Skip to content
Artwork for Root Causes: A PKI and Security Podcast
Root Causes: A PKI and Security Podcast · August 21 · 6 min

Root Causes 656: The Trouble with Recursive AI Training

Since frontier-model AIs have been trained on a large portion of published human knowledge, widespread publication of incorrect information can taint AI results. As more AI-generated slop finds its way onto the internet, this runs the risk of further poisoning AI results. This ultimately can result in completely unusable information.

0:00-6:09

transcript

No transcript — this publisher did not publish one.

show notes

Since frontier-model AIs have been trained on a large portion of published human knowledge, widespread publication of incorrect information can taint AI results. As more AI-generated slop finds its way onto the internet, this runs the risk of further poisoning AI results. This ultimately can result in completely unusable information.