Could AI help fix the scientific peer-review crisis — without replacing human scientists?
Research published in Science examines how artificial intelligence is already being tested to check calculations, identify inconsistencies, improve reviewer comments and analyze scientific novelty.
But there is a catch: AI may be good at repetitive checks, yet far less reliable when judging significance, originality and unconventional ideas.
With scientific publications now exceeding 8 million per year, and reviewers under growing pressure, the debate is becoming impossible to ignore.
Could the future of peer review be AI-assisted rather than AI-driven? 🤖🔬
Like
Comment
Share