Data compiled by Indeed show that US science job postings requiring artificial intelligence skills are climbing sharply while overall science hiring drops, Nature reported.
Quadram Institute chief executive Daniel Figeys said researchers at the Norwich-based food science centre must demonstrate working knowledge of machine learning to handle data from high-throughput screening. Figeys noted that candidates lacking those skills would raise a "red flag" during hiring decisions.
Mayo Clinic dean of research Vijay Shah said recruiters prioritize candidates who independently learn machine-learning concepts over applicants trained on specific tools. Shah, whose Minnesota institution had nearly 30 open research positions in May, warned that benchmark tests tied to platforms such as OpenClaw or AlphaGenome quickly become obsolete. TileBio chief executive Christopher Walsh similarly advised candidates to build curiosity and use language models to learn new material while verifying sources.
Massachusetts Institute of Technology computer scientist Regina Barzilay warned that scientists must verify statistical outputs from automated models rather than accepting probability figures at face value. In unpublished work at MIT, AI specialist Aziz Ayed and his colleagues found that the breast-cancer screening tool MIRAI miscategorized patient risk levels, failing to flag some women for advanced screening.
Moderna engineer Sydney Pham uses machine-learning models to determine testing parameters for pharmaceutical manufacturing after completing Barzilay's course in 2024. Pham corroborates automated recommendations through secondary testing to understand how systems arrive at conclusions. University of Oxford consultant Dominik Lukeš noted that researchers must study how models interact with datasets to avoid erroneous results.
Cancer Research UK head of prevention and early-detection research Talisia Quallo said funding bodies require grant proposals to detail exact algorithm mechanics and operational limits. Quallo explained that review committees favor applicants with prior machine-learning publications or teams that incorporate computational specialists, citing Ke Yuan's AI for Cancer Research group as an example.
