To each their own though.
I've debated this at length with people, because I know it's a popular film, but felt the same way after re-watching it.
It pierces my suspension of disbelief that AI can be created that simultaneously (a) simulates complex behavior & (b) completely lacks self-reflection.
Either give Caleb a character flaw (thus justifying his implicit murder by Ava) or have her successfully escape, reflect in the helicopter, turn around, and end the movie with her conflicted after re-landing at the house.
Otherwise, I don't see how the ending doesn't reveal the entire plot as a Shyamalan-esque mechanical Turk (ha! or inverse, as the metaphor would be). A filmmaker got us to believe a lie by omitting any contrary evidence? How difficult!
But I'm honestly not talking about Star Wars droids who are regarded as just intelligent enough to be useful but still expendable.
I'm talking about AI which are so much more intelligent than humans being placed in human suits and outwitting humans by convincing them they feel something they do not.
That's not to say that they don't actually have emotions. They might, as is the case in Westworld. But the emotions they show humans are entirely "perfect acting" used to achieve a goal.
In Ex Machina that goal winds up being surprisingly mundane, which I really loved.