iris carver, the telos project — ideas
the telos project
the telos project shows us that its somewhat ambiguous what it means for a game to be simulating predictions rather than outcomes, and the text doesnt necessarily clear up that confusion any better since its set-up like an epistemic trap. iris carver, a hyperstitionist prophet from the future heralding the project of dr. oskar sarkon re-emerges through an older text, occultures, as questioning the limits of a new type of crypt, one that has more to do with prediction and less to do with manifestation. the telos project claims that the game (the simulation that tests its own boundaries) is fully free to form? on the one hand this could mean the world itself is already a part of it, a meta-game about its own outcomes? (we will see quickly that this isnt the case). but given that that were true, in that case, we would already be playing it, so what does the machine even do that is meta at all if its already fully meta? might it be that a full meta is just already the world in its self-manifestation?
the goal with this text isn't to expand the question “can teleology want itself?” which the original poses, but to show how ontological closure can emluate a self-desiring predicational machine without collapsing into a self-recursive totality that eliminates the externally-posited differential that makes prediction meaningful (and generative).
around the questions that were posed at the introductory paragraph, a fully “free-form” game in a sense means the boundaries between controller, designer, and player collapse. the difference here is that the world only recursively anticipates itself, whereas hyperstitious anticipation would mould both its operational target and more directly influence successive chains of roles, disposing and shifting the roles and intentions of agents in the game faster than they can produce predictable outcomes. the text finishes through this dilemma as the idea of the collapse of the world's generative power, it either questions the limits of ontological value, or it questions whether epistemology can truly uncover recursivity in a way that ends up productive (simulate outcomes and also manage to govern them without collapse), even without hacking the underlayers of the world (virtuality) itself (if such a thing is even possible).
the problem is, in a world that is fully generated by this machine, anything outside of it is also meta, governance here implies all limits of the virtual are all possible political boundaries. anything outside the machine that is acting in a non-meta manner (and in that case, it must be acting non-meta) prevents the machine from fully simulating and thus from fully generating the world. the classic answer to this question is that externality reinforces and sometimes even imposes boundaries instead of breaking them, and that recursion is solved by higher infinities accounting for them, which always end up open at the top. but in that sense, we simply return to an image of this world's teleology, the one that we already exist under, where simulations are always partial and loops are closable by default before they emerge as meta-paradoxes.
the simulacrum itself and its limits are defined by origin points which themselves have an embedded talisman, a fragment that shows the proof of a higher-tier unalterable world, the proof that even if there is a simulation, its either infinitely recursive or scaling towards upper limits that choose not to define the lower ones. this however in and of itself already disables recursivity from being productive. in deleuze's reading of nietzsche, recursivity becomes productive through a becoming that selects future orientations through power, and the circut is closed from within re-ontologized successivities. a hyperstitious machine would work slightly differently than this.
the real solution to creating axsys would be one where a generative computational simulation is generative not because it causes everything but because it "marks" it for future value (this operation itself is fluid, value and marking are metaphors for higher-tier abstractional functions, one that don't necessarily operate functionally the same way as marking or value, but are a form of open-access accounting, but closed-access simulating - essentially, that events occur determinately but that they are felt indeterminately).
if it has an account, it has externalities. now you might say, ok but the universe is infinite and this and that. but there is actually an important difference between infinity as an abstraction and infinity as a limit, aka infinity as impactful. the infinite isnt even necessarily infinite but "greater than" any meta. the actual "infinity" (worlds greater than world) that exists for us as the universe in a vacuum itself is actually not empirical but abstract if it doesnt impact a future potential meta-generator (meaning the virtual is an empirical curse that may be prevented, a form of generation that can be computationalized away from occuring. the meta-generator reaches ontological finality when it prevents anything external from contacting it in the infinite future that it itself doesnt cause. in that sense, deus ex machina actually stops being a theological concept where god's distance from us is immesurably large when we can't decide it and immesurably small whenever he does decide it, but optionality is an actual predicable possibility, contingency itself isn't a natural law but an imposition, necessitated rather than necessary contigency.
the thing about marking is that it doesnt presuppose valuation or categorization, if you take marking as a metaphor for the universe, marking itself doesnt imply totality, marking doesnt need to be an agential activity it can be a direct form of generation - also known as creation. if it reaches this point there is no contingency, since agentiality is the same as value (as for exampel in spinoza)
you can always create an additional two external agents, if our machine optimizes everything around it by closing it, a probabilistic machine optimizing to infer future threat optimizes to cleanly redirect any contact into the internal machine / make it as if it didnt occur however way it can - and one extra one, where the machine's system optimizes to close the existence of anything else in some causal field outside the possible trajectory of the machine. one becomes the only agent, the other becomes the only boundary, the third becomes the only background, the second (probabilistic actor) keeps the generative machine uncontaminated even if it isnt actually perfect
returning away from speculative ontological synthesis and into the minor-world epistemic problem posed by the text, or otherwise an attempt at simulative generation, lets say that axsys is a state where a bunch of predicting machines exist that predict outcomes of like five things, the existence of future polymarket gambling bets (not their outcomes but their existence), a team that decides that the near future will look a specific way and then designs a story around it, a team that sends that story into the world and mass markets it, a team that already cosplays as aspects of the story, and a team that tries to kill or exploit the fourth team, but also whatever profits it makes from this, it sends them back to polymarket. the text never speculates on a narrative about how dr. oskar sarkon creates the telos machine, but if we could imagine one existing on the world, it wouldn't start as anything too far from a proposal similar to this, given that current institutional synthesizers in the form of investment firms and market predicting conglomarates already behave impositionally about meta-environments, and all they'd require is one more meta-synthesis to loop into being causal agents (or to fully decinegrate all real/realized abstractions, depending on which road this narrative takes)
the two things missing right now are inter-actor synthesis (players becoming controllers and vice versa) and a corruption of all functional systems of value so that intentions intertwine or are obfuscated. in the narrative, as the machine evolves, polymarket switches to a central banking system, a hyperstitious organization, a corporate self-sovereign superpower creating hyperscales that predict the future, it becomes an international space station that oversees the planet, it becomes integrated into future technology, it then redesigns itself as betting on cosmological issues rather than world affairs such as meteor strikes. meanwhile, the teams themselves all switch places, and slowly get totalized as team cosplay starts to kill team exploit, whereas team marketing starts to meta-market itself as a better version of team original, eventually surpassing it. meanwhile, a fifth team emerges that tries to convince the other four roles it can do a better job than them. eventually, after a half dozen years and slow coercion it works, but an organizer in the team slowly starts to meta-cosplay the teams themselves, taking over their objectives. an example of this is that re-formulates the now space-ship back into a ground mission organization (where others are betting on whether he will do this.)
after this, he (likely in the narrative of the original text this man is dr. oskar sarkon himself) succulently bets on himself over various things. including that he will take over anyone who bets on him so that only telos can bet on its own activities. then he bets on himself in the form of a body-double of himself that he will not create a machine to duplicate himself and also that he will not kill himself, after which he loses his own bet, kills himself, body-doubles back into life, and then uses the machine to roleplay as various betting agents about his own meta-tasks. meanwhile, as other betting companies who are now slowly being replaced by his own actions due to having no will other than meta-power prediction succumb to telos, the betting system is outvalued by a system of generative sovereignity that acts on behalf of his own will to conquer future predictions on other predicting software, as this happens, the world itself slowly starts transforming into whatever other form the original polymarket structure is to take.
these processes by themselves already correspond to carver's earlier conceptual developments in "what didn't happen in the millenium?" where body-doubles speculating on eachothers contagion maps to a "lesbovampiric virality" that spreads as engineered crypt-mechanized paleofeminized bodies emerge as realized abstractions of eachothers ensourced speculations, crypt-sex is precisely identical to the infections it transmits", and where the ground zero of the earlier speculated triple-functioned virtuality-destroyer meta-machine is itself operating within a flatlined materialism, where a recursive field is only modelling its own anticipations as if they are immanent rather than speculative, therefore turning speculation into the temporal engine itself rather than something that assumes or tries to model external influence as a byfactor of its existence.
in this instance, the world reaches completion by autopoietic sovereignty, meaning that the telos of the agent corrupts the telos of the world, and the world reaches a state i refer to in my own terminology as thymosic, or a baudrillardian schizophrenization of world-infrastructure into a consistently moulding but no longer identifiable cacophony. however this leaves us with two unresolved dilemmas both found in "occultures". one is over the difference between the zero-level of transcendence and the flatline, or in other words, how intention actually impacts the epistemology or ontology of the crypt (or my graphicatization of it - the meta-machine), which aren't actually related, and the other is over where machinic flow is connected or not connected with the concepts of temporality and predictability.
the "the telos project" ends with this question "in saying this, of course, axsys had not yet reached its conclusion, but it had already begun to question its own purpose. “can teleology want itself to happen?” it asked, awkwardly, in the middle of a game. “what is the telos project’s final end?” y2k helps us bridge dates, acronyms and indices into mechanism, but it stresses that the mechanism itself loops inwards, and that symbols neither work as functions nor as predictations but as encodings or talismanic gates. this causes an issue for ontological construction, because encryption itself has no talismanic basis outside of social organization unless the arcane actually becomes an ontological law with virtual potential - something not entirely certain in the crypt or it's more secularized and centralized computationalist version - or the talismanic basis itself is a computational post-appropriative layer. however, if it is, then higher order symbolic power falls on it either being predicative and intendational or it being presupposed and therefore sterile and naturally innate - essences themselves collapse their potential function into manifestation too quickly.
the solution for fixing markets and intentions is simple, you can have the doctor generate patterns that appear hyperstitionally fresh, that are then offloaded into actual simulations of externality, and then pre-loaded back into prediction-manifestation. the process of prediction-manifestation is one where intention, agentiality and predication are bypassed as laws and become functions, but where determinism is necessarily obfuscated by a virutality that "screens out" (as is well known) parts of the system. but this screening out holds both a rabid potential that exists as its own simulated but very real externality, and a dynamic layer that also obfuscates the pre-emptive results for the generative agents, allowing a type of "posteriori" act of creation.
the problem of or for time is a little different, where encryption can't possibly influence succession without manifestation, but can influence it through predication. the part of the future that re-emerges in the present isn't the "known" aka intended part of it, nor is it recursive, but it appears as the return of a past symbol in a slightly altered form, but cruically one that has no talismanic value. essentially, a ritual encoding that behaves temporally schizophrenic, but clearly demarcates the splits of time, where even if temporality collapses inwards, outwards it exists as a form of accounting.
ontological closure, then, allows teleology (meta-goals) to become self-inherited essences by a process that allows a stable and never-altering ontological insulation through a meta-machine, just enough continuity through a crypt-forming paradigm, an overabundance of faux signification through temporal haunting, a value-shifting process that aligns with a generative computational simulation that shifts correspondences of values correctly depending on the narrative progression of historical simulation, and sometimes even by an ontological exclusivity as enumerated elsewhere, such as in my writings on the halograph.