News for hackers. Stuff that matters.

A comment on The Dual LLM pattern for building AI assistants that can resist prompt injection.

Re:The Dual LLM pattern for building AI assistants that can ... by hackernewds on Monday May 15, 2023 @05:12AM (#35945495)
One need only beat level 2 of gandalf.ai to know that this level of security is hilariously insufficient
[ Parent | on HN ]
Re:The Dual LLM pattern for building AI assistants that can ... by mklond on Monday May 15, 2023 @11:49AM (#35950189)
Gets trickier at the higher levels, but all of Gandalf's defenses are hand crafted at the moment. Can probably be made much more secure. Lots of interesting discussions happening here: https://news.ycombinator.com/item?id=35905876
[ Parent | on HN ]