How do you hack on biology in the same time as computers? Primitives you cannot see, no rollbacks, expensive instruments. Iterative cycles measured in months, if not years. An accumulation of specialization transmitted either by the polished paragraphs and tables of a published paper or (not) as tribal knowledge for only the folks who were there.

Being the user of the real is motivating for sure, but one cannot stay in that context forever. In some cases, one may not survive the cure.

Tough stuff. If we could mechanize some aspects of research, sure: automate the manual things, stand up data pipelines, improve the boring bits. All that is in the lab, though.

Possibly, there is no virtualizing the plastic tube.