AI research papers are getting better, and it’s a big problem for scientists

The Verge describes Peter Degen investigating unusual citations to a 2017 paper: it rose from a few dozen citations over several years to being cited every few days, while the RSS snippet does not disclose the full sample size or review findings.

Why it matters: HKR-H/K/R all pass: the paradoxical angle, named investigation, and citation spike give it signal. The post lacks full sample size, so it stays in the lower featured band rather than becoming must-write.