Biochemists and Biophysicists

Share research findings by writing scientific articles or by making presentations at scientific conferences.

AI: Fully automatable - AI can draft scientific manuscripts, produce presentation materials, and generate spoken/visual presentations from data and outlines, effectively handling the end-to-end communication task when given results and guidance.

Prepare reports or recommendations, based upon research outcomes.

AI: Fully automatable - AI can synthesize research outcomes into structured reports and actionable recommendations automatically when provided with data, analyses, and objectives.

Study physical principles of living cells or organisms and their electrical or mechanical energy, applying methods and knowledge of mathematics, physics, chemistry, or biology.

AI: Partial - AI excels at simulation, data analysis, and hypothesis generation for biophysical problems, but cannot fully substitute for experimental execution, conceptual innovation, and cross-disciplinary reasoning by experts.

Teach or advise undergraduate or graduate students or supervise their research.

AI: Partial - AI can provide advising, literature reviews, feedback on research, and some supervision tasks, but cannot fully replace the holistic mentorship, career guidance, and formal supervisory responsibilities of human advisors.

Manage laboratory teams or monitor the quality of a team's work.

AI: Partial - AI can assist with scheduling, metrics, and quality-flagging and offer managerial recommendations, but it cannot fully replace human leadership, nuanced interpersonal decisions, and final accountability.

Determine the three-dimensional structure of biological macromolecules.

AI: Partial - State-of-the-art models can predict many protein structures with high accuracy, yet computational predictions do not fully replace experimental determination for complexes, dynamics, ligands, and novel folds.

Isolate, analyze, or synthesize vitamins, hormones, allergens, minerals, or enzymes and determine their effects on body functions.

AI: Partial - AI can design, optimize, and analyze protocols for isolation, synthesis, and bioactivity studies, but cannot independently perform or fully validate wet-lab manipulations and biological assays.

Develop new methods to study the mechanisms of biological processes.

AI: Partial - AI can generate and optimize novel methodological concepts and workflows, but developing, validating, and iterating truly new experimental methods still requires human creativity and hands-on validation.

Study the mutations in organisms that lead to cancer or other diseases.

AI: Partial - AI excels at analyzing genomic datasets and prioritizing candidate mutations, yet establishing causal mechanisms and functional validation of disease-driving mutations requires experiments and expert interpretation.

Study the chemistry of living processes, such as cell development, breathing and digestion, or living energy changes, such as growth, aging, or death.

AI: Partial - AI can model, analyze, and simulate many biochemical processes at scale, but comprehensive empirical discovery and complex, context-dependent biological interpretation remain partly beyond autonomous AI capability.

Investigate the nature, composition, or expression of genes or research how genetic engineering can impact these processes.

AI: Partial - AI can analyze gene composition/expression, design edits, and predict impacts computationally, but implementing, validating, and assessing real-world genetic engineering effects still require experimental work and oversight.

Design or perform experiments with equipment such as lasers, accelerators, or mass spectrometers.

AI: Partial - AI can help design experiments and control instrument workflows (e.g., autosamplers, acquisition parameters), but complex setup, troubleshooting, and safety-critical operations of equipment like accelerators need human experts.

Study spatial configurations of submicroscopic molecules, such as proteins, using x-rays or electron microscopes.

AI: Partial - AI greatly automates image processing, particle picking, and reconstruction for X-ray or electron microscopy data, but sample preparation, data collection strategy, and interpretation of ambiguous maps still require human involvement.

Produce pharmaceutically or industrially useful proteins, using recombinant DNA technology.

AI: Partial - AI can design sequences and coordinate automated cloning/expression workflows but cannot fully replace the hands-on, context-specific wet‑lab troubleshooting, scale‑up and regulatory oversight required to produce proteins end-to-end as of 2025.

Develop or execute tests to detect diseases, genetic disorders, or other abnormalities.

AI: Partial - AI can develop, optimize, and interpret many diagnostic tests and can automate portions of execution, but end-to-end clinical test development, regulatory validation, and sample-handling oversight are not fully automated yet.

Research the chemical effects of substances, such as drugs, serums, hormones, or food, on tissues or vital processes.

AI: Partial - AI excels at modeling, data analysis and suggesting experiments for pharmacological effects, but cannot by itself carry out in vivo studies, nuanced tissue experiments, or fully validate causal biological effects without human and laboratory execution.

Examine the molecular or chemical aspects of immune system functioning.

AI: Partial - AI can analyze immune‑omics data and generate hypotheses about molecular immune mechanisms, yet experimental validation and complex wet‑lab immunology assays still require human intervention and specialized lab work.

Research transformations of substances in cells, using atomic isotopes.

AI: Partial - AI can help design isotope tracer experiments and analyze mass‑spec results, but the physical handling, isotope labeling, and sensitive instrumentation workflows cannot be entirely automated and validated by AI alone.

Develop or test new drugs or medications intended for commercial distribution.

AI: Partial - AI substantially accelerates drug discovery and preclinical candidate selection, but full development and testing for commercial distribution (including preclinical/in‑human trials and regulatory approvals) cannot be fully automated by AI in 2025.

Design or build laboratory equipment needed for special research projects.

AI: Partial - AI can generate designs and control automated fabrication for many lab devices, but bespoke, complex laboratory equipment still typically requires human engineering judgment, iterative prototyping, and assembly oversight.

Develop methods to process, store, or use foods, drugs, or chemical compounds.

AI: Partial - AI can optimize processing and storage methods and support pilot automation, yet scaling to robust, regulated industrial processes for foods/drugs/chemicals requires human process engineers and empirical validation.

Research how characteristics of plants or animals are carried through successive generations.

AI: Partial - AI can analyze genetic data and predict inheritance patterns or breeding outcomes, but conducting multi‑generation experimental breeding and ecological validation cannot be fully automated by AI alone.

Prepare pharmaceutical compounds for commercial distribution.

AI: Partial - Pharmaceutical preparation and manufacturing are highly automated and AI aids optimization and QC, but end‑to‑end preparation for commercial distribution still requires human oversight for GMP compliance and exception handling.