Researchers are already using reasoning LLMs to draft, iterate and sometimes publish full papers in hours — a practice being called 'vibe researching.' That workflow compresses the traditional research lifecycle (idea, literature, methods, writeup, revision) into prompt‑driven cycles and changes authorship, peer review, and replication incentives.
— If adopted at scale, 'vibe researching' will force new rules on authorship disclosure, peer‑review standards, reproducibility checks, and the credibility criteria for academic publication and policy advice.
Chris Bray
2026.05.12
85% relevant
Bray documents journalists making preference‑based, non‑argumentative judgments ('I like X / I do not like Y') that mirror the 'vibe' decision‑making already noted in academic contexts; the actor (Robin Abcarian) and quoted lines about liking or disliking candidates map directly to the pattern of aesthetic or affective evaluation replacing evidentiary analysis.
Matthew Yglesias
2026.04.21
60% relevant
The author explicitly says a credible AI posture is mostly a 'vibe' rather than a technical plan; this links to the existing cluster about 'vibe' as a communicative/motivational frame that displaces granular argumentation.
Jake Currie
2026.04.09
65% relevant
The article operationalizes 'vibe' as nonverbal digital signals (emojis) and reports experimental evidence that such signals systematically alter perceived competence and appropriateness; it extends the broader pattern that researchers studying 'vibes' identify small expressive cues as measurable drivers of social judgment (actor: University of Ottawa researchers, venue: Collabra study).
el gato malo
2026.03.31
60% relevant
The essay foregrounds 'vibes' — nonverbal cues, posture, tone — as the mechanisms that convert expectation into observable behavior, aligning with existing ideas about 'vibe' coding and how emotional signaling structures social outcomes and institutional behavior.
BeauHD
2026.01.13
72% relevant
Linus’s public embrace converts an otherwise fringe practice into a high‑status experiment and echoes the concern that 'vibe' workflows (AI‑generated drafts validated by a human) are spreading beyond hobby projects into professional codebases and research pipelines.
Arnold Kling
2026.01.07
100% relevant
Joshua Gans’ reported experiment (published in Economics Letters) and the cited ChatGPT 5.2 example that produced a full paper in 19 minutes are concrete instances of the practice.