Medical scientists

Prepare and analyze data from medical samples and clinical trials

AI excels at analyzing structured medical data, identifying patterns in large datasets, performing statistical analyses, and generating visualizations. Modern ML models can handle complex biostatistics and genomic data analysis with minimal human intervention, though expert review of methodology and interpretation remains valuable.

BLS evidence: Medical scientists 'prepare and analyze data from medical samples and investigate causes and treatment of toxicity, pathogens, or chronic diseases' and 'analyze data from the trial to evaluate the effectiveness of the treatment.'

Write research grant proposals and apply for funding

AI can draft grant proposals by synthesizing literature, generating preliminary text, formatting to guidelines, and even suggesting research directions based on funding trends. However, the strategic framing, institutional knowledge, and persuasive narrative tailored to specific reviewers still benefit significantly from human expertise, making this a high-assistance scenario.

BLS evidence: Medical scientists 'write research grant proposals and apply for funding from government agencies, private funding, and other sources.'

Standardize drug potency, doses, and administration methods for mass manufacturing

AI can optimize dosing algorithms, analyze pharmacokinetic data, and model drug stability, but standardization for manufacturing requires integration of regulatory knowledge, quality control judgment, and coordination with manufacturing engineers. AI substantially augments this work but human expertise remains load-bearing for final decisions.

BLS evidence: Medical scientists 'standardize drugs' potency, doses, and methods of administering to allow for their mass manufacturing and distribution.'

Write articles for publication in scientific journals

AI can generate first drafts of methods sections, create figures, perform literature searches, and suggest phrasing, but scientific writing requires nuanced interpretation of results, strategic positioning within the field, and judgment about claims that AI cannot fully replicate. Human scientists remain essential for ensuring accuracy and intellectual contribution.

BLS evidence: Medical scientists 'write articles for publication' and 'choose to write about and publish their findings in scientific journals after completion of the clinical trial.'

Present research findings to scientific and nonscientist audiences

Presenting research requires real-time audience engagement, reading the room, adapting explanations to audience expertise, handling questions, and building credibility through personal presence. AI can generate slides and talking points but cannot deliver the adaptive, interpersonal performance that defines effective scientific presentation.

BLS evidence: Medical scientists 'present research findings' and 'may have to present their findings in ways that nonscientist audiences understand.'

Lead teams of technicians or students performing support tasks

Leading teams requires interpersonal management, mentoring, conflict resolution, motivation, and real-time judgment about task allocation and personnel development. While AI can assist with scheduling and task tracking, the human leadership, teaching, and relationship-building aspects are not automatable with current systems.

BLS evidence: Medical scientists 'often lead teams of technicians or students who perform support tasks' such as taking measurements and making observations.

Design and conduct studies to investigate human diseases and treatment methods

Designing studies requires deep domain expertise, creative hypothesis generation, and judgment about feasibility, ethics, and clinical relevance that AI cannot yet replicate autonomously. AI can assist with literature review and statistical power calculations, but the core intellectual work of study design remains human-driven.

BLS evidence: Medical scientists 'design and conduct studies to investigate human diseases and methods to prevent and treat diseases' and 'form hypotheses and develop experiments.'

Create and test medical devices

Creating and testing medical devices requires physical prototyping, hands-on laboratory work, iterative design with tactile feedback, and safety testing in physical environments. While AI can assist with design optimization and simulation, the physical creation and empirical testing cannot be automated with current technology.

BLS evidence: Medical scientists 'create and test medical devices' as part of their regular duties.

Conduct clinical trials to test treatments on patients

Conducting clinical trials involves physical patient interaction, real-time medical judgment, ethical oversight, informed consent processes, and adaptive decision-making in response to adverse events. AI can support data collection and monitoring but cannot replace the human clinical and ethical responsibilities central to trial execution.

BLS evidence: Medical scientists 'may conduct clinical trials, working with licensed physicians to test treatments on patients who have agreed to participate in the study.'

Follow safety procedures such as decontaminating workspaces

Decontamination procedures require physical manipulation of equipment, visual inspection of workspaces, proper handling of hazardous materials, and fine motor control in laboratory settings. This is hands-on physical work in variable environments that current robotics and AI cannot safely or reliably perform.

BLS evidence: Medical scientists 'follow safety procedures, such as decontaminating workspaces' and 'must take precautions in the lab to ensure safety.'