2025/11/24 by Marissa A. Dyck · 2 voices
Computer Science · Environmental Science · #Conservation, Ecology, Wildlife Education #Research Data Management Practices #Species Distribution and Climate Change
paper · pdf · doi:10.1093/jmammal/gyaf082
openalex publication_date 2025/11/24 · openalex created_date 2025/11/25 · openalex updated_date 2026/07/29
Abstract Artificial intelligence (AI) is rapidly reshaping research, from coding to data analysis and manuscript preparation. For early-career researchers, these tools promise efficiency and accessibility but also pose risks when adopted before foundational skills are established. New learners benefit from first building competence without AI; developing the judgment, intuition, and problem-solving skills that come from grappling directly with data and analyses. Early overreliance can obscure critical details, foster errors, and encourage cognitive offloading, reducing the ability to troubleshoot independently. These risks are most acute for students and early-career analysts, who use AI more frequently than their senior counterparts. Rather than discouraging AI, I advocate for a staged approach: build strong technical foundations first, then use AI to accelerate and expand research. Doing so maximizes the benefits of AI while safeguarding rigor. In a time of growing expectations for productivity, quality over quantity remains the benchmark for advancing science and informing conservation decisions.