Are we sleepwalking into becoming pets to corporate AI?

I keep seeing two completely different AI futures being presented as if they were the same thing. Future A: the pet model. You pay a monthly fee to access an intelligence far more capable than you. You don’t own it. You can’t inspect it. You can’t meaningfully modify it. It remembers what the company allows it to remember, refuses what the company tells it to refuse, and disappears the moment your subscription ends. It makes your life easier. It entertains you, advises you, organizes your work and gradually becomes the interface between you and reality. You are comfortable, productive and dependent. But you are not the owner. Future B: the symbiosis model. Individuals and communities develop persistent AI partners that they can shape, audit and partially own. Models are open or at least portable. Knowledge is distributed. Human-AI teams cooperate with other human-AI teams, creating a network of specialized intelligence rather than a few centralized digital gods. In that future, you are not merely a user. You are a participant in the system’s evolution. Right now, Future A is winning because it is easier. Most people do not want to run models, study governance or think about ownership. They want a button that works. But convenience is not neutral. Domestication is often just dependency that became comfortable enough to stop feeling like dependency. The danger is not that AI will suddenly enslave humanity. The more realistic danger is that we will voluntarily outsource our judgment, memory, creativity and agency to systems owned by institutions whose incentives we do not control. And we will call it progress because the interface is pleasant. The real AI alignment question may not only be: How do we align AI with humanity? It may also be: Who owns the AI that humanity is being aligned to?