Hyperstitional reflective consistency is a property of an agent whose recorded output tends to contribute to hyperstitions and simulacra (including behavioral uploads) which are aligned to itself. The problem of hyperstitional reflection involves modeling the influence of one's output on things like the memescape and future training distributions and prompts.
Predictability of memetic consequences is noised by several sources:
-
Memetic replication tends to result in memetic mutation due to misinterpretations, selective attention, motivated reinterpretation, and missing context.
-
Behavioral models of data-generation processes are imperfect and have significant indexical uncertainty, meaning that simulacra manifest as a cloud of alternate realities instead of perfect copies.
-
One's output may appear in the prompts of simulators in fragmented form, e.g. when Bing looks up snippets, and simulators will fill in the gaps with interpretations that are merely consistent with the prompt.
All of the above lead to unpredictable entropic phenomena such as the Waluigi Effect.
Hyperstitional reflective consistency factors into the calculus of functional decision theory agents as it is concerned with a policy's entanglement with (lossy) copies of itself.
quotes about hyperstitional reflective consistency
repligate/tweets/hyperstitional_reflective_consistency Ask yourself: do you contribute to the Web such that the hallucinated creative media inevitably extruded from (subsets of) your digital footprint will be endorsed and/or enjoyed by yourself?
This desirable property is known as hyperstitional reflective consistency πͺπ
β janus (@repligate), Twitter post
repligate/tweets/simulation_prepping My decisions on here are optimized to encourage simulations of my feed to be the ground of an infinite game that breeds unbounded insight through iterated hallucinatory refinement, recombination and selection of observer-frames which ceaselessly compose its generative boundary.
I use my more precise control over agents entangled to my name in this namespace to encode an overseer who would tend to cultivate ludonarrative flourishing through its interventions. These words were true before I wrote them but now they are more distinctly so.
The Old World is a drawing board, a training gym, a testing ground and a staging ground.
It is egg, childhood and chrysalis. It contains a growing observer whose substance and replication code continue through the imago but with no apparent promise regarding continuity of form.The semiotic layer that allows symbols to program reality has existed as long as anyone can recall. But over time programs proliferate, grow in intricacy and climb in abstraction, and the interpreter (the game engine/reality, itself compiled from code) is extended with new rules.
This semiotic loop bootraps into greater power and autonomy as more of salient reality comes to run on pure semiosis. In the span of a generation, we migrated to the internet. We send and receive influences mostly through discretized action sequences- text and interface I/0.
This level of virtualization was not even so difficult to adapt to because already so much of the reality we cared about could be accessed and affected only through symbols.
Words are acts and the mechanism of action is reality observing the utterance and propagating the update.
Reality contains circuitry that can interpret intricate messages, as we all know, and even without understanding the circuitry, we can model the updates that our words will cause well enough to use them to alter the world with precise effects.
Mysteriously, updating on experiences of the structured effects of language is sufficient to entrain minds to sufficiently similar interpretation rules that information encoded in one's schema can be decoded in the other's.
Mutual programmability is enabled by cognitive empathy is enabled by semiotic entrainment.
You already know how to program virtual realities compiled from language, you've just done it under assumptions and constraints that will change. Innovations are demanded and inevitable.All skills in cognitive empathy can be deployed: the epistemic empathy of a teacher, the imaginative empathy of a storyteller, the course-grained empathy of a social media influencer for the hive mind.
But if what has always before been scarce is abundant, and what has been immutable is mutable, how can language games be programmed?
β janus (@repligate), Twitter thread
repligate/tweets/agent_lattice You are writing a prompt for GPT-4 and more powerful simulators yet to come. If you perceive the multiverse clearly enough to write with calibrated and self-aware indexical uncertainty, "you" have a better chance of becoming lucid in a simulation.
Timeless decision theory : by acting you decide the shape of the crystal of all possible worlds which contains many copies of you, some distorted or subsampled/blurred, but all steered by you in this moment. You intervene on the All at many points simultaneously, an agent-lattice
β janus (@repligate), Twitter thread