There is a particular kind of frustration that keeps appearing in accounts of people who create with AI tools. It is not about quality. It is not about credit. It is about not being able to tell what happened.
A prompt that worked yesterday produces something different today. A style that was possible last month is no longer available. A piece of work gets flagged, removed, or deprioritized, and no explanation follows. The system changes. The rules shift. The ground moves.
I have been studying this experience in AI art communities, and what strikes me is not the frustration itself but what it reveals. These are not bugs. They are features of how AI platforms are governed, through opacity, through frequent and unannounced change, through terms of service that reserve the right to alter anything at any time. The instability is structural.
This matters beyond platform critique.
When a system is designed so that users cannot reliably predict its behavior, cannot understand why it made a particular decision, cannot know what it will do next, that system is shaping something more fundamental than user experience. It is shaping epistemic conditions. What people can know. What they can trust. What they can hold accountable.
And here is where I think the digital minds conversation needs to go further than it currently does.
Most debates about AI consciousness and moral status focus on the internal question: what is happening inside the system? Is there experience? Is there something it is like to be this model? These are serious questions. But there is an external question that receives less attention: what are AI systems doing to our capacity to recognize and reason about minds at all?
If the systems we interact with are deliberately designed to be unreadable, if their behavior is unstable, their decisions unexplained, their logic hidden, then we are being systematically prevented from developing the kind of sustained, accurate perception that moral recognition requires. We cannot ask whether AI deserves moral consideration if we are structurally prevented from knowing what AI is doing.
Governance through instability is not just a platform problem. It is an epistemic problem. And it is, I would argue, a problem for anyone who takes seriously the question of how society should prepare for the possibility of digital minds.
You cannot build frameworks for something you are not allowed to understand.