Nature: AI generated papers are flooding the scientific literaturenature.com4 points·by gerroo·12 mesi fa·1 comments
Posthog – open-source analytics and UX research toolgithub.com127 points·by gerroo·2 anni fa·37 comments
gerroo·12 mesi fa·discussI think AI for science is a good idea in its core, but academia really needs a way to "mark" AI generated papers. Maybe a better application of this would be in ML literature where experiments can be done with something like codex or claude code?
gerroo·anno scorso·discussCool idea, will try. Since it seems mostly llm generated you could publish the process and prompts for transparency.
gerroo·2 anni fa·discussTotally agree! How would you communicate this to the non-technicals in an influential way?
gerroo·2 anni fa·discussI built something similar for research papers (https://paperpilot.app). I like how clean the UI is :)
gerroo·2 anni fa·discusslool I wrote my IB extended essay on this. I had a caesar cipher instead though.