Everybody needs a personal AI policy. Just ask Hank Green.
TL;DR
YouTuber and science communicator Hank Green has admitted leaning too heavily on AI as a research aid for his work. It gave him fast access to papers he did not know existed, but it also narrowed his own engagement with each topic. Green stresses that the words in his videos are his own, while describing his relationship with AI as unhealthy and bad for his work. The case shows how a useful research shortcut can quietly turn into a dependency.
Nauti's Take
Green's admission is valuable because it names the upside honestly: AI gave him fast access to a large body of papers, and that advantage stays real for any research workflow. The same case shows the risk, since the shortcut quietly turned into a dependency that crowded out his own engagement with the material.
Anyone doing heavy research should decide in advance what AI handles and where primary source checking starts, then log the papers they actually read.
Video
Sources