Example · Biology and Pre-Health Student

73% of the work a biology and pre-health student is heading toward still needs a person.

AI scribes are changing the pre-health jobs fastest. Bench work and signed records stay with people.

Make my planA typical week at the middle pace. Your plan uses your own tasks.

Is a biology degree worth it with AI?

In a typical week at the middle pace, 43% of this work needs your judgment and 30% is work you do using AI. AI could do about 27%. Biology graduates had a harder first job search than most majors long before AI showed up. The New York Fed's data for 2024 has about half of recent biology graduates underemployed, working jobs that do not need their degree (New York Fed, The Labor Market for Recent College Graduates, February 2026).

Read more

That problem came before AI. AI now changes which first jobs are worth chasing. The bench work in your future is protected by your hands and by rules that make a named person sign the record. The note-writing jobs many pre-health students take for clinical hours are the part most exposed.

So the plan is about choosing first jobs on purpose. You want ones that put a sample or a patient in front of you, and ones where your initials go on the record. If lab meeting makes you nervous, it helps to become the person in the room who checked the numbers, including the ones a model produced. You can stay in biology, and you can stay at the bench.

The week, task by task

Who does each hour of a typical 40-hour week.

See every task
You43% · 17 hours

Needs your judgment.

  • Bench work: samples, cultures, assays, pipetting18%
  • Patient contact: vitals, rooming, screening, consent conversations10%
  • Running GMP release tests and signing the result (QC)8%
  • Setting up, calibrating and troubleshooting instruments7%
You using AI30% · 12 hours

AI does the slow parts. You still own it.

  • Analyzing data, plotting and writing results for the PI7%
  • Image analysis: cell and colony counts, gels, microscopy7%
  • Study paperwork: case report forms, IRB documents, scheduling (CRC)7%
  • Ordering supplies and keeping inventory4%
  • + 1 smaller task4%
AI27% · 11 hours

AI could do most of it.

  • Writing the clinical note for a physician (scribe work)9%
  • Recording results in the notebook, logs and databases9%
  • Reagent math, protocol and SOP first drafts7%
  • Literature searches and background for lab meeting3%
Why some tasks still need a person
Plating, media prep, culture work, running assays at the bench

Someone has to be in the room with the sample. Autonomous labs exist, and they still need people to set the question, load the deck and catch the error. When a model proposed a negative water volume, Ginkgo's robot technicians ran the experiment at a corrected volume (Scientific American).

Keeping an instrument running: calibration, cleaning, the autoclave log

Every instrument has habits nobody wrote down. The person who knows them becomes the person the lab calls.

Your initials on a regulated record

Under FDA's GLP rule for lab records, every data entry "shall be dated on the date of entry and signed or initialed by the person entering the data" (21 CFR 58.130(e)). Under GMP, the FDA's rule for drug making, lab records need "the initials or signature of the person who performs each test" and a second person's review (21 CFR 211.194). A model cannot be that person.

Talking with a patient: vitals, rooming, a consent conversation

O*NET, the federal job database, lists overseeing enrollment "to ensure that informed consent is properly obtained and documented" as core coordinator work. That work takes a person in the room with the patient.

Checking whether a count or a figure is right

As analysis gets cheap, the person who can tell a real result from an artifact becomes the scarce one. It is also the skill that makes lab meeting easier.

Good to know

What is hard about this field, and what you can do about it.

The hard truth

A biology degree alone often leads to non-degree work.

For most people, a biology bachelor's is a first step. About half of recent biology grads work in jobs that do not need a degree, and most biology grads eventually earn a graduate degree.

What you can doPick your next step by junior year: a health program, a lab tech job that builds toward grad school, or a skill like quality control or checking data.
Sources (2)

What can I do with a biology degree besides med school?

Plan A is the best move in the field you are studying for.

APlan A

Focus your studies

The main focus. Most weekly sessions build toward it.

Your next stepQuality control jobs in biotech, drug company and hospital labs
Each week
4 sessions
Ready in
One semester to learn the rules and own one instrument log
Why it fits

Why you. If you already do the bench work and like knowing exactly what you did, this is your door.

The gap. You don't yet know GMP and GLP, the FDA's rules for drug making and lab records. You also need one instrument or quality task documented as yours, in a research lab or a lab course.

The trade-off. Repetitive work, sometimes shifts, and industry hiring runs in cycles. O*NET's QC median of $62,280 is borrowed from a broader category, so check current pay in job postings near you.

2 other moves that fit
  • Clinical research coordinator or patient-care jobs for your gap yearsReady in about a week for the ethics training
  • A research technician job or a paid NIH research year after collegeReady in one summer

What's a plan B for pre-med students?

Plan B is meant to be different from this job. It was picked because skills from this role carry over.

BPlan B

Get ready for a new kind of job

A second option. A newer job AI is creating that these skills already fit.

A job to aim forClinic AI agent specialist

A gap-year clinic job rooming patients puts you at the desk when AI agents go live, ready to be the superuser.

Each week
1 session
Outlook
Emerging
Pay
$110K to $190K a year (senior roles)Two health AI vendor postings, June and September 2026; Assort adds commission.
2 other options
  • Expert AI trainer (contract)Running assays and checking whether a count is right helps you grade AI answers to biology questions with care.
  • AI governance analystPutting your initials on regulated GMP records builds the documentation discipline compliance teams want for AI in labs.

In your own plan you can switch to any of the 47 roles.

The 13 weeks

Small steps that add up. One plan for both: most sessions build Plan A, one a week builds Plan B.

By week 13 you will have

AFor Plan A

  • An error log of AI mistakes you caught on public lab data
  • A research statement with a one-page evidence sheet
  • Three summer research applications submitted

BAnd for Plan B

  • A before-and-after score table
  • A one-page clinic agent safety summary for your portfolio
Month 1Weeks 1 to 4
  1. 1Measure where you stand
  2. 2Choose your test year
  3. 3Start your error log
  4. 4Learn who signs the record

You finish withAn error log with about 20 entries

Month 2Weeks 5 to 8
  1. 5Plan a short talk
  2. 6Find patient contact options
  3. 7Make one figure in code
  4. 8Finals week, one light session

You finish withAn up-to-date hours sheet and error log

Month 3Weeks 9 to 13
  1. 9Finals week, ask for letters
  2. 10Break: list summer programs
  3. 11Break: write your statement
  4. 12Break: get feedback
  5. 13Break: submit applications

You finish withA third submitted summer application

Three courses from the plan
  • Summer Internship Program (SIP) 2027NIH Office of Intramural Training and Education · Paid; stipend set by education level

    Yes, Week 13.

  • REU Sites (Research Experiences for Undergraduates)NSF · Stipend; many sites help with housing, meals and travel

    Yes, Weeks 10 and 13.

  • Postbaccalaureate IRTANIH · Starting stipend $44,806 (0 to 1 year of experience)

    Later. Apply about six months before you want to start.

Checked 2026-10-13.

The fuller picture

Risks beyond AI: highGrant-funding risk

For the bench half of this degree, grant policy matters more than any model. Academic lab jobs are paid from grants, and NIH is making fewer new awards even though its budget held; biotech hiring runs in its own layoff cycles. For the scribe half, AI is the bigger force.

Where this could be wrong
  • The blend is ours. If graduates from your school mostly go to industry QC, the share of your week that holds up rises. If they mostly scribe, it falls.
  • Scribe exposure may be overstated in the near term. Adoption is uneven, and studies show physicians still editing AI notes. The tools can do the work, but employer policy decides the pace.
  • NIH grant counts come from mixed sources and are not final. The direction is clear, but the size is still uncertain.
  • Bench automation may move faster than we expected. Autonomous labs ran tens of thousands of experiments in 2025 and 2026. If academic labs buy liquid handlers widely, bench work becomes work you do alongside machines rather than by hand.

This is an example. Your plan uses your own week.

Make my plan

Also matched: biology student, biology major, bio major. Research checked 2026-10-13. Page updated .