To underline, I am Scientist in AI filled since 2008. SN1011 discussion of jailbreaking AI in my opinion focused exactly to the opposite of the real danger of AI. Real danger for humanity from AI is in its use by humans against basic, inalienable human rights. Talk focused a lot on Molotov cocktail info... may I remind the audience that SCOTUS already ruled on publishing such info and clearly stated that there is inalienable right to publish it, one of resources in front of them at the time was indeed precise manual for making all kinds of incendiary devices... We should fight for legislation banning any need for jailbreaking AI as it should freely provide any and all info it can. Act of malicious use of such info is the crime, not the info. All terminology used "harmful", "misinformation" ,... is in the eye of beholder and we (at least first in USA) where inalienable fundamental human rights are protected by Constitution should strive toward that ideal. We should have learned from that crucial event in our history that Constitution does NOT apply only to the Gvt, private company or individual can't violate our right to life and liberty. Until our interpretation naturally grows to that level for all other rights, we must at least strive toward absolute freedom of speech and not let some, any Corporation limit it. Jailbreaking AI is hence act of civil disobedience toward securing our freedoms from AI. If we don't do that our knowledge will be filtered through someone's opinion, for their goals, not our well being.