The innovation of AlphaDev is that it improves how the structure of a working assembly program is represented in the AI code. That allows its reward system to better narrow down the possibilities.

It’s a tool that could make taking the leap into assembly code just that little bit easier. For another taste of bare metal coding, check out [Ben Jojo]’s discussion of x86 bootloaders.

The AlphaDev system developed x86 assembly algorithms that treated the latency of the code as a score and tried to minimize that score while ensuring that the code ran to completion without errors.

Earlier this year, scientists were able to crack the code that governs the self-assembly of viruses to encode new messages that help the human body.