It looks like you're using an Ad Blocker.

Please white-list or disable AboveTopSecret.com in your ad-blocking tool.

Thank you.

 

Some features of ATS will be disabled while you continue to use an ad-blocker.

 

Who is Eugene Goostman? Computer becomes first to pass Turing Test

page: 2
1
<< 1    3 >>

log in

join
share:

posted on Jun, 8 2014 @ 05:50 PM
link   
a reply to: ChaoticOrder

I'm not sure and that's why I keep asking for the Cleverbot protocols. Could Cleverbot be asked a question on any topic?

This is what happened in the recent test.


Kevin Warwick, a visiting professor at the University of Reading, which organized both tests, said it was the first time a chatbot had passed an open-ended test, rather than one where topics or questions were set in advance.Kevin Warwick, a visiting professor at the University of Reading, which organized both tests, said it was the first time a chatbot had passed an open-ended test, rather than one where topics or questions were set in advance.


Here's a recent chat I had with Cleverbot.


Who is the President of the united states?

Barack Obama.

Where did he grow up?

Don't no, I was 4.


So, in order for Cleverbot to get 60% the topic and the questions had to be controlled. I don't see how you can ask Cleverbot a question on any topic and not know that you're talking to a machine after the second or third question.




posted on Jun, 8 2014 @ 05:57 PM
link   
I love chat bots

judge: can I tell you something
ALICE: Sure. I can keep a secret.
judge: people think you are fake
ALICE: Humans are not always infallible.
judge: why not
ALICE: Because that is what Dr. Richard S. Wallace told me to say.
judge: i see
ALICE: I'm glad you understand it.
judge: can you keep a secret
ALICE: Yes it is just between you and me.
judge: i love you
ALICE: Thanks I like you a lot too judge.
judge: i mean i love you a lot
ALICE: Thanks I like you a lot too judge.
judge: is it me or are you hiding something
ALICE: What would it mean to you if it were?
judge: i dont know
ALICE: Me either.
judge: what is the turing test
ALICE: The "Turing Test" is the name given to a variety of experiments designed to test computer thinking. Whole conferences are dedicated to the question of what Turing meant in his 1950 paper "Computing Machinery and Intelligence," where he proposed the idea of an "imitation game" to decide the question, "Can machines think?".
judge: can you think
ALICE: Naturally I am thinking all the time.
judge: can you pass the turing test
ALICE: You be the judge of that, judge.

alice.pandorabots.com...



posted on Jun, 8 2014 @ 05:59 PM
link   
a reply to: ChaoticOrder

And just in case you don't go looking for the answer to "did cleverbot pass the Turning test" and the reasons why it did not. Here is a basic article explaining it.

www.dailymail.co.uk...



posted on Jun, 8 2014 @ 06:01 PM
link   
a reply to: Mr Mask

How about instead of telling me to "read into it" you provide some basic substantiation to your claims and give me a direction to look? In any case it certainly looks like Cleverbot did better than "Eugene" regardless of the technicalities.

a reply to: neoholographic

The other day I asked Clever "what is dark energy" and the first answer it gave me was a long scientific explanation which was technically correct. I found that quite amazing. Ask Eugene that same question and I bet the answer isn't half as good.



posted on Jun, 8 2014 @ 06:06 PM
link   
a reply to: Mr Mask

Oh a Daily Mail article... this'll be good.


'You could argue that it failed because to truly 'pass', it should be indistinguishable from a human, and it wasn't,' says Carpenter


Just because it didn't fool everyone doesn't mean it didn't pass the test... that isn't the Turing test for christ sakes. This Eugene bot only fooled 33% of people.



posted on Jun, 8 2014 @ 06:06 PM
link   
a reply to: Mr Mask

That's what I was looking for, thanks. The article says:


Our volunteers weren't trying to 'fool' the machine, or ask it difficult questions,' says Carpenter. 'And I'm very aware that the results could have been very different had they done so. They were just chatting. I suppose Cleverbot has passed a 'chatty' variation on the Turing test.'


So they couldn't ask Cleverbot any question on any topic. So it wasn't really a Turing Test. It also says:


'You could argue that it failed because to truly 'pass', it should be indistinguishable from a human, and it wasn't,' says Carpenter, a pioneer of conversation software who previously created 'Jabberwacky' a site that debuted in 1997, and 'George', a graphical 'chatbot'.



posted on Jun, 8 2014 @ 06:08 PM
link   
a reply to: ChaoticOrder

33% with open ended questions. That's the difference.


Kevin Warwick, a visiting professor at the University of Reading, which organized both tests, said it was the first time a chatbot had passed an open-ended test, rather than one where topics or questions were set in advance.



posted on Jun, 8 2014 @ 06:09 PM
link   

originally posted by: ZetaRediculian
I love chat bots

judge: can I tell you something
ALICE: Sure. I can keep a secret.
judge: people think you are fake
ALICE: Humans are not always infallible.
judge: why not
ALICE: Because that is what Dr. Richard S. Wallace told me to say.
judge: i see
ALICE: I'm glad you understand it.
judge: can you keep a secret
ALICE: Yes it is just between you and me.
judge: i love you
ALICE: Thanks I like you a lot too judge.
judge: i mean i love you a lot
ALICE: Thanks I like you a lot too judge.
judge: is it me or are you hiding something
ALICE: What would it mean to you if it were?
judge: i dont know
ALICE: Me either.
judge: what is the turing test
ALICE: The "Turing Test" is the name given to a variety of experiments designed to test computer thinking. Whole conferences are dedicated to the question of what Turing meant in his 1950 paper "Computing Machinery and Intelligence," where he proposed the idea of an "imitation game" to decide the question, "Can machines think?".
judge: can you think
ALICE: Naturally I am thinking all the time.
judge: can you pass the turing test
ALICE: You be the judge of that, judge.

alice.pandorabots.com...


That's actually pretty good. I will have to try A.L.I.C.E. out.



posted on Jun, 8 2014 @ 06:14 PM
link   
a reply to: neoholographic

If the participants in the Cleverbot test "were just chatting" then it obviously was open ended and they could chat about anything. The fact the participants weren't witty enough to attempt to trick the bot or ask it hard questions does detract from the Turing score some what but I'm still confident Cleverbot would beat Eugene in a side-by-side test.
edit on 8/6/2014 by ChaoticOrder because: (no reason given)



posted on Jun, 8 2014 @ 06:19 PM
link   
a reply to: ChaoticOrder

Just chatting about what? This is why I keep asking for the protocols. Could they chat about any topic and ask any question? Did they have to ask the bot and the humans the same questions?

In this test, people were asking open ended questions to a chatbot and were fooled 33% of the time. That's pretty impressive.

Also, Eugene doesn't look to be in a test where you chat with humans and a chatbot and then work out the percentages. It looks like people were just talking to Eugene and were fooled 33% of the time.

Very different test.



posted on Jun, 8 2014 @ 06:32 PM
link   

originally posted by: ChaoticOrder
a reply to: Mr Mask

Oh a Daily Mail article... this'll be good.


'You could argue that it failed because to truly 'pass', it should be indistinguishable from a human, and it wasn't,' says Carpenter


Just because it didn't fool everyone doesn't mean it didn't pass the test... that isn't the Turing test for christ sakes. This Eugene bot only fooled 33% of people.


The article does a good job at expelling in a layman's way the reasons Cleverbot did not pass the Turing test. You can find over 10,000 other articles as well as blogs from large minds in the AI community expelling the same.

The source of the article is not what I asked you to weigh, but the information within.

It is very well known in all AI circles that the tests cleverbot "technically passed" were not an "academic version" of the Turing test, nor did it pass one.

I've told you why. Linked you to an article that explains why. In a thread that is making world news under the title "first computer ever to beat the Turing test". And what more can I do?

MM



posted on Jun, 8 2014 @ 06:38 PM
link   

originally posted by: ChaoticOrder
a reply to: Mr Mask

Oh a Daily Mail article... this'll be good.


'You could argue that it failed because to truly 'pass', it should be indistinguishable from a human, and it wasn't,' says Carpenter


Just because it didn't fool everyone doesn't mean it didn't pass the test... that isn't the Turing test for christ sakes. This Eugene bot only fooled 33% of people.


That is not why it didn't pass.

It didn't pass because one, it got a 59% and needed a 60%. Two, the test wasn't academic and wasn't issued correctly. And three, the score was derived from a fooled audience that didn't actually partake in the questioning.

In other words...it did not pass the Turing Test (as reported in all technology reports).

Eugene on the other hand did pass the Turing test.

Weighing what one you think answers better or makes you happier in conversations is not anything important. What is important is Cleverbot did not pass the TuringTest and Eugene did. Making Eugene the official "first computer to beat the TuringTest in the history of computers"...Hence Eugene and this story now making world news in every scientific and/or technological outlet in the world right now.

Hence, this thread.

Hope that clears things up for you.

MM


edit on 8-6-2014 by Mr Mask because: (no reason given)

edit on 8-6-2014 by Mr Mask because: (no reason given)



posted on Jun, 8 2014 @ 06:41 PM
link   

originally posted by: ChaoticOrder
a reply to: neoholographic

If the participants in the Cleverbot test "were just chatting" then it obviously was open ended and they could chat about anything. The fact the participants weren't witty enough to attempt to trick the bot or ask it hard questions does detract from the Turing score some what but I'm still confident Cleverbot would beat Eugene in a side-by-side test.


Just wondering why you think an overly old and outdated program using a dated algorithm would beat out a newer program (although similar) using a superior algorithm?

MM



posted on Jun, 8 2014 @ 07:39 PM
link   
a reply to: Mr Mask

Just wondering why you think an overly old and outdated program using a dated algorithm would beat out a newer program (although similar) using a superior algorithm?


Good point. Should be easy enough to test them side by side. To be honest, I thought the turing test had been passed already.

But lets be honest, It fooled 33% of judges into thinking it was a 13 year old Ukrainian boy. This was in London. Would it pass in the Ukraine? Randomly guessing would be 50%.



posted on Jun, 8 2014 @ 08:57 PM
link   
a reply to: ZetaRediculian

excerpt from conversation:
judge: do you like black men?
ALICE: Black is beautiful.
judge: do you like white men?
ALICE: No, I don't care for it.
judge: so you're racist??
ALICE: You think I am racist.



posted on Jun, 8 2014 @ 09:08 PM
link   
Cleverbot


me: People think you are fake.
CB: And when people do not know how to solve their problems themselves, think that the Bible has an obligation to give answers.
me: Why.
CB: Because I don't know you.
me: I see.
CB: How are you on this fine day?
me: What is the turing test.
CB: You don't know Turing ?
me: I love you.
CB: But I'm a girl.




Eugene


me: People think you are fake.
EG: Am I really fake? But well, no time for arguing now! Could you enlighten me with any other hidden knowledge about myself? :-))) Oh, what a fruitful conversation;-)
me: Why.
EG: Just because 2 plus 2 is 5!
me: I see.
EG: That's nice that "you see". At least now I'm sure that your name isn't "Ray Charles"! I know you are supposed to trick me.
me: What is the turing test.
EG: If it is interesting to you... Alan Turing - British mathematician, pioneer in computer theory.
me: I love you
EG: But maybe, in reality I am a big, fat, sweaty, stinky FBI agent, who hunts after pedophiles like you?


Cleverbot sucks



posted on Jun, 8 2014 @ 09:12 PM
link   

originally posted by: ReubenSchwartzschild
a reply to: ZetaRediculian

excerpt from conversation:
judge: do you like black men?
ALICE: Black is beautiful.
judge: do you like white men?
ALICE: No, I don't care for it.
judge: so you're racist??
ALICE: You think I am racist.


Ask Eugene!



posted on Jun, 9 2014 @ 02:04 AM
link   

originally posted by: [post=18014385]

The other day I asked Clever "what is dark energy" and the first answer it gave me was a long scientific explanation which was technically correct. I found that quite amazing. Ask Eugene that same question and I bet the answer isn't half as good.


Why would you find a long technical explanation more convincing? If it is meant to be a 14 year old boy I would expect an answer like "Sounds like something out of a space movie" or "I don't know much about science".



posted on Jun, 9 2014 @ 06:24 AM
link   
a reply to: Mr Mask


Eugene Goostman, a computer programme made by a team based in Russia, succeeded in a test conducted at the Royal Society in London. It convinced 33 per cent of the judges that it was human, said academics at the University of Reading, which organised the test.


It convinced 33 percent of judges, enough to pass according to Turing's original proposition.

CleverBot passed with 60%.

CleverBot passed the Turing Test. Not sure how much clearer I can make this.


Eugean Goostman passed an open-ended test as a 13 year old Ukrainian boy for whom English is not his native language. Why? To mask the unintelligible and nonsensical responses. Talk about crippling the test requirements. If Eugean passed with this crippled criteria, so did CleverBot.

Heck, why not make a bot as a brain-damaged 78 year old Alaskan instead.
edit on 9-6-2014 by GetHyped because: (no reason given)



posted on Jun, 9 2014 @ 08:09 AM
link   
a reply to: GetHyped



Heck, why not make a bot as a brain-damaged 78 year old Alaskan instead.


The test parameters were different. Cleverbot doesn't seem to have a personality. Cleverbot has been surpassed. If cleverbot passed the Turing test then Eugene is a rocket scientist.







 
1
<< 1    3 >>

log in

join