The Turing Test – Review
25 Aug, 2016
In The Turing Test, a research team on Europa has gone missing, and you must now solve puzzles and find out what happened. Here’s what we think about it.
In 1950, Alan Turing, the father of artificial intelligence, proposed a test to determine if machines can think the way humans can. Called the Turing test, it involves a human judge who converses with a human and a machine, without knowing if the entity they’re talking to is a human or a machine. If the machine can successfully convince the judge that is a human, based on its text responses, it is said to have passed the Turing test.
Fast forward to the 23rd century. Ava Turing is an engineer in stasis on board a spaceship orbiting Jupiter’s moon, Europa. The research team on Europa has seemingly gone missing and has retreated into the depths of the moon. Ava is woken up by an artificial intelligence named T.O.M. and off she goes on a journey to find out what exactly has happened to the ground crew.
What should be a straightforward waltz through the research base proves to be something of a challenge, however, as the base has been repurposed into a linear collection of puzzles. This, according to T.O.M., is a Turing test, designed to make sure you’re human. Will you be able to pass it?
While still on the ship, you get armed with the E.M.T., the engineering tool that will be your companion throughout the game (apart from T.O.M.’s grandfatherly baritone, of course). With a name like that, you’d think it’s probably a matter-dissolving ray gun like the Multi-Tool in No Man’s Sky. Instead, all it does is collect up to three energy ‘orbs’ that are used to power various throughout the Europa base. You know, things that would be useful on any extraterrestrial research base, such as pressure pads, massive anti-gravity magnets and airborne power beams that look like a string of pearls.
Armed with the E.M.T., you’ll mostly be turning power switches on and off by transferring energy orbs or power boxes in the game’s puzzle sectors. And of course, the E.M.T. looks like a gun, in case you’re having first-person shooter withdrawal symptoms. As the game stresses, these puzzles require lateral thinking, which means that an AI like T.O.M. cannot solve them. It would be amusing, then, if one day, someone specifically develops an AI that can solve The Turing Test’s puzzles.
But therein lies the philosophical debate that T.O.M. and Ava discuss across dozens of puzzle rooms. A machine programmed in a certain way can simulate behaving like a human, and accomplish whatever goals can be set forth to prove that it is ‘human’. But is it the same thing as the real deal? Can it ever think like humans do?
A recurring motif the game uses to point this out is the Chinese Room thought experiment (which the excellent indie development studio takes its name from). In the Chinese Room, a person is inside a closed room, receiving messages in Chinese, a language the person does not understand. Using a reference book, however, they can respond to the messages using Chinese characters, and convincingly hold a conversation with the person outside the room, despite not understanding Chinese.
The Turing Test loves to wax philosophical like this, questioning whether morality should be governed by empathy or logic, whether artificial intelligence and human intelligence are similar in nature, as well as whether free will exists—and if it does, whether it is even desirable at all. This philosophising does not ever translate to any actual gameplay, of course, which follows in the good old Portal tradition of throwing you into a virtual mouse cage and having you solve a puzzle to get to the next room.
Back when Portal was released around nine years ago, it directly addressed this by having its token sinister AI, GLaDOS, repeatedly demean the player character, Chell. This was then subverted by Chell breaking free of the AI’s grasp and struggling to escape (even if that meant mostly going through more puzzle rooms, only less polished and clean than before).
The Turing Test uses a different tack. Rather than treating the Ava as a human guinea pig, it strokes the player’s ego by emphasising how it requires a human being’s creativity to solve these puzzles. T.O.M. being an ever-watchful companion and not the creator of the puzzles makes for a curious relationship between (wo)man and machine.
This set-up sounds like it should get rather rote, especially considering T.O.M. is no GlaDOS. And yet, surprisingly, the conceit works. At the end of each 10-sector chapter, you get a deeper glimpse into the ground crew’s lives as well gain hints into what might have happened to them. The pace at which you’re rewarded with new information felt just right.
Apart from this, The Turing Test sprinkles each chapter with an optional puzzle room. I came to really enjoy these optional puzzles, as they required some really imaginative thinking. Although I beat all the standard puzzles, a couple of the optional puzzles continue to niggle at the problem-solving section of my brain.
The game also excels at adding new puzzle mechanics at just the right pace. Rather than bunching them up together in batches of same-y puzzles, the mechanics are spread out over the game, keeping the gameplay fresh. In the eight hours it took me to complete the game, I don’t recall using the same trick to solve two puzzles, except when a puzzle actively built upon that trick.
At the end of each level, you bathe in the satisfying glow of what appears to be a decontamination corridor of some sort. It’s never really explained what it does, but the satisfying thump of magic blue lasers scanning Ava never got old. If the player is a mouse in a cage, this is the crumb of cheese you get at the end of it.
The endlessly-clean and reflective surfaces of the Europa base avoid looking too Portal-y by adding a splash of colour. Even if the research scientists were in a hurry to construct the Turing test, they certainly accounted for their creation looking pretty. Strategically-positioned lights bathe rooms in vivid colours, and combined with the reflective surfaces, have a weirdly drinkable quality to them. And expect lens flares. Lots of lens flares.
The game’s voice acting can be hit or miss, with Ava’s voice actor being somewhat unconvincing, particularly in the early stages of the game. For someone woken up in an emergency and rushed to a moon, Ava has the uncanny knack of talking like she’s reading a script. T.O.M. on the other hand, is played masterfully, commanding a voice that matches its personality: not quite authoritarian, not quite submissive either and also not quite the wise and knowledgeable senior. T.O.M. is not quite anything—something it knows and is tormented by.
The Turing Test is polished—both literally and figuratively. The puzzles are never overwhelming, and the game’s intriguing, hard sci-fi story is told with a suitable air of mystery. In the end, it delivers a satisfying yarn while upturning thought-provoking questions about the nature of thought, understanding, the mind, and whether it is better to use a red orb or a blue orb for this socket.
This review is based on a review copy we received from the game’s developer. The Turing Test will release on Steam (Windows only) and Xbox One on 30 August. It is priced at $19.99, £14.99 and €19.99, but there’s a discount if you pre-order on Steam. On Xbox One, the UK price is £15.99.