Stanford professor declares paper-based academic evaluation dead, sparking debate

On September 20, Stanford professor Anshul Kundaje argued in a thread that papers are effectively dead as a measure of academic productivity, expertise, and achievement: acceptance at top venues is rapidly approaching randomness while the review time needed to maintain standards grows exponentially, so traditional gatekeeping has failed. He urged the academic community to seriously and urgently discuss and test alternative evaluation systems, prompting responses from multiple scholars.

Confirmed

• Kundaje clarified he is not against recording and publishing research; he opposes papers as the central currency of academic evaluation.
• His thread opened with rhetorical questions: whether every paper submitted and published over the past 30+ years was truly gold-medal quality; whether anyone has actually read (methods sections included) all their pre-2024 papers; why top PhD admissions hinge heavily on paper counts; and whether researchers without a stack of glossy papers can even get faculty interviews at top departments.
• He suggested the review system may have been misaligned from the start, naming chronic practices—cherry-picking favorable results, overselling conclusions, hype, dishonesty about limitations, not sharing data and code—and questioning whether paper counts, citations, h-index, and venue prestige are the right currency for identifying talent.
• In discussion with Stanford statistician Jessica Hullman, he said he is not against improving the current system, but most proposals he has seen are desperate short-term bandaids; he welcomed discussion of any approach.
• He argued the system was never designed for the drastic changes of the past year—it was already strained, and now "the dam has broken and the flood is advancing," requiring a better, stronger rebuild.
• Hullman responded that while many are certain the system cannot hold, she believes it is too early to conclude either way.
• Scholar sokrypton offered a simpler fix: once gatekeepers in admissions and promotion stop judging mainly by "glam papers," the problem largely eases.

Why it matters

• With AI-era paper volume surging, near-random acceptance and exploding review costs threaten to strip paper counts and venues of signal value, directly affecting hiring, promotion, and resource allocation.
• The debate signals academia beginning to confront structural failure in traditional publishing-based evaluation and to explore contribution- and capability-centered alternatives.

2026-09-20 ~ 2026-09-21 · 16 related posts

• [source] Stanford prof Anshul Kundaje: papers are finished as a measure of academic productivity — anshulkundaje · 2026-09-20
• Stanford's Kundaje urges academia to invent new ways to evaluate scientific impact — anshulkundaje · 2026-09-20
• [source] Stanford professor: papers are dead as a measure of expertise, academia must rebuild evaluation — anshulkundaje · 2026-09-20
• AI researcher: academia's glam-paper obsession dies the day admissions stop rewarding it — sokrypton · 2026-09-20
• [source] Scientists debate alternatives to academic publishing as AI shakes the system — TuhinChakr · 2026-09-20
• Stanford professor calls most peer review fixes 'desperate short-term bandaids' — anshulkundaje · 2026-09-20
• Professor: the academic system's 'dam is broken' amid AI-driven upheaval — anshulkundaje · 2026-09-20
• Were 30 years of published papers all gold star? Professor opens critique thread — anshulkundaje · 2026-09-20
• Why do elite PhD programs judge admits by paper count? Professor asks — anshulkundaje · 2026-09-20
• No glam papers, no faculty interview? Professor's thread on academic gatekeeping — anshulkundaje · 2026-09-20
• Professor calls out cherry-picked results, hype, and withheld code in academia — anshulkundaje · 2026-09-20
• Did anyone actually read the methods sections? Professor's peer review jab — anshulkundaje · 2026-09-20
• Stanford professor questions whether peer review system was misaligned all along — anshulkundaje · 2026-09-20
• Plugging holes with fingers while the dam breaks: professor on academia's denial — anshulkundaje · 2026-09-20
• Stanford professor: papers are dead as a measure of productivity as top-venue odds go random — anshulkundaje · 2026-09-21

1 near-duplicate retellings: anshulkundaje