Neural Computers

TL;DR

This paper introduces Neural Computers, a new machine paradigm that integrates computation, memory, and I/O in a learned runtime, aiming for a general-purpose, reprogrammable system.

Contribution

It demonstrates that neural models can learn elementary interface primitives from I/O traces without instrumented program state.

Findings

Neural Computers can learn I/O alignment and short-horizon control.

Routine reuse and symbolic stability remain challenging.

Outlines a roadmap toward fully realized Neural Computers.

Abstract

We propose a new frontier: Neural Computers (NCs) that unify computation, memory, and I/O of traditional computers in a learned runtime state. Our long-term goal is the Completely Neural Computer (CNC): the mature, general-purpose realization of this emerging machine form, with stable execution, explicit reprogramming, and durable capability reuse. As an initial step, we study whether elementary NC primitives can be learned solely from collected I/O traces, without instrumented program state. Concretely, we instantiate NCs as video models that roll out screen frames from instructions, pixels, and user actions (when available) in CLI and GUI settings. We show that NCs can acquire elementary interface primitives, especially I/O alignment and short-horizon control, while routine reuse, controlled updates, and symbolic stability remain challenging. We outline a roadmap toward CNCs, to establish a new computing paradigm beyond today's agents and conventional computers.