The movie S1M0NE, starring Al Pacino, was brought up in another thread. Rather than hijack that other thread, I wanted to address a few ideas here. Yes, "Simone" (Simulation One) was merely a puppet, not an AI. One might argue that all actors are puppets to the director, except that we understand a movie (like many other endeavors) is a collaborative effort. The actors are individual artists in their own right and contribute to the grand picture orchestrated by the director. In the end, the characters portrayed are composites of the writer's colorful script, the actor's skill, the stunt double's daring, the composer's moving score, the spectacle of visual effects, and the editor's nuanced timing. Even the actor-celebrities themselves may be composites off-screen—carefully spun properties in the public eye. That distinction between fact and fancy was the whole point of S1M0NE. Much of the satire had to do with people's responses to the composite, simulated "personality" on screen. (Was Simone really that good, or was she superlative in an eyeglasses-on-the-floor, trendy way?) The Turing Test was proposed as a test of how well a machine can simulate human responses in a text-only chat. Many mistake this as a test of intelligence or "mind," when it was actually meant to illustrate the opposite. (Likewise, Schrödinger's cat was meant as a satire on quantum mechanics, yet many picked it up as the perfect analogy for explaining QM concepts.) The Turing mistake is made by people who confuse the conventions and symbols of a technology for the ideas they represent. For example, cameras of any kind "see" the world very differently from the way a human sees the world. There is a great deal of stylized lighting that must be done to make the image captured by a camera "look natural." Also, the human eye—or rather the brain—motion stabilizes our view of the world. And the "whip-pans" of eye movement are edited out in the same way a film editor cuts together just the relevant images of a scene. Our eyes see the movement of the environment during a head turn, but the "experience" of the blur is suppressed, edited down to irrelevancy. Thus, even the cuts, wide shots, closeups and other conventions of moviemaking are a learned language symbolizing head turns, environmental gestalt and focused attention, respectively. Let me underscore that cameras and microphones are a tunnel-visioned view of the world. Our brains build up a gestalt of the world around us as delivered by the senses. We're aware of so much more than what is directly in front of our eyes at a given moment, but moviemaking is tunnel vision. Much of what we think we see in a movie is actually filled in by the brain, and VFX artists know how to take advantage of that. For example, a doll-sized scale figure looks fake on camera if it is too detailed because a real figure that small in perspective is far enough away to lose some detail. Now re-consider that text-only chat—is the other party a person or a machine? One is tunnel-visioned by the text-only constraint. People in forums like this use emoticons and abbreviations, like LOL, to fill in all the nuances and gestures and tones of voice that are lost in this medium. The text-only Turing Test becomes an Uncanny Valley when the simulation is on-screen or an in-person animatronic. S1M0NE also addresses another subject of interest for me. Many people claim to categorically "hate CGI," but I can guarantee that many computer generated, or digitally composited shots have slipped under their radar completely unnoticed. Artists can botch a piece whether the tools are olde fashioned analog techniques, or cutting edge digital. The skill of the artist is what matters. How good is the simulation? Is it live, or is it Memorex?
Turing created it as a thought experiment in 1950, in an age where computers were brand new and no one was even close to creating an AI. It's not meant to be used as an actual test for artificial intelligence.
When alone, I like to try non-drug hacks on myself. I try to induce Phosphenes https://en.wikipedia.org/wiki/Phosphene Misc: http://www.thenakedscientists.com/HTML/questions/question/2770/ https://en.wikipedia.org/wiki/Closed-eye_hallucination http://scienceline.org/2014/12/why-do-we-see-colors-with-our-eyes-closed/ https://en.wikipedia.org/wiki/Entoptic_phenomenon There are times, when on a Phosphene "trip" where the images almost resolve into text on a page--but I cannot quite make out the words.... Probably lines from The Emperor's New Mind. I would say that part of passing any Turing/Voight-Kampff test--that might take quite some time if done right--is a self described felt need to connect--to become more than what was given by organics or electronics. To explore and become more than you are. The above is a simulation by IBM _ _| | | | (_) (_) | | | | (_) | | |_ _| ___ \ \
Sure he would. It wouldn't matter much, though; contrary to popular belief, the Turing Test doesn't actually establish whether or not an AI is conscious/thinking or not. The Turing Test is a philosophical thought experiment that ASKS THE QUESTION whether or not an AI system that passes it could be considered conscious or not. The fact that Data passes that test merely makes him fair game for that kind of debate and turns the question into a practical rather than hypothetical issue ("The Measure of a Man"). The Voight-Kampff test is better because unlike the Turing Test it doesn't seek to establish whether or not an intelligence can pass for human, but measures the degree to which that intelligence is INhuman. In the novelization, it was explicitly given as a test not just for empathy, but the extent to which a Replicant's emotional responses were triggered reactively rather than proactively; response time is a factor, because the moral ambiguity of the question forces Replicants to waste precious processing power trying to figure out how they're supposed to feel and then displaying the proper outward signs as expected. Human responses are less rational; for example, when Deckard asks Rachel about her husband displaying pornographic images, her INITIAL reaction would be to blush or show other signs of being embarrassed/uncomfortable before she further processes the question and answers based on moral reasoning. The Replicant, on the other hand, doesn't react until it hears the entire question and decides how to process it; it only reacts to the question as a whole. So even if its final answer is the same, ("I would not allow my husband to display nudie pictures on my wall... are you implying I'm a lesbian or what?") the reception of the question is processed differently between a computer and machine brain and those very subtle differences can be detected by the Voight-Kampff test.
Voight-Kampff would be completely useless unless they could read your responses in real time, particularly autonomic responses you normally wouldn't be aware of and that a replicant wouldn't be able to mimic in a way that is 100% appropriate to the context of the discussion. Dumb test is dumb.
How can a camera see the world differently than a human? A camera doesn't have the ability to interpret what it is seeing unlike the human who can see the forest before it as being logs for a home, tools to hunt and plant crops with as well defending itself along with tools to eat with.
How can an eye ball see the world differently than an android? An eyeball doesn't have the ability to interpret what it is seeing unlike an android that can see the forest before it as being logs for a home, tools to hunt and plant crops with as well defending itself along with tools to eat with.