A Path Towards Autonomous Machine Intelligence
Source
- Raw Markdown: paper_lecun-autonomous-machine-intelligence-2022.md
- PDF: paper_lecun-autonomous-machine-intelligence-2022.pdf
Core Claim
LeCun proposes an autonomous intelligence architecture built from configurable predictive world models, intrinsic objectives, hierarchical planning, and joint embedding architectures trained by self-supervised learning.
Key Contributions
- Frames world models as the missing substrate for human-like sample efficiency, reasoning, and planning.
- Argues for prediction in representation space rather than direct pixel-level prediction.
- Connects intrinsic motivation, actor modules, cost modules, and latent variables into one agent architecture.
Method Notes
This is a position paper rather than a narrow empirical result. It provides the conceptual root for JEPA, Energy-Based Models, and World Models in this wiki.
Evidence And Results
The evidence is architectural and argumentative: the paper compares limits of supervised learning, reinforcement learning, and generative modeling, then motivates hierarchical predictive representations.
Limitations
The proposal is broad and leaves many training details unresolved; later sources such as LeJEPA and LeWorldModel instantiate pieces of it.
Links Into The Wiki
Open Questions
- Which parts of the proposed architecture are necessary versus optional?
- How should hierarchical prediction be trained at large scale without collapse?