The Horizon of AGI
General intelligence is a disputed threshold, not a settled product label.
The Mechanism: Artificial general intelligence refers to a hypothetical system able to perform a wide range of cognitive tasks across unfamiliar domains. No agreed test establishes when breadth, transfer, or autonomy crosses that threshold.
The Philosophical Pivot: Calling a system intelligent changes how people distribute trust and responsibility, even before anyone agrees what intelligence means. A label can describe capability, advertise ambition, or quietly invite moral recognition.
The Horizon: If systems become more capable, human worth cannot rest only on being the smartest problem-solvers in the room. Institutions will have to distinguish competence from consciousness, and both from permission to decide for others.
A further shift may be social rather than technical: people could begin treating machine judgments as more impartial than human ones simply because they are difficult to inspect. That appearance of neutrality can conceal the values embedded in training and deployment.
The question of AGI therefore changes with the institution around it. Capability without accountability can magnify old hierarchies, while public oversight can make powerful systems answerable to the people whose lives they shape.
A threshold we cannot define can still reorganize the world around the people who claim to have crossed it.
If a machine could do most things I value, what would still make a human life matter?