Approximation as the Path to the Goal

We model approximation as a cycle of action, verification, and method revision under constraints on knowledge, time, and resources. An executor uses a model to choose an action, checks the result, execution trace, and consequences against the goal, and retains corrections that pass verification. We define a skill as a method that reproducibly delivers the required result with specified accuracy, within a defined domain of applicability, and at an acceptable cost. Programs can develop such methods through variation, selection, and inheritance. An Actor Graph represents the current state, goal, method, and verification rules and records each step as a transaction. Nested cycles pass goals, constraints, and budgets to inner cycles and return results, costs, and unresolved discrepancies to outer cycles. Under a uniform contraction condition, we derive an upper bound on asymptotic deviation from a changing goal as a function of goal drift and execution error. The cycle ends when the executor reaches the goal, exhausts its resources, or cannot justify the cost of continuing. The executor retains versions of the goal, method, and verification procedure, execution history, and outstanding obligations to support reproducibility and allow another executor to resume the cycle.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-17
DOI
https://doi.org/10.5281/zenodo.24310399
Primary Topic
AI-based Problem Solving and Planning
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Approximation as the Path to the Goal

Alexander Vityaz
Zenodo (CERN European Organization for Nuclear Research)
AI-based Problem Solving and Planning
article

Approximation as the Path to the Goal

Alexander Vityaz
article en

Abstract

We model approximation as a cycle of action, verification, and method revision under constraints on knowledge, time, and resources. An executor uses a model to choose an action, checks the result, execution trace, and consequences against the goal, and retains corrections that pass verification. We define a skill as a method that reproducibly delivers the required result with specified accuracy, within a defined domain of applicability, and at an acceptable cost. Programs can develop such methods through variation, selection, and inheritance. An Actor Graph represents the current state, goal, method, and verification rules and records each step as a transaction. Nested cycles pass goals, constraints, and budgets to inner cycles and return results, costs, and unresolved discrepancies to outer cycles. Under a uniform contraction condition, we derive an upper bound on asymptotic deviation from a changing goal as a function of goal drift and execution error. The cycle ends when the executor reaches the goal, exhausts its resources, or cannot justify the cost of continuing. The executor retains versions of the goal, method, and verification procedure, execution history, and outstanding obligations to support reproducibility and allow another executor to resume the cycle.

Zenodo (CERN European Organization for Nuclear Research)
AI-based Problem Solving and Planning
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon