We are at just the start of AI for visual interactions.  This test is a first stumble showing just how complicated the issue is.  Its analogue.  Not mathematical.  Its interpretive not computational.

OK, first for visuals, perhaps. But, we haven't gotten a computer that can pass the first test dreamed up by Allen Turing. And that test was a whole lot better and more meaningful than this silly thing that at least one human failed at and no human on this thread got 100% at yet.

So, this test is a heaping steaming mound of dung. It is an insult to Turing to put his name on it.

Certainly there is much to do in AI. Let's start with getting a computer to pass the original Turing Test. Then, let's invent another great test for the computer to pass. This is most certainly not it. That's my point. This test isn't even a progression along the way. It's a complete and utter failure.

When a good test is invented, I'm just guessing that the author will put his/her own name on it.

That name might even be the name of the computer who thinks of it. Perhaps it'll be the Mycroft or Minerva test.

BTW, passing the real Turing Test alone would be a strong argument to me that the passing computer should be granted the rights of other sentiences. Ditto for any dolphin, ape, or parrot who can pass the test.

Oh SCOTT!!!!!!!!!!

I was going to ignore you until that last sentence.  Where is your appreciation for the subtlety of the self conscious and aware "entity?"  Do you think it will ever be expressed in a series of machine code and one's and zero's? 

The Turing Test isn't even about sentient being or intelligence.  Its about reaching a first rung in that process of getting a computer that can "fool" a human being in an exchange of text.  NOT EVEN THE SAME QUESTION YOU ARE READY TO GIVE VOTING RIGHTS TO.  Ha, ha.  No wonder you don't understand the visual Turing  test issues.

Still an interesting debate going on whether or not awareness is possible with machines.  No one I think is "that" sure one way or the other.  Course, some think their stuffed Teddy Bear is sentient as it is always available to listen.  No doubt at all when computers can fool the human being---many humans will be fooled.

While typing this reply, it struck me what the Bobbo Test for Sentient Beings might be but I think I am ripping off a forgotten episode of Twilight Zone:  its when the computer says it wants to die.  Ha, ha.

Good one on the Quantus Thread.  I must be getting crusty?

Scott, I am surprised you omitted Octopi and Squid from the list!
Bobbo, how can you say that self awareness is critical step? Surely some of the mammals have some degree of self awareness, and how do you define self awareness anyway? Can't you visualize a very large/powerful computer with a specially designed operating system using a self review program of its memory of interactions being at some lower level of "self awareness"?
I don't mean to belittle this feature of the human mind, but don't we design our computers to mimic ourselves?
Finally, should the test for what is human versus what is AI not be a test designed to find the Human Error(s) in Thinking? [I'm referencing Yochelson and Bernstein tome on the Criminal Personality. They listed some 56 or 58 pathological errors in thinking of criminal minds.]

Obtuser---whats your point?  You are only repeating what the issue is:  whats the definition of sentient, with the subset of can computers reach it?

MY POINT, in correcting Scott, is that a computer passing the Turing Test, is not in theory anywhere near that feat.  It is only a first step--to fool a human into thinking he is talking to a human.  That evidences the ARTIFICIAL in AI.  When the computer on its own asks:  "Won't anybody talk to me, I'm lonely" then maybe the conversation can start up again======to find out who put that code into the system.

Obtuser---whats your point?  You are only repeating what the issue is:  whats the definition of sentient, with the subset of can computers reach it?

MY POINT, in correcting Scott, is that a computer passing the Turing Test, is not in theory anywhere near that feat.  It is only a first step--to fool a human into thinking he is talking to a human.  That evidences the ARTIFICIAL in AI.  When the computer on its own asks:  "Won't anybody talk to me, I'm lonely" then maybe the conversation can start up again======to find out who put that code into the system.

Bobbo, if you extend the logic one step further.........sentient computers will write their own code, not have it written for them.
Do you agree that higher animals are to a degree sentient, in that they remember what they have experienced, and exhibit a degree of personality versus others of their kind? And therefore for a computer to qualify for the title Intelligent it must exhibit personality traits as well as retained experienced situational memories! Furthermore, a percentage of those computers would have to present "rogue" personalities, too. Conversely, an unintelligent computer would ask the question Will anybody have a conversation with me? Then proceed to talk to itself as if it were two separate entities, n' est pas?

I was going to ignore you until that last sentence.  Where is your appreciation for the subtlety of the self conscious and aware "entity?"  Do you think it will ever be expressed in a series of machine code and one's and zero's? 

I don't know. So far, certainly not.

Why do you believe electrical impulses in neurons fundamentally different than those in wires and semi-conductors? Because neurons are slower?

Scott, I am surprised you omitted Octopi and Squid from the list!

So am I. And, many more, I would think.

\MY POINT, in correcting Scott, is that a computer passing the Turing Test, is not in theory anywhere near that feat.  It is only a first step--to fool a human into thinking he is talking to a human.  That evidences the ARTIFICIAL in AI.  When the computer on its own asks:  "Won't anybody talk to me, I'm lonely" then maybe the conversation can start up again======to find out who put that code into the system.

Are you so sure? Perhaps when a computer does not make statements such as "I feel lonely" or "I'm bored" or "I want to die" or "I have this pain down all the diodes in my left side"* it will not fool humans into thinking that it is human. If the definition of passing the Turing Test is that you and I can't tell the difference between a human and the computer, which is actually the definition, then perhaps it is the only test necessary.

Regardless, it is a much better test than this pile of garbage you posted.

Scott--please provide a more detailed explanation how the Turning Test is a better test for visual performance than the test we just took.

I'll wait.

Scott--please provide a more detailed explanation how the Turning Test is a better test for visual performance than the test we just took.

I'll wait.

I didn't say it's better at testing visual performance. If you think I did, please quote exactly what I said that makes you think so.

What I said was that this test was totally fucking worthless.

Why is a visual test so important? Do you deny that people blind from birth are intelligent and sentient beings? What I said was that the original test of fooling people into thinking that they were typing to a human, as we presumably are now, is a test that probably obviates the need for this stupid test. Further, I believe that the visual test adds nothing. Even typing back and forth to each other, we can still discuss things like "the person behind the sign" and other such spatial constructs.

So, to pass the Turing Test, a computer must still get these things right in such a conversation. If I said to the computer that there was a bus leaving in five minutes and it should be under it, the computer would still have to generate an appropriate response, rather than thanking me for the instruction/advice. But, the visual is not the important bit. A blind person would still get this right.

So, A) why do you place such importance on the visual? and B) why do you even think this test is a good test of visual/mental acuity?

I deny both.

Language always needs to be interpreted.  Its why Turning made his test?  Ha, ha.

"I didn't say it's better at testing visual performance.  If you think I did, please quote exactly what I said that makes you think so." ////  why here it is, the very last thing you said.  What up Scotty, watching too much Sponge Bob?:  "Regardless, it is a much better test than this pile of garbage you posted."  What did you mean by that statement then?

Really Scott, you are being silly.  Its just the start of a long journey and you want to poo poo the efforts by comparing it to a different test that has been worked on for 50 years?  How much "sense" does that make?  Beginnings are always rough.

Silly.

Language always needs to be interpreted.  Its why Turning made his test?  Ha, ha.

No. This is why I repeatedly point out that you use your own definitions that have nothing to do with dictionaries.

"I didn't say it's better at testing visual performance.  If you think I did, please quote exactly what I said that makes you think so." ////  why here it is, the very last thing you said.  What up Scotty, watching too much Sponge Bob?:  "Regardless, it is a much better test than this pile of garbage you posted."  What did you mean by that statement then?

I said it was a better test. I stand by that statement. It is not a better test of visual acuity. It is a better test of intelligence. The subject is artificial intelligence, not artificial vision, unless I'm mistaken. So, read a dictionary. Read my posts again. Tell me where I said it was a better test of visual anything.

Really Scott, you are being silly.  Its just the start of a long journey and you want to poo poo the efforts by comparing it to a different test that has been worked on for 50 years?  How much "sense" does that make?  Beginnings are always rough.

Silly.

No. I'm pointing out that the point is artificial intelligence. Not one computer yet has passed this test. So, before we go about creating a test that would deny that blind people are intelligent, let's work on passing a real test.

So, again, please don't ignore the important questions I asked.

Do you deny that people blind from birth are intelligent and sentient beings?

When you answer the above question, I think you'll see that this test is an insult to the intelligence of human beings rather than a test of the intelligence of computers.

There is always a bugger factor in any test regime.Just how would Dyslexics score on this test and on the Turig test? I have known some Dyslexics casually, and suspect that on certain questions they would be total failures.
Another side point, I suspect that if and when a computer is designed, built and tested that qualifies for some level of AI, it will lack much of the diversity found in human intelligence.

There is always a bugger factor in any test regime.Just how would Dyslexics score on this test and on the Turig test? I have known some Dyslexics casually, and suspect that on certain questions they would be total failures.
Another side point, I suspect that if and when a computer is designed, built and tested that qualifies for some level of AI, it will lack much of the diversity found in human intelligence.

Actually, I've had a rather lengthy conversation with a dyslexic on my blog site, all typing, no audio. Though, a couple of times I misunderstood the meaning, at no point did I ever question whether the person was indeed a human being. So, dyslexics have no trouble with the Touring Test in my experience. Remember what the test is, not grammar and spelling, but whether the individual is a person. Did you ever suspect a dyslexic of being a computer??!!? I doubt it.

A computer that fails in the diversity, would likely also fail the test. You'd notice it repeat itself and suspect it of being a computer. You're probably doing that with me right now. Sorry, just being an (_o_) in hopes that it keeps me passing the Turing Test.

Scott--blind people are completely irrelevant to the Turing Test for Visual Competency.

You also are over equating language skills with intelligence.

does raise the whole other issue of what intelligence is and can it be unrelated to self awareness, to self directed inquiries and at what level would a second person not know artificial from real, and really==what difference would that make?

A lot of it comes down to what either of us thinks AI actually, as in "ACTUALLY" is.  You seem to equate it much more closely with sentient self awareness whereas I see it as not even close and many stumbling blocks away from that goal.

Blind indeed.