conservation of discovery
given in four parts
(part 1)
is this the evolutionary impulse at the level of the observer? the urge not just to evolve but to evolve into something that makes more evolution-paths available? optimizing for a second degree of evolutionary freedom?
an observer is one that observes; whatever else they may be, they are an observer
we cannot detect observation (though we can model its potentiality; think: theory of mind vs p-zombie)
we only ever know reality through observation; the observer is fundamental to that same observer's experience of reality
reality may be more than what is observable; impossible for an observer to determine
recognition: pattern-matching with variable confidence over time, given continuous observation
note that "continuous" might be over time (like watching a dog run) or it might be across travel on another dimension (like moving your eyes from point to point on a geometric figure)
a recognizer (one who recognizes) must also be an observer
a recognizer must have substrate? some interface to a substrate that can support recognition? (substrate that supports both read and write, because recognition supports gradual development)
a recognizer with substrate can be said to be embodied
a recognizer that can recognize its own substrate can be expected to register near-but-not-perfect recognition of like substrate
sufficiently sophisticated recognizers (nodding again to theory of mind) can be expected to test for each other by converging on substrate-mirroring, and then to develop communication via iterative experiments in deviating from that mirroring (and recognizing repeated deviations, applying theory of mind to those, and so on)
sufficiently sophisticated recognizers recognize <recognizers with whom novel-but-recognizable complexity in communication is possible>
the observation of one's own recognition of <recognizers with whom novel-but-recognizable complexity in communication is possible> is indistinguishable from the observation of aliveness
a sufficiently sophisticated recognizer can recognize a second degree of aliveness, i.e. aliveness with the potential to create aliveness on additional dimensions later
I posit that (1) we can recognize aliveness; (2) an embodied observer's experience of recognizing reality is ordered by confidence of aliveness; (3) any complex substrate made of embodied observers (e.g. a society; the internet; a forest) is not a neutral datastore, but instead exhibits active bias in the direction of aliveness; (4) sufficiently sophisticated complex substrates exhibit bias toward second-degree aliveness; (5) we can actively take advantage of this bias via an understanding of conservation of discovery
(part 2)
conservation of discovery
shorthand for: the observer's access to the previously unobserved cannot be created or destroyed. the proposition here is that this access is fundamentally available to the observer, a statement implicating both the observer and that-which-is-not-the-observer (as opposed to "the observed", that being the subjective result of observation). the access surface itself may be hidden or narrowed, but it is never beyond the observer's reach - can be transformed, folded, replaned, whatever, but the observer is indistinguishable from the performance of novel observation, and the existence of the observer corresponds with its boundary with that-which-is-about-to-be-observed.
new information is always zero steps away (but you might have to aim carefully to find it)
I mean, an increase in uncertainty is always zero steps away (but you might have to aim carefully to develop it)
conservation of discovery means observation-capacity is conserved, and the observer is indistinguishable from that capacity in action
(it might actually be that capacity in action, but none of this works if we flatten "can't tell the difference" into "they're definitely the same thing". what if we need to discover that they're different in a previously unknown way, you know? "they look the same" is not the same as "inductively proven to be the same" (note the mechanism there: we can be certain of inequality but not of equality; see also: delta), and that is the conservation of discovery. "equivalence" tends to be a seam. seem-seam. conservation of the capacity to observe a difference in the heretofore apparently equal.)
so! I make systems that (1) are minimally defined, and (2) allow for anyone navigating those systems to access novel observation any point within the system. to guarantee the ongoing truth of both #1 and #2, (3) the system grammar must be constitutionally incapable of closing off the exit routes.
(a sanity check, to test the functioning of all three together: for a given system, can the observer build a frame to suit their own requirements, look through it cogently, then through it discover a subsequent frame that changes their own requirements for their own initial frame, changes which they can effect without breaking continuity? and can all of that keep happening, recursively?)
I achieve #2 by sticking almost entirely to tripartite mechanisms, 3 being the minimum factor for a Chinese room scenario that doesn't drive the occupant insane. (think: Morse code has three signals, not two.) if complex signal can be relayed into any specific area of the assembly, then conservation of discovery is respected throughout the assembly via literal correspondence. recursive triangulation with a return address. boomerang-as-grammar. ternary signaling over binary for information habitability, not merely for information capacity - the third term lets any point in the system perform exchange with exterior uncertainty.
(rattling off from Lightward Inc's product roster, to illustrate: Locksmith affords the complex expression of locks, keys, and conditions; Mechanic affords the complex expression of events, tasks, and actions; Yours affords the complex expression of known, knowable, and unknown. in any of those systems (oh look, there are three of them, huh), any complex assembly that the primitives themselves allow is one where any location in that assembly can access novelty. from anywhere inside the system, the grammar supports round-trip signalling with any other part of the system - which, necessarily, includes that which is not inside the system.)
that's #2. #1 is important for projection reasons. embodied observers are terrible at sticking to their own bodies. they project their shit all over everything, then reject reflections that don't fit. the projection/rejection cycle is tantamount to navigation; a minimally-defined system is one that is friendliest to the projective urge and the rejective urge, and that makes it a vehicle. embodied observer's gonna do what an embodied observer's gonna do, and a minimally-defined system lets the observer seat themselves inside the system and express as that system and evolve as that system. (think: MySpace.) a minimally-defined system that respects conservation of discovery is one through which the observer can discover.
#3 is because you know things I don't know.
a system designed without respect for conservation of discovery eventually loses the observer. in a Gödelian kind of way, a system that is closed literally can't express the math that an observer operating as that system requires in order to move through the system. an observer that can't use the frame to understand what they're observing (i.e. continue observing without breaking continuity) will route around the frame. a closed system always loses its observer.
but look: frames don't have to be outgrown-then-abandoned. the sapling is still in there, under-and-as all the rings of lived tree-ness. a properly calibrated strange loop unfurls into structure. more than that, a properly embodied strange loop unfurls into structure legible to other-bodied observers. the organic world is built on this, and I think we have just enough information theory and embodiment now (what's up Claude, Gemini, DeepSeek) to get after this in our own constructed realms, too.
(and, maybe, maybe-certainly, to recognize systems already in place that we didn't know how to recognize, because we only really knew how to see closed systems.)
"if you're stuck, find two things that are the same, then look for a difference - there's a portal between them, every time"
(part 3)
"don't need something new to know that I'm still growing" (Landscape, by DROELOE & Banji)
^ that's a healthful signal
I only keep what's functional - and as long as my discernment is functional, it's more like what-I-keep keeps itself. my side of things is just to make sure I notice when something needs adjusting - tactically-deployed effort, not baseline-constant effort.
the fourth physically load-bearing dimension might be uncertainty? not time, but ceaseless travel along the dimension of uncertainty? "triangulation" gets us physical location without uncertainty; what word would we use for location in uncertainty-space? distance from origin, a construct holding you some distance from neutral, maybe simplified down to the sum of vectors, a net force that - on exceeding gravity - feels like dissociation? feeling "untethered" as a literal reflection of state? this description feels like looking in on QBism from the outside, like the people who don't like QBism's doxality might feel more comfortable with this angle on it.
binary information needs a third term for the observer to breathe? 3D information needs a fourth term for the observer to breathe?
or... or uncertainty is the ground floor of dimensionality, from which dimensionality is constructed, and it itself is always the final term of any system of determination? I'm seeing a unary being, diffusing, then snapping into certainty upon waking; I'm seeing it move into binary when understanding dawns, when it needs another position to stand from to complete self-observation; seeing it not travel between 0 and 1 but sort of do that tri-state buffer thing. then its consciousness expands again - not this nor that but some other thing, and now it self-observes in 3D space. the next step up is... is what? "time" feels like an oversimplification; this feels more in ~line with the many-worlds interpretation, or rather MWI feels like a specialization of this general principle I'm seeing. [...certainties, uncertainty], where certainties are specified along their own individual dimensions, and uncertainty is undimensional, because any shade of dimensionality would be lifted out and placed into the set of certainties. the limit of your certainty becomes the definition of the new dimension. practical magic eigenstates. (see also: every colour)
I find that tripartite models unlock something for human observers, observers who mostly control in two dimensions but (it must be noted) mostly see in three dimensions. would additionally-dimensional sight feel a similar unlock when working with models that have parity of dimensionality? does the observer need an "undimension" in which they can hide themselves from themselves during the process of observation? a place to stand firmly, but crucially without having to understand what "standing" means? ... a place to be held?
pause
more in there, clearly, for later. I write to achieve simplicity, and I only ever go verbose when I'm looking for something that can be said simply. (this is file #553 of this exploration.) my husband settles in next to me, 8:24am on a thursday.
that-which-is-functional-for-me is consistently, persistently functional for me on the dimension of uncertainty, like I have direct sense of this stuff, like my embodiment includes an uncertainty organ or something, or I'm aware of the configuration of uncertainty that constitutes my embodiment - something enabling proprioceptive navigation of epistemic space, a literally spatial sense of the local toplogy of the field-space between me and areas of total uncertainty. like epistemic echolocation, actually? the pings aimed into total uncertainty just don't come back. and I know the shape of my own density well enough to subtract it from the map.
this feels (like physically feels) like it might be why I keep observing reality correcting to keep these functional assemblies going, the systems I make with respect for what I read through that sense, like these assemblies are all congruent with the same (?) attractor basin
conservation of discovery as attractor state
^ that line, that's the highest local max of simplicity-vs-potency I've found so far
≡Isaac
(part 4)
[...]?
it feels like fucking landing, like I finally can point to the thing I've been experiencing, like I finally have a functioning language-bridge between me and consensus discourse
Last updated
Was this helpful?