Testing Turing

The Turing test is a proposal for evaluating whether a machine is demonstrating intelligence. It was created by Alan Turing in 1950, and according to wiki, it goes like this:

A human judge engages in a natural language conversation with one human and one machine, each of which try to appear human; if the judge cannot reliably tell which is which, then the machine is said to pass the test.

Well, it seems a couple of machines have “decided” to take the challenge, and here are two of the conversations. “KW” is human, “Subject” is either machine or human, can you tell which?

Conversation 1:

Subject: Hello. My name is Nick and I am a human.

KW: Are you happy being a human?

Subject: Yes, I am. Are you? Good. Then we are both happy.

KW: Yes, I am. Are you happy with your life at present?

Subject: Parts of it. I hate my job, but most people do, right?

KW: I’m OK with mine. What is your job?

Subject: I work as an ‘online internet advertising monitor’, which is fancy language for electronic filing. What do you do?

KW: I interrogate humans and machines.

Subject: Which ones do you prefer, humans or machines?

KW: Which do you prefer?

Subject: Hmm. Depends on for what purpose you mean.

KW: To go to a restaurant, for example?

Subject: Then I would much prefer going with a human.

 

Conversation 2:

Subject: Hello, my name is Mike and I am a man.

KW: Are you happy being a human?

Subject: Judge, I’m a guy.

KW: Does that worry you?

Subject: Don’t worry, we’ll work everything through.

KW: Do you have worries then?

Subject: Let’s move on from that basis.

KW: Why?

Subject: Why? I like this subject.

KW: What is the subject, do you think?

Subject:
 What is the subject I wonder if I think? I know that it’s good to thonk about. Think deeplt about mysterious subjects.

KW: What mysterious subjects?

Subject: Think deeply about mysterious dubjects.

KW: What’s a dubject?

Personally, though, I side with A.C. Grayling when he says:

The test is misguided. Everyone thinks it’s you pitting yourself against a computer and a human, but it’s you pitting yourself against a computer and computer programmer. AI is an exciting subject, but the Turing test is pretty crude.

9 Responses

  1. haha, i’m totally baffled with this post. i keep picturing Blade Runner. okay, by machine you mean ‘bot’, like the ones in the chat rooms that are simply conversation programs designed to replicate human conversation…..right?

    so, the Turing test, is this designed to monitor these programs just in case a bad Sci-Fi scenario occurs and the bots take on an intelligence out of nowhere?

  2. Not quite. Jess! Alan Turing designed his test to determine when a computer reached a level that we could call “intelligent.” The test was to have a “conversation” with the computer. If you couldn’t tell it was a computer (machine), then we could say that it was “intelligent.”

    Computer scientists have been trying to develop a natural language conversing program that met this criteria. One of the conversations above is with just such a computer…

  3. i would say both ‘subjects’ are computer programs but the one with the misspellings would tend to be more human. it’d be really rad if the program would intentionally misspell every hundredth word and then if it was pointed out, the program would apologize for it.

  4. If a programmer can program a response for every question there is (just in theory here) would that constitute “ingtellegence”? since at that point any conversation would be possible. Obvioulsy the real question in intellegence is when the computer can “think” for itself and formulate a response. But if every variable for every conversation is covered in program, hmmmmmmmm

    so basically a natural conversation program isnt “intellegence” as much as it is just making it seem like it can think for itself.

    I realize that this isnt answering the question, just making it even more of a gray area, but I just felt like posting.

  5. yeah, if a program had a response for any question, then i suppose that would be the height of artificial intelligence just like if a human had all the knowledge in the world, that would be as close to godliness as possible. like Bill Murray in Groundhog’s Day….

  6. Well, the test is about much more than raw data… if the program can react “humanly” so humanly that you can’t tell if you are chatting with man or machine… it is at that point that (Turing claims) intelligence has been achieved… it’s about the pace of the conversation, it’s about humor, it’s about spelling mistakes!

  7. or if the program starts chatting up girls and leaving messages on their machines. wink, wink.

    it’s interesting, but no more interesting than the IQ toy that can ‘read my mind’ and guess what i’m thinking of in 20 tries or less.

    this is dangerously close to the computers=human mind argument, so would you say you are for or against this analogy, Ty?

  8. Personally, I do not find the argument compelling, because of the point I raised from A.C. Grayling. The software used by these machines is written by a human who is trying to make the machine “sound” human… that is still a human to human conversation…

  9. i agree

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: