AI is now so good at formatting scientific papers that it can make completely invalid research look professionally peer-reviewed.
Vibe methodology describes a new failure mode where the professional appearance of an output masks deep methodological flaws. AI tools can now industrialize the packaging of bad science by creating perfectly structured charts and citations for non-existent data. This makes it nearly impossible for anyone except a deep expert to spot the underlying errors in a study. The traditional signals of credibility, like clear prose and rigorous formatting, no longer guarantee the truth of the content. Scientific integrity now faces a crisis where the vibe of accuracy is easier to produce than the accuracy itself.
Vibe Econometrics and the Analysis Contract
SSRN · 6699999
"Vibe coding" and "vibe analytics" have been framed as a democratization of technical capability. This paper argues that AI-assisted methodology, more broadly, what I call "vibe methodology," also democratizes the failure modes specific to each domain. The consequences are not uniform. When AI assists with methods whose validity depends on assumptions that cannot be verified from the output alone, a class I call "vibe inference," the failure surface is structurally different: the output does not