The Turing Test Now Available For Closer Examination

We are all familiar with the Turing Test, in which a computer and a man must engage in a freestyle rap battle to determine which one is “thinking”. But there’s also a videogame of the same name, which is out today. The Turing Test [official site] is a first-person puzzler set in a research base on Europa, in which you play astronaut Eva Turing as she tries to find out where everyone has gone. It’s part of Squeenix’s indie claw machine, the Square Enix Collective, and it looks like this.

That’s very shiny. Most of the game is about transferring power from one thing to another using Ava’s energy spanner, we’re told. But you also have to take control of machines, while taking orders from a robot voice, so yes, there’s a lot of Portal and Talos Principle flavour to it. But it also promises to tell a story about “what it means to be human”. You can see the robo-philosophical direction this is heading. Anway, it’s on Steam now for £14.99/$19.99.

Joe played a demo and liked it, so maybe there’s more here than just another tale of space station abandonment. Of course, if you come away unsatisfied, but are still interested in those kinds of questions (“Why do humans think things?”, “How does my coffee machine know things about me?”) then there’s always SOMA, a much dirtier and wetter exploration of these things, with a stonking ending.

From this site

12 Comments

  1. Babymech says:

    I always found the Turing test pretty easy to beat, perhaps because of my overwhelming humanity: link to tinyurl.com

  2. DrazharLn says:

    > Of course, if you come away unsatisfied, but are still interested in those kinds of questions (“Why do humans think things?”, “How does my coffee machine know things about me?”) then there’s always SOMA, a much dirtier and wetter exploration of these things, with a stonking ending.

    The philosophical questions posed by SOMA are interesting enough, made actually really quite satisfying by how they’re not overwrought. But SPOILERS SPOILERS the ending sequence where the player character doesn’t understand that they’re making a copy of themselves was pretty annoying. OTOH, loads of people on reddit were confused about it, so maybe the developers thought they had to make it clearer that copies of Simon and Catherine would be left behind. SPOILER SPOILER

    • Babymech says:

      The main character in Soma had like 8 opportunities to get the central premise, and never understood it. Also, everything in the game was overwrought. Trying to repeatedly explain the mind-body duality with the pedagogic aid of screaming insane souls trapped in robots is as over as wrought gets.

      • Serenegoose says:

        TOTAL SPOILERS FOR SOMA RIGHT UP INS ALL THROUGHOUT And yet, the game states that old scans (y’know, like the main character) are ‘flat’ compared to people, unconvincing and without the capability for being dynamic, to the extent where if placed in contact with newer scans, the fakeness bothers them. Even after repeated explanation, why should the main character be able to comprehend and internalise new information of that complexity in any way? The game’s already said they’re a pretty unconvincing fake person!

        • Babymech says:

          That could technically be a very good explanation for his flatness, though I don’t think they thought of it. It would also be an incredibly depressing answer – controlling a protagonist who is inherently unable to learn, develop, or improve in any meaningful way. Also, just as annoying in practice as controlling a badly written protagonist.

          MY DUMB THEORY throughout the game was that all of it took place in a therapeutic simulation designed to restore the main character’s brain functions after the accident. In the apartment in the beginning there’s a book on therapeutic simulation, ‘stress testing’ the brain to identify and repair damaged areas. I thought that would have been a neat explanation for the implausible disasters and dangerous scifi scenarios you face, and a nice internalized metaphor for the WAU… but that never panned out.

          • MajorLag says:

            Actually that is a lot more interesting than what the game actually was. Shame the developers didn’t think of it, or at least didn’t go that direction.

    • MajorLag says:

      I watched an LP of SOMA and it confirmed pretty much everything I’d suspected from everything else I heard about it: being bashed over the head with Baby’s First Ship of Theseus while annoying and pointless monsters waste your time.

  3. Urthman says:

    I love the idea of a single-player game asking you to figure out which of the NPCs are supposedly human and which NPCs are just an AI/computer. Sort of like adding a single-player mode to Spy Party.

  4. Monggerel says:

    I thought John Searle pretty much disproved the validity of the Turing Test in 1980.

    • Premium User Badge

      Angstsmurf says:

      I wouldn’t say that the Chinese room really disproves anything. If the person in the room can translate Chinese as if she understood it, what does it matter what our intuitions tell us about whether she “really” understands it or not? The result is what matters.

      • phlebas says:

        At that point the question is no longer whether the person in the room can be said to understand Chinese but whether the room itself can.