Posts: 42
Threads: 10
Joined: Oct 2016
So i'm writing a story, and i'm not sure about what kind of computronium certain brains would use...
So let's say one of the really powerful archai wants to create the fastest brain possible and it has to be below the first singularity. Basically, if a major god decides that it wants to create the smartest, fastest superbright/superturing it can without making the thing massive, plus keeping it below the first singularity level, what kind of computronium/tech would be used?
I was sort of confused, the site says that those of a lower toposcopic level cannot use/understand tech created by a higher toposcopic entity, that part i get. But if a S6 or S5 archialect wants to give/teach a lower entity how to install their technology into their brains to be smarter, is it possible?
I ask this question because in the story i'm making one of the characters is a Proxav for Farview, and i'm not sure about how it would function...
Posts: 11,450
Threads: 450
Joined: Apr 2013
Hi! Welcome to OA!
It's not really true that modosophonts can't use technology created by transapients and archai; there is a whole class of transapientech that is specially designed to be used by modosophonts, even though they can't understand it. See the article on clarketech.
http://www.orionsarm.com/eg-article/4b0d5fc969b5f
Some kinds of transapientech can be recreated by modosophonts; this is called Ultratech, and the assumption here is that the modos would probably have invented this themselves, but the transapients got there first.
http://www.orionsarm.com/eg-topic/4b0d54e136586
As far as the largest and fastest superturing possible, the sky is the limit; an archai could make a planet-sized brain full of plasma processors, but they could place restrictions on the toposophic level of the mind within that computronium. The mind could be prevented from even thinking about transcending, or mentally incapable of doing so; it might even be rigged so it would shut down or explode.
What would a modosophont moon-brain be like? Well, one definition of a transapient is that it can think consciously about all its data, inputs and memories, as a single consciousness or as an arbitrary number of subordinate consciousnesses. A modosophont moonbrain could not do that. I would imagine a very large modosophont to be less than the sum of its parts, always cross-checking between different segments of its mind and getting confused or conflicted. In short, it would be like a civilisation, or perhaps several, always arguing with itself.
A modosophont superturing could be much smaller and still be prone to this sort of self-division and paralysis; we call this the Bloatware syndrome, or King Gnuff's Curse.
http://www.orionsarm.com/eg-article/460e7c6d840e5
Posts: 2,110
Threads: 129
Joined: Sep 2012
(10-18-2016, 12:26 AM)stevebowers Wrote: Hi! Welcome to OA!
It's not really true that modosophonts can't use technology created by transapients and archai; there is a whole class of transapientech that is specially designed to be used by modosophonts, even though they can't understand it. See the article on clarketech.
http://www.orionsarm.com/eg-article/4b0d5fc969b5f
Some kinds of transapientech can be recreated by modosophonts; this is called Ultratech, and the assumption here is that the modos would probably have invented this themselves, but the transapients got there first.
http://www.orionsarm.com/eg-topic/4b0d54e136586
There's also this article on how ordinary sophonts relate to transpient-invented technologies:
http://www.orionsarm.com/eg-article/557846d4b952f
Stephen
Posts: 15,963
Threads: 730
Joined: Sep 2012
I think what's being asked here is how small a package an archailect could fit a superbright/superturing into rather than how big a processor would be needed.
Assuming that it is the case, please see my answer below...
(10-17-2016, 11:53 PM)Qualitum Dragon Wrote: So i'm writing a story, and i'm not sure about what kind of computronium certain brains would use...
So let's say one of the really powerful archai wants to create the fastest brain possible and it has to be below the first singularity. Basically, if a major god decides that it wants to create the smartest, fastest superbright/superturing it can without making the thing massive, plus keeping it below the first singularity level, what kind of computronium/tech would be used?
It would somewhat depend on the environment the created mind was expected to operate in. Currently we have S4 and higher minds able to make 'plasma processors' that make use of superhot plasma to perform computation. This is very fast, but tends to be a bit hard on the local environment (unless that environment is the inside of a star) unless shielding of some kind is used. Similarly, we've indicated that the archai can make some kind of processor using ultra dense matter (magmatter), but this would generally operate at temperatures and energies where gamma rays would be the waste output - bring your sunblock 7 zillion :p.
If we're talking a human comfortable environment (more or less) than the entity would probably run on something like an advanced form of Ultimate Chip.
Note that a single UC is about 488 times as powerful as a human level mind in the setting (and about the size of a postage stamp). So an array of 32 or so Ultimate Chips could run a First Singularity mind.
S1 minds are about 15,000x as powerful as a human level mind. We've never really quantified how powerful a superturing/superbright mind is in comparison to a baseline human, but usually talk in terms of a superbright being something like a Leonardo da Vinci combined with all the abilities of an idiot savant with none of the limitations. The page on Superior Genemods also gives some indications.
In a nutshell, while superbrights/superturings are certainly smarter than a baseline human, they aren't going to be thousands, or even hundreds of times smarter, most likely.
So you could probably run your characters mind on a single Ultimate Chip (1cm x 1cm x 1mm) with capacity to spare - or make them a heavily augmented character on top of their 'superbright/superturing' level of intelligence.
(10-17-2016, 11:53 PM)Qualitum Dragon Wrote: I was sort of confused, the site says that those of a lower toposcopic level cannot use/understand tech created by a higher toposcopic entity, that part i get. But if a S6 or S5 archialect wants to give/teach a lower entity how to install their technology into their brains to be smarter, is it possible?
While the tech of a given S-level is not generally operable by a lower S-level, it is possible for the higher S-level mind to create specialized devices that the lower S can operate, even if they don't understand how they work.
It is also possible for a transapient or archai (or even a modosophont) to make a device that can just operate itself to take care of the whole implantation/installation thing. Something like telling the lower S user 'put this under your tongue and suck on it until it dissolves'. This introduces transapientech constructors into their system which proceed to construct the necessary device inside them by scavenging materials out of their bloodstream or the like. They aren't really 'installing it' but it's taking care of taht job for them.
(10-17-2016, 11:53 PM)Qualitum Dragon Wrote: I ask this question because in the story i'm making one of the characters is a Proxav for Farview, and i'm not sure about how it would function...
I look forward to reading it
Hope this helps,
Todd
Posts: 42
Threads: 10
Joined: Oct 2016
10-18-2016, 06:36 AM
(This post was last modified: 10-18-2016, 06:56 AM by Qualitum Dragon.)
Thanks, that helps a lot. just one other question though...
The article about gnuff's curse, its meaning is that a hyperturing is better off with less brute force intelligence and better organization, opposed to just adding on more computing power, right? Just want to understand.
The story is about a ToPoGo tournament, where each of the great hexadecimal archai create one champion to compete in the tournament. The rules state that each champion can only take up a few meters in unaltered time-space(no wormholes), and that each champion must not meet or exceed the first singularity. The main character is the proxav representing Red Star M'pire, but he's at a bit of a disadvantage considering that everyone else was designed by a S6, while Farview is only a S5.
Thanks, and i hope you like it
Posts: 11,450
Threads: 450
Joined: Apr 2013
Brute-force data processing is always useful, but once a certain level of complexity is passed it is better to have better organisation than more processing capacity.
Posts: 15,963
Threads: 730
Joined: Sep 2012
(10-18-2016, 06:36 AM)Qualitum Dragon Wrote: Thanks, that helps a lot. just one other question though...
The article about gnuff's curse, its meaning is that a hyperturing is better off with less brute force intelligence and better organization, opposed to just adding on more computing power, right? Just want to understand.
Essentially yes. OA aims to avoid the overly simplistic (in our view) approach to achieving superhuman intelligence just by adding more and more processing power and nothing else.
This is an editorial choice grounded in various factors:
a) If becoming transapient is too easy then everybody does it and we rapidly end up without a setting to play around in and write stories about. In that sense, this limitation is something like Vinge's Zones of Thought.
b) It seems more realistic to us that 'merely' adding more processing power is not an automatic be all and end all for any problem. For one thing you have to coordinate all that processing power so it doesn't just metaphorically (or literally) run around in circles chasing its own tail. Also, things in the real world rarely work out to be that simple. There are always costs and downsides and complications. To use an example: If you strap a jet engine to the roof of your care with the intent of making it go faster (and do nothing else) - the reality is that that is not the only effect you are going to get - and the odds are very good that many of the effects you get are going to be negative ones.
Becoming transapient means not only adding the processing power, but learning/figuring out how to use it most effectively.
To use another example - if you obsessively lift weights and take every performance enhancing drug in the book you may indeed get very big muscles and become very strong - but you will also experience a number of downsides and will likely still get your butt kicked if you get into a fight with a martial arts master (or a skinny little person armed with a gun) - even if they don't have the same brute power as you do.
Note that transapients can augment themselves to have more capability than a 'baseline' transapient of their S-level - but what their version of augmentation looks like might be very different from what modosophonts engage in (we haven't really explored that aspect of the setting much). And they would still need to exercise care when augmenting themselves to ensure they don't bog themselves down or suffer other side effects from the augmentations not being properly coordinated and balanced as they are added into their mind and physical structure.
(10-18-2016, 06:36 AM)Qualitum Dragon Wrote: The story is about a ToPoGo tournament, where each of the great hexadecimal archai create one champion to compete in the tournament. The rules state that each champion can only take up a few meters in unaltered time-space(no wormholes), and that each champion must not meet or exceed the first singularity. The main character is the proxav representing Red Star M'pire, but he's at a bit of a disadvantage considering that everyone else was designed by a S6, while Farview is only a S5.
Thanks, and i hope you like it
Sounds interesting
Todd
Posts: 7,299
Threads: 295
Joined: Jan 2013
(10-18-2016, 09:04 AM)Drashner1 Wrote: Essentially yes. OA aims to avoid the overly simplistic (in our view) approach to achieving superhuman intelligence just by adding more and more processing power and nothing else.
This is an editorial choice grounded in various factors...
To add a point C), from what we observe about differences in intelligence and consciousness IRL it's very clear that organisation is just as important, if not more, than processing mass. Elephants have brains ~3x larger than humans but are less intelligent. Various primates have brains nearly identical to humans but again, aren't as intelligent. Hell there are humans born that are very mentally disabled despite having brains that are just very, very slightly different.
OA Wish list:
- DNI
- Internal medical system
- A dormbot, because domestic chores suck!
|