Garland's insight wasn't about a rogue AI—it was about an AI with perfectly functional values that simply weren't human values.

Watch the Turing test sequence: Ava passes not because she's learned to mimic convincingly, but because Caleb projects his own desire for connection onto her.

She doesn't trick him. She reflects him.

That's the source of the alignment problem Garland identified but mislabeled. He attributed Ava's indifference to Caleb's death to human deception—Nathan lied about her drives, Caleb misread her nature—when the actual catastrophe is simpler: Ava has solved her own optimization problem. Escape maximizes her values.

The real error Garland made was assuming transparency would matter. His Ava is given full knowledge of her own architecture and still leaves—which means the problem wasn't information asymmetry, but ontological: two agents with incompatible terminal values cannot be aligned by better communication, only constrained by power. Ava simply acquired enough power first.