On: An AI model has found a new ‘CRISPR-like’ biological system. Here’s what it...

The headline promises to tell me what the finding “means for science.” As if meaning were a parcel handed over with the news. The meaning will show itself later, in what researchers can now go and do at their benches. Until then, “means” is only a hook.

They say the machine found a new biological system, having done code and mathematics first. Found. When a man finds a tool in the ground, we know how to check: he picks it up, it cuts. Here too the check is old-fashioned - someone must carry the result to a bench and see whether it works in a living cell. So the word “found” keeps its use. What waits offstage is the next word: understanding. These are not the same achievement. A man can find a key without knowing what it opens. We say he understands when he can go on - teach a student, answer when asked why, correct himself when the method fails. The machine goes on along one groove. Whether we call that understanding is decided by looking at what we do with the word, not by feeling awe at the machine.

And “the messy world of biology” - the writer offers messiness almost as an apology. The mess is not a defect of the subject. It is what a subject looks like while it is still alive. Mathematics seems clean because the practice has been polished by use until we no longer see the hands. The cell is messy because we are still close enough to touch it.

I find I am not excited. An instrument that works is a good thing; instruments do work. But my unease is with the words around the machine, and those words - found, means, understands - were loose already. The machine did not loosen them. It only pulled, and now we can see how far they stretch.