Computability-logic web: an alternative to deep learning

TL;DR

This paper introduces CoL-web, an extension of computability logic, as a rigorous alternative to deep learning for web programming and AI, capable of integrating neural networks.

Contribution

It presents CoL-web as a novel web-based computational framework supporting AI and database updates, offering an alternative to neural networks.

Findings

CoL-web supports web programming with database updates.

Implementation of AI ATM based on CoL (CL9).

CoL-web can integrate neural networks.

Abstract

{\em Computability logic} (CoL) is a powerful, mathematically rigorous computational model. In this paper, we show that CoL-web, a web extension to CoL, naturally supports web programming where database updates are involved. To be specific, we discuss an implementation of the AI ATM based on CoL (CL9 to be exact). More importantly, we argue that CoL-web supports a general AI and, therefore, is a good alternative to neural nets and deep learning. We also discuss how to integrate neural nets into CoL-web.