AGI-Systems

DOI

Design Notes on Memory, Context, Orchestration, and Self-Directed Evolution in AI and AGI Systems

Written by Hajime Tsui (@hajimetwi3) - January 2026

Preprint

Abstract (from the Preprint)

The abstract below is quoted from the preprint version 1.2. For the most up-to-date version, please refer to the Zenodo record.

This paper presents opinionated design notes on memory, context management, orchestration, 
and self-directed evolution in AI and AGI systems. In this context, AI and AGI are primarily
discussed with current LLM-based systems in mind.

As interest in Artificial General Intelligence (AGI) continues to grow, it becomes increasingly
important to examine not only model scaling or training methods, but also how systems manage
memory over long time horizons, revise their own internal structures, and coordinate complex
internal processes.

Rather than proposing a finalized architecture or implementation, this work explores a set of
aggressive design hypotheses, including memory maps, selective context retrieval, adaptive
forgetting mechanisms, greed and dynamic personality, controlled and extensible orchestration,
modular skill representations, multi-model utilization, and self-directed internal updates.

References

1. Preprint : Post Engineering for AI: Benevolent Contextual Guidance for Debiasing Large Language Models

Preprint (Zenodo, Concept DOI - latest version): https://doi.org/10.5281/zenodo.17896136

2. Essay : Hello seed (English version)

Preprint (Zenodo, Concept DOI - latest version): https://doi.org/10.5281/zenodo.18089594

Author

Written by Hajime Tsui
GitHubPage: https://hajimetwi3.github.io/hajimetwi3/
GitHub: https://github.com/hajimetwi3
X (Twitter): https://x.com/hajimetwi3
Reddit: https://www.reddit.com/user/hajimetwi3/
BlueSky: https://bsky.app/profile/hajimetwi3.bsky.social
Zenodo

Contact

For questions regarding this project, you may contact me via on X (formerly Twitter) at @hajimetwi3.
Please note that I may not always be able to respond, and your understanding is appreciated.

This project is a personal research record.
Suggestions from external users may be reviewed as reference information, but they do not constitute co-authorship or co-invention.

No intellectual property claims regarding submitted suggestions will be recognized. Even if a suggestion is adopted, it will be incorporated only after independent evaluation and restructuring, and will not imply authorship of the original submitter.

If you have an interest in formal collaboration or joint research, you may express that interest.