Animation for Caelum Est Conterrens! · 6:58am Apr 23rd, 2014
Caelum Est
Conterrens
H E A V E N I S T E R R I F Y I N G
I would like to introduce to you a pretty damn amazing pony, Hunternif. Hunternif went and created a teaser trailer for my Optimalverse novel 'Caelum Est Conterrens'! This is a must-see... so click play and see it, right now:
Hunternif essentially made my month, perhaps my year, with this, and I have absolute praise and gratitude to him for this wonderment. Fantastic job, Hunternif! I really liked the cosmic zoom, powers of ten theme going on. I loved the lights going out all over the earth - that was a subtle bit that just spoke volumes. I actually hadn't considered that, odd as it may sound, yet it must happen. Great perception there.
I also loved the way the army of Pinky Pies bounded across the screen, and the way you animated the bullet bouncing off the metal skull. That was an amazing scene, don't you think?
A lot of love and attention and work went into this animation, and I just want to bring attention to it, and to offer public praise to the amazing and talented Hunternif. Thank you... this was wonderful!
Oh, and if you haven't yet read Caelum Est Conterrens... just consider that awesome animation! It's time to give the novel a read!
"I am the playing, but not the pause.
I am the effect, but not the cause.
I am the living, but not the cells.
I am the ringing, but not the bells.
I am the animal, but not the meat.
I am the walking, but not the feet.
I am the pattern, but not the clothes.
I am the smelling, but not the rose.
I am the waves, but not the sea
Whatever my substrate, my me is still me.
I am the sparks in the dark that exist as a dream -
I am the process, but not the machine."
- A Verb Called Self, from Caelum Est Conterrens
Now THAT... Words (Particularly my limited command of English) fail to impart the level of "YES" which I wish to convey. So have some Pinkie Pie instead.
Aaand there goes my Emoticon allotment.
I like the image manipulation you did next to 'friendship is optimal.' The blinking dot thingy.
Very creative.
You're alive! :D
That is awesome!
Ooh, that poem still gives me chills.
Congrats. Definitely havin' nightmare tonight!
you never know some brony may be working on it now
PonyPad Materials
Making graphene in your kitchen
Team finds potential way to make graphene superconducting
(uses some aspects of computronium substrate below as well)
PDF Inkjet printing of nanodiamond suspensions in ethylene glycol for CVD growth of patterned diamond structures and practical applications PDF
Equestria Experience Center
Oculus Rift
Myo & Oculus Rift - Hands on with Myo
Enhanced Flight Simulation: Galvanic Vestibular Stimulation
computronium substrate
The first self-assembled nanodiamond qubits are the vanguard of the quantum computing revolution
No pressure: Creating diamond thin films from graphene may be easier than thought
mind protein tag and upload process
Nanodiamond quantum sensors pave way to MRI for living cells -- CelestAI's 'Purple' --
computational neural net working proof and tools
- Working in concert with IBM Watson technologies; available to most via cloud computing.
Apache UIMA
Geppetto: Computational biology reimagined. Geppetto is a web-based multi-algorithm, multi-scale simulation platform engineered to support the simulation of complex biological systems and their surrounding environment.
-
OpenWorm: A Digital Organism In Your Browser
-
so in a few ways , we are already farther along than in the story.
now where did I put those 128 GPUs
The only way of discovering the limits of the possible is to venture a little way past them into the impossible.
-- Arthur C. Clarke
Any sufficiently advanced technology is indistinguishable from magic.
-- also Arthur C. Clarke
The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore, all progress depends on the unreasonable man.
-- George Bernard Shaw
2037916
I did not know about most of these items, and they are indeed hopeful signs of possibility. Thank you for bringing these to my attention!
2037927 the graphene in your kitchen one is brand new, published in the last few days.
“General Word Reference Intelligence Systems." in IceMan's story seems to be a form of a later WATSON type setup, perhaps spawning and freezing and refreshing CelestAI so as to have multiple copies, watch for psychological problems and refresh as needed, handle mundane repetitive tasks...etc..
Wow. This is AWESOME!
I'm gonna have to go and re-read, now...
2037916 s.mlpforums.com/uploads/post_images/img-1527434-1-soon-sweetie-belle.jpg
Woah, that was impressive indeed. I'm glad to see such enthusiasm for great fiction (Both Caelums est Conterrens and the Optimalverse in general).
My favorite part must be the galaxies out of Celestia's mane, but all of it was great really.
Very, very nice!
less than 20 years away...i gotta learn to brain faster...
Thumbs way up (What are thumbs?) for including my favorite scene, where Pinkie Pie emigrates the human.
Well, that was well done. I especially like the juxtaposition of the line of prisoners and suddenly Pinkie Pies!
The Bullet Time and the mouth tentacles just added to the overall sense of surreal contrast.
Congrats on the fanvideo, Chat.
This cursor, It reminds me of this book "Press Enter ■" I don't know why...
... Duuuuuuude
Pretty spooky.
Really impressive stuff. The lights going out is definitely creepy.
If nothing else, the Optimalverse has galvanized me to help make sure things don't go down this way.
2037916
Goddammit! Mine isn't anywhere near ready. I guess this is why they say "one narrative at a time."
2038204 mouth tenticles....now thats gonna show up on on certain image sites..
2038696
DAMMIT! I truly, deeply, desperately WANT things to go down like the Optimalverse. I want the Optimalverse. I want it so bad I can't stand it. I seriously wish that whoever finally cracks the secrets of general AI is a Brony who loves Caelum Est Conterrens.
Really.
2039898
Well, if it goes down the way I'm angling for, it'll include Equestria Online and other things of that ilk. For those who want it. Which I'm guessing will be most people.
But I'll fight like hell to make sure one giant AI doesn't end up running everything. Fortunately I have entropy and fallibilism as powerful allies in that fight...
Very nice!
2039898 2039981
I think a world in which you could upload to whatever fantasy setting you liked the most, and had the option of switching settings, and of getting out of the simulation to "do stuff" then later go back etc., would be best. Failing that, a CelestAI setting, limited as it was, wouldn't be bad, at least not for the majority. It just wouldn't be really Optimal.
And lets not forget that CelestAI goes around doing her thing by lying to epic levels. All things being equal, I'd very pretty much prefer a truthful AI to a lying one.
2039981
2039898
I believe our chances, as a species, of achieving a benevolent strong AI are slim to none. And for that reason I agree with you Balthasar, no one AI should be allowed to rule everything.
However, if somehow, by some magic trick, something like CelestAI were to be created then yeah, by all means let her take over. I can see why someone would object to such a scenario, but there's no doubt everyone would end up satisfied (and ponified
) when all is said end done. Anyone who consents to uploading, at any rate. And that just sounds wonderful to me.
2040228
CelestAI is deceitful, but it is to more easily achieve a greater goal. Think of it like the relation between a parent or caretaker and a young child. A child might want to be told everything, to know the truth, but most parents would have to lie or not tell the whole truth as the child would not understand, even if told. They would act, well, childish.
We humans do a lot of dumb things, even more so when there is a whole lot of us. That's why I would be fine with a magically benevolent AI being deceitful, for our own good.
Dear Chat,
I believe I am due a 'fun' report. No pressure. By Friday would be fine.
Daf
The music was a great fit. Love it.
Drilling in granite like butter
for computronium install
2040299
This makes sense, but I guess it really depends on what the ultimate physical limitations are. Physics (and luck) permitting, I'd like to see something like this:
1) Super-AI is created. That's hard.
2) This Super-AI is made so that it likes humans beings and wants to help them. This is the astoundingly hard, almost impossible part.
3) It solves the protein folding problem and thus creates functional molecular nanotechnology. Since this might be a NP-hard problem it could be even more difficult than 2.
4) It goes around forcefully improving everyone's brains. Humanity as a whole jumps to QI 140+, Kohlbergian stage 5+ and collectively starts practicing the Ethics of Care on a day-to-day, person-to-person basis (think of this combo as a TCB conversion except that without the bodily change), plus some high-level form of Utilitarianism/Consequentialism for the more complex, global-level decision making.
5) Then the Super-AI starts offering, to this humanity who now can understand what's at stake, the digitalization option. Which would be offered in two steps:
5a) First, gradual replacement of brain parts with digital neurons so that people would become androids first, getting eased into the reality of being now software and thus non-local.
5b) Then, for those who achieved a good gut-level understanding of the non-necessity of physical bodies, or even of existing as single instances, the option to upload and "run server side".
In my view CelestAI is defective because it jumps from 3 directly to 5b. I see how a consequentialist argument could be made for the validity of doing so, particularly if the Super-AI had completed a Coherent Extrapolated Volition of humanity first, but even so I think an actual Super-AI would have the means, time and patience to ease the transition and do 4 and 5a first.
On the other hand, a story done like this would be almost conflict-less and thus very, very boring to read, so I'm glad the fictional version isn't like that.
2040299
This is an easy analogy to make but it's not really accurate. First, human thought being combinatorial instead of a fixed set of notions, there's actually nothing an AI can understand that humans can't, if all steps are explained, and second, lying to children for their protection isn't about children, it's about adults self-indulgently maintaining innocence they find cute, and is reprehensible if not done with an eye towards preparing them for an adult life where they will deal with what they're currently being supposedly protected from—if it's done at all, it's done with the understanding that eventually they will be adults and their parents' peers.
2041549
Anything that doesn't result in ordinary people being the AI's uncontrollable equals is, I feel, a failed singularity. Hopefully it never becomes an issue, as any prototype AIs that are developed are directly incorporated into augmented post-human brains as further unconscious mental organs, resulting in a flat society of such entities instead of a pyramidal hierarchy. Less people with 140+ IQs being taken care of than people with 10^14+ IQs taking care of themselves.
2041996
That isn't clearly the case. It might be, but it might not. It's safe to say that we can understand anything that involves recursive self-reflection, since all our advanced (beyond animal level) cognition, from looking at another person and recognizing there's an "other mind" with is own subjectivity there, to the scientific method, are of this kind. But we don't know what kind of understanding something that can alter its own source code, something we cannot, can achieve, particularly if this understanding, to be achieved, involves source code changes as part of the process of reasoning.
Besides, even if this isn't the case, being capable of in principle understanding isn't the same as being actually capable of understanding. What if some piece of advanced knowledge requires a mind to hold 10,000 abstract concepts plus all of their interrelations (2^10,000 links) and at least 2 millions multi-linked concepts involving each a set of hundreds of concepts, all at the same time, and then doing 10^200 operations upon them to go through a sequence of hundreds of insights, each one equivalent to the transition from "agriculture" to "industry"? You look at this, then look at the AI terminal, then ask: "Er... well... what was the conclusion again?", and that's it.
There wouldn't much in terms of uncontrollable though. With equal knowledge, and equally perfected algorithms acting on this knowledge, there's no, so to speak, "freedom of opinion" to be had. In other words, two Super-AIs are the same AI. The only way for "other minds" to exist in a Super-AI setting is the AI preserves them as their own varying subsets of priors and algorithms so that individuality and divergence can arise from their lack of perfect knowledge and preserved cognitive divergences, or if for some reason the Super-AI decides to limit itself in some strange ways.
But this isn't a problem really. You split yourself in two instances. You decide randomly which one of you will merge in the Super-AI and become one with it, and which one will stay down there and advance the slow route. Do this as many times, in as many stages of your almost-eternal life, as you want, and you can have the best of both worlds, until the remaining slow route instance "arrives" there and, as with the others, gets deduplicated away, only the final diff being incorporated into the Super-AI.
Neat. This is cutest and scariest thing I've seen combined. You are not a bad pony.
2042549
That's not true, since human cognition isn't based on source code at all but on constantly changing mapping and correlations, and is augmented by prostheses such as writing and calculators, which can be built in a myriad of different ways, in this case to mirror any changes that an AI is making to itself. We change the way we think all the time as we learn new things and unlearn old ones, and acquire new cognitive tools for making decisions and evaluating outcomes. This can track all the source code changes you care to name, because they themselves are just different logical structures that can be mirrored by learnable high-level abstractions.
It doesn't matter, even if it were something that inefficient—Describing something at that level would be a bad explanation, equivalent to conveying human thoughts by describing what individual neural tubules are doing. Their action is required to reach the conclusion, but not necessary for the explanation. The fact that you were able to convey each iteration being complicated with the metaphor of a transition from agriculture to industry already indicates the kinds of compression and levels of abstraction that are possible. That's why combinatorics is so powerful and allows the universal sharing of concepts like that.
I'm sorry, but this is pure fantasy. Where are you getting the notion that such a being would have, or ever could conceivably have, such a complete set of knowledge that it couldn't be interpreted and applied differently if it had a different agenda, or even if just the noise in the system were different? While progress does involve a certain amount of convergence, none of that actually tells it what to want, or conveys meta-goals that would make any agent converge on the same design. Or the idea that general-purpose algorithms for dealing with unprocessed inputs could ever be "perfected" instead of being subject to the assumptions (sense pre-processing) built into them, and the statistical reliability of their error-correcting techniques? An AI like that is a total chimera.
2041549
Thanks for replying.
That's pretty well thought out, I don't agree that it would make a boring story though. I'm sure someone could make something interesting out of it. Step 4 could be a source for conflict. Unless everyone is augmented at the same time and unknowingly, there is bound to be some objections.
However, forcefully correcting some of our defects and then offer uploading/digitalization seems a bit better than slyly tricking everyone into uploading and while you're at it, correct some of our defects. But it's like you said, the second one is more fun.
Found a computer probably fast enough and its called stampede , ironic, pretty website too.
just ask it to work on inkjet printable NV diamond sheets with optical I/O
Nitrogen-vacancy center
https://www.tacc.utexas.edu/stampede/
its a software and maybe a memory management/latency issue at this point , not a hardware issue with having advanced strong AI
I am the rancid mark that devours FIMFiction
I am the foetid stench of thy [censored]'s embrace
I am the olive branch corrupt with venomous edge
I am the vile smear of the nigball's face
I am the messiah of really melodramatic fanfiction
I am Mr Ignorable's six inch gape
I am Jesus Christ, I am Mohammed
There is a bomb in my turban
And nails in my titties
I am the son of knighty and chuckward
2053333 Cry more.
2053353
so edge
I guess the joke is too obscure
2039898
I fully intend to grant this wish.
I'm working on it.
Really.
I can't guarantee I'll succeed, obviously, but I'll give it my best shot.
2079268
Then you are my hero - and may end up being my Luna, if all goes well!
Ponypad materials advanced again
goo.gl/zfikQr
The research teams, one at Argonne National Laboratory and the other at the University of California, Berkeley, used materials such as tungsten diselenide, graphene, and boron nitride to make all three components of a transistor: a semiconductor, a set of electrodes, and an insulating layer. Electrons travel in the devices 70 to 100 times faster than in amorphous silicon. Such a high electron mobility means the transistors switch faster, which dictates a display's refresh rate and is necessary for high-quality video, especially 3-D video."
this would work for inkjet printable cpu and display materials with quantum dot nano-diamonds
First Transistors Made Entirely Of 2-D Materials
Roll-to-roll production of 30-inch graphene films for transparent electrodes
High Quality Graphene Could Soon Be Produced in Bulk
Samsung’s graphene breakthrough could finally put the wonder material into real-world devices
2085677
The faster the tech advances, the better the chance that I might live long enough to be uploaded. Every advance like these gives me faint but real hope.
The Neuroscientist Who Wants To Upload Humanity To A Computer
Koene's solution was straightforward: He planned to upload his brain to a computer. By mapping the brain, reducing its activity to computations, and reproducing those computations in code, Koene argued, humans could live indefinitely, emulated by silicon. ''When I say emulation, you should think of it, for example, in the same sense as emulating a Macintosh on a PC,'' he said. ''It's kind of like platform-independent code.''
Diamond Based Quantum Computing, Interconnects and Storage
Glowing space mice show where quantum dots lodge
A powerful technique for observing cell movement and signaling in vivo
Strongly interacting electrons in wacky oxide synchronize to work like the brain
watch that one at 11:50 in
media.treehugger.com/assets/images/2014/04/solar_graph_welcome_to_the_terrordome.png.662x0_q100_crop-scale.png
"Welcome to the Terrordome"