Intelligent Learning Laboratory Theory: An AI-Augmented Architecture for Inquiry, Experimentation, and Knowledge Construction
- 1Supreme National Defense University, Tehran, Iran
Abstract
The rapid convergence of artificial intelligence, learning analytics, adaptive learning, virtual and physical experimentation, and agentic scientific systems creates an opportunity to move beyond conventional intelligent learning environments toward learning architectures that can support complete cycles of inquiry. This paper develops the Intelligent Learning Laboratory (ILL), a human-centered theoretical architecture in which learners interact with context-aware AI agents, experimental infrastructures, digital resources, and epistemic learning analytics to formulate questions, generate and critique hypotheses, design experiments, interpret evidence, revise explanations, and construct knowledge iteratively. The theory is developed through evidence-informed synthesis across six previously fragmented research streams: intelligent learning environments, inquiry-based learning, laboratory learning, learning analytics, artificial intelligence in education, and AI-supported scientific discovery. Recent evidence shows that intelligent learning environments can improve learning outcomes, while systematic reviews identify persistent limitations in learning analytics for capturing hypothesis generation and evidence interpretation. Emerging scientific AI systems demonstrate increasingly autonomous cycles of hypothesis generation, debate, experimental planning, and validation, but these systems have not yet been adequately translated into educational architectures. ILL addresses this gap through a five-layer architecture contextual intelligence, adaptive learning intelligence, AI inquiry agents, intelligent experimentation, and knowledge construction—and a recursive epistemic cycle. The paper introduces three candidate constructs: epistemic learning analytics, AI-supported co-inquiry, and epistemic agency. Eight propositions and twelve testable hypotheses are derived. The manuscript specifies a sequential mixed-method empirical validation design and delineates governance conditions including human oversight, uncertainty communication, verification, AI literacy, and data governance. The contribution is therefore not the claim that AI can replace teachers or researchers, but a theoretically bounded architecture for integrating AI into learning environments while preserving human epistemic agency.
Keywords: agentic AI, epistemic agency, inquiry-based learning, intelligent learning environments, learning analytics, knowledge construction
Related
- Design and analysis of an intelligent audit risk assessment system based on retrieval-augmented large language model
- GEO-BERT: A Retrieval augmented LLM Framework for Intelligent Dataset Retrieval from the Gene Expression Omnibus
- An agentic framework for intelligent clinical laboratory medicine.
- QCSA: LLM-driven progressive quadruple construction and prompt-augmented learning for robust citation sentiment analysis
- SuperJEPA: Pretraining with a Supervision-Augmented Joint-Embedding Predictive Architecture for High Energy Physics
BibTeX
@misc{sadeghi2026intelligent,
title = {{Intelligent Learning Laboratory Theory: An AI-Augmented Architecture for Inquiry, Experimentation, and Knowledge Construction}},
author = {Sayed Hadi Sadeghi},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23503817},
url = {https://zenodo.org/records/23503817},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Intelligent Learning Laboratory Theory: An AI-Augmented Architecture for Inquiry, Experimentation, and Knowledge Construction
AU - Sayed Hadi Sadeghi
PY - 2026
DA - 2026/10/20
PB - Zenodo
DO - 10.5281/zenodo.23503817
UR - https://zenodo.org/records/23503817
AB - The rapid convergence of artificial intelligence, learning analytics, adaptive learning, virtual and physical experimentation, and agentic scientific systems creates an opportunity to move beyond conventional intelligent learning environments toward learning architectures that can support complete cycles of inquiry. This paper develops the Intelligent Learning Laboratory (ILL), a human-centered theoretical architecture in which learners interact with context-aware AI agents, experimental infrastructures, digital resources, and epistemic learning analytics to formulate questions, generate and critique hypotheses, design experiments, interpret evidence, revise explanations, and construct knowledge iteratively. The theory is developed through evidence-informed synthesis across six previously fragmented research streams: intelligent learning environments, inquiry-based learning, laboratory learning, learning analytics, artificial intelligence in education, and AI-supported scientific discovery. Recent evidence shows that intelligent learning environments can improve learning outcomes, while systematic reviews identify persistent limitations in learning analytics for capturing hypothesis generation and evidence interpretation. Emerging scientific AI systems demonstrate increasingly autonomous cycles of hypothesis generation, debate, experimental planning, and validation, but these systems have not yet been adequately translated into educational architectures. ILL addresses this gap through a five-layer architecture contextual intelligence, adaptive learning intelligence, AI inquiry agents, intelligent experimentation, and knowledge construction—and a recursive epistemic cycle. The paper introduces three candidate constructs: epistemic learning analytics, AI-supported co-inquiry, and epistemic agency. Eight propositions and twelve testable hypotheses are derived. The manuscript specifies a sequential mixed-method empirical validation design and delineates governance conditions including human oversight, uncertainty communication, verification, AI literacy, and data governance. The contribution is therefore not the claim that AI can replace teachers or researchers, but a theoretically bounded architecture for integrating AI into learning environments while preserving human epistemic agency. Keywords: agentic AI, epistemic agency, inquiry-based learning, intelligent learning environments, learning analytics, knowledge construction
ER -