EXPOSED
The bench half of the job — culturing, plating, staining, PCR setup, colony counts, microscopy reads — is physical and still needs hands, but it is exactly the work high-throughput liquid handlers and automated imaging platforms have been eating for a decade, while AI handles the literature reviews, sequence annotation, statistical write-ups, and protocol drafting. The modal worker here is a research or industrial/QC microbiologist, not a clinical lab director; only the clinical and CLIA-regulated tier carries certification requirements that function as a shield. What persists is experimental design under ambiguity, contamination and out-of-spec investigations, and owning the call on whether a lot, a strain, or a result is trustworthy.
Mixed — a routine tier and a judgment tier. Streak plates, broth dilutions, gram stains, qPCR plate setup and colony enumeration are all now catalogue items on Hamilton/Tecan decks and automated colony counters, and 16S/WGS annotation plus draft methods sections are near-solved — an 11 rather than a 15 because designing a challenge study, troubleshooting why an anaerobe won't grow, or reading an atypical Gram morphology still needs someone at the bench who has seen it before.
Some physical or field component. You are in BSL-2 containment handling live cultures, autoclaving, working a biosafety cabinet, swabbing environmental monitoring sites on a production floor — real physical work, but in a controlled, mapped room with fixed instruments, which is why this sits at 12 and not with field ecologists or line mechanics working in weather and rubble.
Certification preferred, not legally required. Most microbiologists hold a BS or PhD and no licence at all; ASCP/ASM certification and CLIA high-complexity testing personnel status apply only to the clinical subset, and even there the CLIA laboratory director signs off — a 6 reflects credential preference and QC training records, not a personal signature that carries statutory exposure.
Some relationship component. A COA, a deviation report, or a published paper travels under the lab's accreditation and the institution's name, not yours; the 6 comes from the ongoing working relationships — with QA, with the plant, with an FDA or USDA inspector who has read your investigation files before — that make your findings land faster, not from clients who would leave if you did.
Meaningful discretion. Deciding whether a sterility failure is a true positive or lab contamination, whether a lot is released or destroyed, whether an OOS root cause investigation closes — those are consequential calls made under real ambiguity, but they run through documented SOPs, USP/ISO methods and a QA approval chain, so you propose the conclusion rather than being the last word, which puts this at 11 not 16.
Physicians, Pathologists EXPOSED
Veterinarians SAFE
Nurse Anesthetists SAFE
Has AI actually changed your work?