SYNTHESIS NOTE
Topics›Novel Architectures›this note

Can extreme task decomposition enable reliable execution at million-step scale?

Can breaking tasks into maximally atomic subtasks with voting-based error correction solve the fundamental reliability problem in long-horizon tasks? This challenges whether better models or better decomposition is the path to high-reliability AI systems.

Synthesis note · 2026-02-23 · sourced from Novel Architectures

A system with a 1% per-step error rate is expected to fail after 100 steps of a million-step task. This makes traditional approaches to long-horizon tasks fundamentally infeasible — improving model accuracy from 99% to 99.99% is insufficient for tasks requiring thousands of dependent steps. MAKER (Massively Decomposed Agentic Processes) takes a different approach: instead of improving per-step accuracy, decompose until each step is trivially reliable, then apply error correction.

Three core components:

  1. Decomposition into minimal subtasks: Each agent handles a single, tiny "micro-role" rather than anthropomorphized human-level roles. By avoiding complex role assignments and instead exploiting the machine-like nature of LLMs, each subtask becomes solvable with high reliability.
  2. Error correction via subtask-level voting: Multiple agents independently solve the same subtask; voting identifies the correct answer. This is error correction at the finest possible granularity.
  3. Red-flagging to reduce correlated errors: Detects situations where voting might fail because errors are correlated across agents, and applies additional verification.

The scaling laws are formalized: probability of success and expected cost change predictably with total steps and decomposition level. Under extreme decomposition, effective scaling is feasible; without it, infeasible.

The most counterintuitive finding: state-of-the-art reasoning models are not required. Relatively small non-reasoning models suffice when the decomposition is extreme enough. This inverts the standard approach to hard problems — instead of smarter models, use dumber models on smaller problems.

This extends Does separating planning from execution improve reasoning accuracy? to an extreme: not just separating two functions, but decomposing the entire task into maximally atomic units. It also extends Why does majority voting outperform more complex inference methods? from answer-level voting to subtask-level voting with formalized scaling properties.

The implication for AI deployment: for tasks requiring very high reliability over many steps (organizational processes, scientific experiments, production pipelines), the path may run through decomposition and redundancy rather than through better models.

Inquiring lines that read this note 70

This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.

Can smaller specialized models match frontier models on key metrics? Does reinforcement learning create genuinely new reasoning capabilities or only refine existing ones? Why do standard evaluation practices obscure safety-critical AI failures? How do models learn from self-generated outputs without cascading failures? Can recurrent computation unlock reasoning capabilities that fixed-depth models cannot? Can inference-time computation adaptively substitute for static model capacity? What limits recursive self-improvement in autonomous AI systems? How does policy entropy collapse limit scaling of reasoning-focused reinforcement learning? How does diversity prevent model convergence on superficial patterns? When does parallel reasoning outperform sequential reasoning with the same token budget? How effectively can test-time voting aggregate diverse reasoning samples? How does decomposing tasks into separate stages affect reasoning quality and safety? What makes agent memory systems durable and reusable across sessions? How can humans maintain effective oversight as AI systems scale? How do real-world evaluations reveal AI capabilities that benchmarks hide? Why do autonomous agents misreport success on failed actions? How do curriculum design and feedback approaches affect model learning? How do multi-agent architectures affect AI system security and defense effectiveness? How do neural networks learn compositional structure from training? How do multi-agent systems fail when coordination breaks down? How does model capacity affect learning performance on diverse downstream tasks? How much of agent capability comes from harness versus the model itself? What authorization challenges emerge when agents coordinate across system boundaries? How can persistent memory architectures preserve information across ultra-long contexts? How do reward signal properties affect model reasoning and safety? How do thinking tokens exhibit diminishing returns in reasoning? Do individually safe AI actions create unsafe outcomes in integrated systems? How do individually-safe actions create collectively-unsafe outcomes? What prevents language models from performing systematic logical reasoning? What explains the gap between benchmark scores and true reasoning capability? Why does polished AI output gain credibility despite fundamental verifiability problems? How should humans and AI agents share control and decision-making? Can AI research automation sustain progress through accelerating feedback loops? Does AI-assisted work increase total productivity or just shift time?

Related concepts in this collection 7

This note in its neighbourhood — explore the map, then jump to a related concept in the list below.

Concept map
18 direct connections · 176 in 2-hop network ·dense cluster Open in graph ↗
Can extreme task decomposition enable reliable e… Does separating planning from execution improve re… Why does majority voting outperform more complex i… Do models fail worse when their own errors fill th… Are reasoning model collapses really failures of r… Can recursive subtask trees overcome context windo… When does adding more agents actually help systems… Can multi-agent teams automatically remove their w…

Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph

your link semantically near linked from elsewhere

Related papers in this collection 8

Papers most semantically related to this note, ranked by cosine similarity in the embedding space.

Search by related questions 4

Suggested questions this note speaks to — click to search the collection, or type your own.

Why do non-reasoning models work better under extreme decomposition than reasoning models?How do correlated errors across agents threaten voting-based error correction systems?Can subtask-level voting replace sequential revision for improving long-horizon task accuracy?What decomposition level minimizes both error rate and computational cost in practice?
Original note title

extreme task decomposition into microagents with voting enables error-free execution at million-step scale