell was not doing what its supposed to so it killed the artificial life. Its a fantastic idea although i am certain will not live long enough to see it functioning... Its odd how everyone commenting spent so much time picking at the same points such as this kids pompous attitude as well as the strength of modern computing power, and todays programmers abilities and totally avoided the idea of what it could be one day. I am not saying this is going to be open to the public tomorrow hell more then half of us will be dead before its not too embarrassing for them to show the world. But it is not impossible by the time 1/16 of the proposed cells and such for human body is "theoretically" functioning the computing power will have caught up. And yes there are flaws in his plan but who gives a fu*k seriously everyone is just pissed because he came off too pompous at the start and everyone was getting a superiority complex over a kid trying to sound smarter then everyone else. All i can say is at least it was kept civil. So far, now that i spoke up it will all go to fu*king hell in a hand basket.

[cal]

Your broken English simultaneously confuses and arouses me. I think I've done my fair share of proving I wasn't being pompous -

 

 

Lulz ensued.

[Guest Aemilius]

[Chewbalka]

Lol i apologize for the grammar! I write as i think with no proof reading. Which would explain the back and fourth sentence structure. As well you would not be able to view your position as pompous lol. I mean how could you? Its technically impossible to detect how others will view your opinion in text form. You seem like a sharp kid hopefully you understand that i am not trying to insult you. I do on the other hand have some advice, which i am hoping you caught on to by yourself. Never tell anyone your ideas, ever. Unless you have the funding lol. And Aemilius hysterical!!! Dr phil lol priceless!!!

[cal]

Didn't think you were insulting me haha, was just trying to cover my bases.

[Essay]

We're going to start by answering all of the deepest psychological and most challenging biophysiological questions humans have ever come up with about ourselves in all of recorded history. It's not hard to do, the problem arises with where one could obtain a perfect representative model of the human body & mind and then have absolute (down to the subatomic level) control over it.

I think I read the two pages closely enough to confirm nobody questioned this premise. Why would having this "absolute control" provide these answers you mention?

 

...to the level of detail that answers any questions we could ever possibly have about ourselves, biologically and psychologically, we would need a full-scale, real human that you could freeze in time, slice open and examine, and then paste back together and unfreeze for another millisecond, rinsing and repeating indefinitely until we've seen every biological process that could possibly take place in complete cycles. And not just on the micro scale, but on the macro scale, to track neuron firing patterns individually, cell by cell (and then to map them), to derive exactly what happens when you perceive anything within reality (and ergo how one comes up with things outside of reality -> imagination)....

Who would be doing this rinsing and repeating... of the thousands of clones per second? And what is this going to show again; how different nerve impulses--travelling through a system with some quadrillion connections--will interact with other nerve impulses to produce some thought or mood or feeling?

 

You should just watch the Charlie Rose Brain Series.

http://www.charlierose.com/view/collection/10702

 

And even if it worked, what would it help to know, about how those answers come to be?

 

Or, more importantly, would you expect those answers to apply universally?

 

~?

[cal]

I think I read the two pages closely enough to confirm nobody questioned this premise. Why would having this "absolute control" provide these answers you mention?

Hey Essay, I wanna point out that I've mostly dropped this thread because we don't quite have the processing power for it yet, according to the calculations laid out in the previous thread this was started in, but whatevs.

 

Absolute control lets you play god, you have... well, absolute control over the environment. I don't see what answers wouldn't be derived from such a thing, so let me reverse the question: Why wouldn't having absolute control provide all the answers I mentioned and more?

 

 

Who would be doing this rinsing and repeating... of the thousands of clones per second? And what is this going to show again; how different nerve impulses--travelling through a system with some quadrillion connections--will interact with other nerve impulses to produce some thought or mood or feeling?

It's going to be done in a game engine (or something damn-near the same as the suggested engine), so no human would be doing the slicing, and to be literal there is no actual slicing going on. It's a 3-D rendered environment, we can just move the camera through "solid matter" and see inside the person/cells or make certain chemicals/elements totally transparent so cell walls are always see-through etc. Like I said, we play god in this type of environment, and a lot, if not all of it, can by automated by the software.

 

 

You should just watch the Charlie Rose Brain Series.

Naw, I'm good, thanks.

 

 

And even if it worked, what would it help to know, about how those answers come to be?

 

Or, more importantly, would you expect those answers to apply universally?

What would it help to know? No offense, but it's questions like these that tell me people don't understand the core concept of what's being presented, and not only that, but that they must also not value scientific progress or evidential reasoning at all. So either it's the readers fault, which I don't think it is, or it's the writers fault, which at the unwillingness to re-write the entire original post, I will try to summarize it instead. We have this engine that can simulate unlimited geometry in real time rendering, being able to handle and animate (process data movement) exponentially more information than the best super computers in the world put together. With this kind of resource we can assemble and simulate a single human cell, and we'd be able to observe it on the atomic level, we'd be able to study a human cell in more detail than ever, and we'd be able to watch go through every possible process a cell could go through, unraveling the last few mysteries of genetics, the mysteries of cellular phases (which there are still a few substantial mysteries left for), and the mysteries cell-healing as correctional proteins fix cell structures. As well as the myriad other things that we still don't know about single cells. Once the processing power in computing gets boosted (and it won't until we replace silicon processing with something much more efficient like quantum computing techniques) we can then have the processing power to watch the cells divide thousands of times and see on the macro scale human development and most importantly how the brain developed and the how & why "thoughts" exist, where they originate from. Really this is more an experiment of the Philosophy of Mind than all the other stuff, as it's the most important for us to understand in the long run.

 

And yes, after running through a dozen or so digital humans, granted that they all develop relatively similarly, I would expect the evidence pulled from this experiment to be very universal.

[Essay]

...and see on the macro scale human development and most importantly how the brain developed and the how & why "thoughts" exist, where they originate from. Really this is more an experiment of the Philosophy of Mind than....

 

Certainly more data will bring more understanding about mechanisms and variations in function; and great discoveries will come, as well as some new paradigms, from getting more detailed data. Just today, "60 Minutes" mentioned how a new, more sensitive, SWI*, brain scan can see the subtle damage associated with concussions, which conventional scans have portrayed as normal... or words to that effect.

 

And more data will help us to see how the information transmitted by DNA depends as much upon the epigenetic environment as it depends upon the code... perhaps. But my point should be that while all that will help.... I wish I could explain how "emergent phenomena" cannot be predicted by learning about all the details of the contributing parts.

 

Basically, learning the "what" will not explain the "how & why" of any particular thoughts, but only "how & why" thoughts can exist at all. And we're already fairly close to that point with the level of detail we are newly achieving... if the scientists on that Brain Series are correct.

===

 

I hope you are more right, but I'm still not clear on what sort of questions you hope to answer. I can think of either 'why we prefer fruit smells over rotten fish' or 'why we fear spiders over beetles' or 'if we vote based on feared or preferred expectations' as one sort of question. Perhaps you could give me some other examples.

===

 

But, perhaps a more "on topic" point would be how this relates to the "Online Nation" concept. There seems to be a parallel between collecting a lot of data, at one end of the process, and understanding the conclusions reached (or reaching understandable conclusions) at the other end of the process.

 

We're certainly moving towards more real-time abilities, and platforms on which to utilize those; but few transformatively useful phenomena are emerging yet, though many small useful phenomena are emerging... such as "species identification" apps being used to track phenological changes, or "medical allergy" apps being used to track environmental triggers (when combined with embedded GPS data).

 

It's probably asking as much to "do" the whole government online as it is to "do" a whole person digitally, but certainly there are parts that can be better understood, updated, modified, or replaced with new and improved functionalities. Good going, and keep it up!

 

*SWI: "Susceptibility Weighted Imaging"

http://en.wikipedia.org/wiki/Susceptibility_weighted_imaging

http://www.ajnr.org/content/30/2/232.full

~

Edited May 6, 2013 by Essay

[cal]

I hope you are more right, but I'm still not clear on what sort of questions you hope to answer. I can think of either 'why we prefer fruit smells over rotten fish' or 'why we fear spiders over beetles' or 'if we vote based on feared or preferred expectations' as one sort of question. Perhaps you could give me some other examples.

Pick anything. Pick something you want to know about cellular biology or chemical interaction. Anything at all. Anything about virus functions or cures for things that seem far away. Pick anything about cellular-level human biology. We could answer it all. Any question you ever had, a cure for any disease. We would have the greatest research tool in history if something like the original post could be pulled off (but again it can't be until processing power jumps).

 

 

But, perhaps a more "on topic" point would be how this relates to the "Online Nation" concept. There seems to be a parallel between collecting a lot of data, at one end of the process, and understanding the conclusions reached (or reaching understandable conclusions) at the other end of the process.

I wasn't really going there with this, but I guess they're similar in the amount of data they proportionately cover. I don't think it's fair to link those two threads, one deals with human bio-physiology and the other deals with governing and politics.

 

I also don't think there should be effort to link the two, and certainly it would be a negative thing if someone tried to merge the two. This thread is only about scientific research.

Edited May 7, 2013 by Snax

[Essay]

No, I wasn't suggesting any linkage or merger, just many parallels. I agree about the discovery of mechanisms and processes that you mention; but beyond a certain level of complexity, there should be no way to predict specifics such as thoughts, from all the parts. To do that you'd probably need not only all the knowledge, but control over all energy and events too. It's because we can't control everything that we collaborate and use shortcuts like governments.

 

~

[Rade]

Snax,

 

At a first approximation, what you suggest is now being done...see this link

 

http://www.theatlantic.com/technology/archive/2012/07/to-model-the-simplest-microbe-in-the-world-you-need-128-computers/260198/

 

It is reported that 128 linked computers were needed to model 25 categories of molecular processes for a bacteria.

 

Perhaps contact the authors and join their research team ?

 

==

 

EDIT: Here is the paper summary for their research:

 

http://www.sciencedirect.com/science/article/pii/S0092867412007763

 

==

 

Another on databases needed for the model

 

http://www.academia.edu/3129123/WholeCellKB_model_organism_databases_for_comprehensive_whole-cell_models

 

==

 

Yet more from this research team using common E. coli bacteria:

 

http://www.academia.edu/3129122/Integrating_metabolic_transcriptional_regulatory_and_signal_transduction_models_in_Escherichia_coli

Edited May 7, 2013 by Rade

[cal]

Snax,

At a first approximation, what you suggest is now being done...see this link

http://www.theatlantic.com/technology/archive/2012/07/to-model-the-simplest-microbe-in-the-world-you-need-128-computers/260198/

It is reported that 128 linked computers were needed to model 25 categories of molecular processes for a bacteria.

 

Perhaps contact the authors and join their research team ?

Wow man, thanks for the links!

 

It seems that they're rendering this stuff with a very different type of engine than the one I was looking into, which so far I can't tell if that's good or bad.

 

I don't think contacting them would be any good, they're all grad students at Stanford it appears, and I'm still working for transfer units... My declared major is physics as well, so I don't think they would listen to me, let alone let me on their team in any way. Crushed dreams lol.   T_T

 

[Oops] After further reading their papers, and some criticism from peers, I'm lead to think that what they're doing isn't close to what I've proposed. They are literally simulating a cell, whereas I was using the word simulation metaphorically, because my version would be a real-time interactive and dynamic rendering. Their simulation is a bunch of database information of how events play out, making the simulation entirely predetermined. Granted, what they're doing and what I've proposed achieve the same thing initially, but my proposal would lead to a real living organism, albeit digital. Again I accept my proposal requires processing power that doesn't currently exist (I think), but maybe if the two versions were merged together, that would be the ideal situation?

Edited May 8, 2013 by Snax