While on my flight to Boston for ReWorkDL I watched Ex Machina the “must see” latest AI film. I’d been warned that it wasn’t very good by my husband (who’d just flown home the day before!) but I thought that since he’d already seen it, I’d better take the chance to watch it since it’s unlikely to be something we’d watch together in the future. If you haven’t seen it, then please be aware that this post does contain spoilers so read on with caution.
I actually really enjoyed watching Ex Machina, and the special effects were very well done, instilling a sense of realism about the near future of the film’s setting. I’ve got better at being less bothered by the abstractions of science and programming that occur in films and TV and instead try to focus on the message that’s being sent out, and this film does raise some interesting ethical questions on several levels.
We’re dropped in to Caleb’s world as he wins a competition to spend a week at his boss’s house. Nathan, the CEO of the Google-esque company Caleb works for, is a prodigy coder who has been amassing wealth from his skills since 13. We’re given very little background on Caleb and expected to discover him and his intentions as the film progresses. It can leave you feeling somewhat distanced from the character in the first half of the film but does help to see him almost through Ava’s eyes – is he a good person? Ava is the AI – made humanoid and with a very human face – that Caleb is there to test.
Some of the questions Caleb asks himself and Nathan as he’s defining the test are quite inciteful – he comments on the chess computer being good at chess but without any real understanding of what it’s doing, yet he fails to dig deeper into the questions I would have asked in his position when Nathan asks him to back off on the technical side (to avoid the film turning off many of its viewers) such as:
- who did you model Ava’s face on? This does come up near then end of the film, but it was a question I was hoping he’d ask much earlier
- did Ava have any role in choosing the winner of the competition? Again, the choice of winner was resolved but in Caleb’s position I would have asked this early on
- what stimuli has Ava had? knowing that she’d never left her room, how has she learned – was she exposed to multiple soources of language, images? While this is resolved to some extent with the phone hacking revelation, this just gives context to facial expressions and language – not enough for understanding in any AI book
The revelation that Nathan has “hacked the planet’s phones” is fascinating, alongside the similarities of Nathan’s company with Google (“the most popular search engine”), data such as this is already being amassed on us on a global scale and, while it’s possible all our conversations and facial expressions are being monitored at present, I’d say it’s unlikely. But all our searches, our facebook interactions, what adverts we watch and even with Smart TV, which channels we watch, games we play are all being built up into a digital fingerprint of us that can then be used to predict what we want before we know it ourselves. Sometimes code gets this right and sometimes it doesn’t – how often has Amazon recommended something to you that was irrelevant because you’d bought a present for a friend or family member with interests different to your own? Maybe we all ought to do this once in a while to keep the machines guessing :). In any case, we cannot afford to be ignorant about our digital footprint and that it is being stored, and abstracted and used to target us for marketing right now. This is not science fiction and will only get more invasive. Data privacy is a big thing and I suggest that you read up on it as we do need to make decisions about what is private and what we are happy to let go. The safest thing to assume is that anything you do online is publicly available, because security vulnerabilities crop up every day and it seems like there are constant news items of private photos or messages being leaked.
Fairly early in the film I clocked that Keiko was one of Nathan’s AI – her lack of English and (brilliantly acted) slightly artificial movements gave it away. This combined with the voyeuristic elements of the film through the surveillance system gave me the feeling I was watching a clever remake of Swordfish, there would be twists and I was wondering whether Ava and Nathan wore working together or indeed which of them was pulling the strings. As the film progressed and it’s obvious that Caleb is falling for Ava, we’re presented with a question about consciousness. Is Ava self-aware? While she would not feel pain at being switched off, would “formatting” her brain be an act of murder even if the only thing that is removed is memories? We are on a path that will create thinking machines and we need to answer these questions well before we are presented with the actuality of the decision. I think most people, if presented with something they perceive as human, would have difficulty switching it off. Similarly, if presented with something they think is not human then they would have no problem powering it down.
When Caleb discovers Nathan’s horde of ex AI experiments in his bedroom (closets full of “dead” naked female robots being unsurprising about that character) and that Keiko is also a machine he does go a little crazy. I was really hoping at this point that he wouldn’t turn out to be a machine himself as that had been a possibility up until the point he cuts open his arm (perfect background – no family after a car accident, wakes up in hospital – this could easily have been artificial memories and I’m glad that this wasn’t the route that the film took). What I did find interesting was that he set up the escape for Ava by reprogramming the doors before he discovered Keiko and the closets of nudity. However what I do find intriguing is that he punches the mirror after cutting his arm and leaving a pattern similar to the one he sees when he first arrives. Is this an indication of someone previous to Caleb undertaking the same test with an earlier model? Or does it raise a question over whether Caleb is just an AI that bleeds and maybe his memories have been reset, but his reactions are similar…
The escape is intriguing. I really wasn’t sure whether Ava would release Caleb or not right until the moment she steps into the lift. The slight glance she gives him – is that regret, or is she simply a cold calculating machine that has used the tools in front of her to get what she wanted? When she gets into the helicopter that had arrived for Caleb, did the pilot not wonder where his real passenger was? I suppose that he could have been ordered just to collect a “guest” but I was half expecting to see a body on the grass as the helicopter left the valley (somewhat erratically to my eye) (this was the altered-for-the-plane version so forgive me if there was a body in the cinematic version!)
The final scenes with Ava people watching and her long shadow on the ground were, in some respects, frightening – Ava had killed at least one person (possibly also the pilot and probably indirectly Caleb as well) and now walks among us without us being any the wiser.
While seemless human-like robots of the persona synthetics type are far off, they are already in progress – we have many teams working on humanoid AI that will play football, do domestic chores etc, and research in Japan is looking into simulating human facial expressions. With the advent of deep learning and the explosion in the field of that this has provided, I think that we will have humanoid shaped robots with good synthetic voices very soon. Fake skin and our range of movement will be a while afterwards, although with 3d printing and advances in prosthetics for lifelike limbs, I doubt it will be long until the two fields are combined. Welcome to the future.
Ex Machina is available on bluray from 1st June, and the book by Alex Gardner is also available on Kindle. If you want more clever AI and a gradual path for how humans could become the thinking machines then I’d recommend the work of Greg Egan, particularly Axiomatic.