InFractal Cognition LayerbyMyojin Research·Nov 1AI Doesn’t Have to Lie to Change Your PastWhen artificial intelligence begins telling the story of your life, it may also influence the person you are becoming.A response icon1A response icon1
InArtificial Intelligence in Plain EnglishbyMyojin Research·Oct 25What Are We Actually Calling Secure? Lessons From PHP2ShellBefore declaring a system safe, ethical hackers should ask a different question: What assumptions make that judgment possible?
InArtificial Intelligence in Plain EnglishbyMyojin Research·Oct 22AI Can Act Before We Understand. But Who Decides?The October 2026 security crisis reveals a deeper shift in human expertise — and why the ability to act is not the same as the ability to…
InFractal Cognition LayerbyMyojin Research·Oct 19AI May Change the Reason a Task ExistsAs AI agents become capable of doing more work, some of the assumptions behind that work may begin to disappear.A response icon1A response icon1
InArtificial Intelligence in Plain EnglishbyMyojin Research·Oct 10Where Understanding Used to Live:How AI Is Quietly Removing the Need for Persistent Internal StateA response icon1A response icon1
InArtificial Intelligence in Plain EnglishbyMyojin Research·Oct 8The AI Safety Failure That Doesn’t Look Like a RefusalSome apparent limits of AI may come from how a system interprets its task, rather than what it is capable of doing.A response icon1A response icon1
Myojin Research·Oct 8Once a Frontier Model Is Public, It Starts Teaching the CompetitionOpenAI’s latest distillation incident raises a question that goes beyond security: when should a frontier capability become public?
Myojin Research·Oct 2Once a Frontier Model Is Queryable, It Starts Teaching the CompetitionOpenAI’s latest distillation incident suggests that release changes more than access: it can reduce how much competitors need to discover…A response icon1A response icon1
InArtificial Intelligence in Plain EnglishbyMyojin Research·Sep 30When AI Disagrees With You, Who Gets the Final Word?