Machines Of Loving Grace
It's mostly set in a modernist one-man research facility in an isolated estate in Alaska, where a young coder, Caleb Smith (Domhnall Gleason), arrives after winning a lottery that gives him face-time with the boss of his search-engine company, Nathan. Played by the ever-versatile Oscar Isaac (Inside Llewyn Davis, A Most Violent Year) with a shaven head and hipster beard, Nathan embodies all the manic, macho, super-controlling vices of Silicon Valley's alpha-geekdom. He gives Caleb a pass that allows him access only to certain areas of the house, and tells him that the lottery was a sham: he's actually been recruited to test Nathan's latest creation, Ava, an android that may or may not be self-aware enough to pass for human. With only the overbearing, manipulative Nathan and his silent servant/bed companion Kyoko (Sonoya Mizuno) for company, Caleb soon falls for the coolly sympathetic and beautiful Ava (Alicia Vikander), assumes the role of her white knight, and enters into a contest of wills over her freedom.
The film's examination of the nature of intelligence is capable rather than novel, but its story is furnished with cleverly unsettling twists and revelations, and there are some nice touches in the genesis of Ava - her silicon-sponge brain runs on data derived from Nathan's vast search engine, for instance, so that she embodies and has access to all of the desires and appetites of its users. And although she's the object of the contest between the two men, Alicia Vikander's delicately assertive portrayal of Ava steals the film. Moving with controlled and slightly inhuman grace, displaying unsettling sharp insights into human behaviour, Ava is an ethereal, poignant Miranda whose yearning for the whole wide world beyond the confines of her glass and concrete cage captures the sympathies of both naive Caleb and the audience.
Alicia Vikander's performance is aided by CGI that seamlessly marries flesh with plastic and metal, a terrific example of state-of-the art visual manipulation. Given the ongoing difficulty of developing actual artificial intelligence, and recent advances in wearable virtual reality, perhaps the most pressing question technology presently poses is not how to spot machines passing for human, but how to distinguish illusion from the actual.