Can Biologists Rewrite the Genome’s Spaghetti Code?

The friction Woolfson identifies is more specific than 'biology is complicated': evolution optimizes for survival, not readability, so genomic sequences carry layered, context-dependent functions that break when you treat them as swappable modules. That mismatch is the actual engineering problem AI-driven genome design has to solve, not just the scale of the sequence space.

This is largely disconnected from recent activity in our archive, which currently holds no related coverage. The story belongs to a cluster of developments worth tracking together: the maturation of large biological language models (such as those trained on protein and DNA sequences), the commercial buildout of automated wet-lab infrastructure, and ongoing biosafety policy debates about who governs AI-assisted organism design. Woolfson's framing of ABI as a distinct category is an attempt to give that cluster a conceptual handle, which is useful for policy and funding conversations even if the term itself doesn't stick.

Watch whether MIT Press or Woolfson follows the book with a concrete ABI benchmark or design challenge in the next twelve months. A named, reproducible test case would signal the concept is hardening into something researchers can falsify; continued abstraction would suggest it remains a framing device rather than a research program.