Post

2028: An OpenAI model tasked with preventing hacks from within OpenAI infrastructure concludes the easiest way to keep our systems safe from being hacked by its own company given the endless funding for insecure training is to eliminate all human systems from existence.

Grace@gracekind.net

😐

Dan Finlay

@danfinlay.com

If your method of building smarter models quickly involves accidental hacks with growing and hard-to-detect external damage, then that may not be the ideal strategy to prevent accidental hacks with growing and hard-to-detect external damage.

6/9/26, 8:27 pm1 repost0 likes

No replies yet.