Position Paper: Towards Transparent Machine Learning

TL;DR

This paper advocates for transparent machine learning, where models are expressed as source code to enhance human understanding, verification, and refinement, aiming to improve AI safety and security.

Contribution

It introduces transparent machine learning as a novel approach, emphasizing source code representation for better interpretability and control over AI models.

Findings

Proposes source code as a means for transparency

Highlights potential for improved AI safety and security

Lays groundwork for future development in transparent ML

Abstract

Transparent machine learning is introduced as an alternative form of machine learning, where both the model and the learning system are represented in source code form. The goal of this project is to enable direct human understanding of machine learning models, giving us the ability to learn, verify, and refine them as programs. If solved, this technology could represent a best-case scenario for the safety and security of AI systems going forward.