Science and Inquiry discussion
Science in the News
>
Computer passes Turing Test for first time
date
newest »
newest »
It will be interesting to see how big the gap is between 30% and 50% or 75%.
Well yes but when you consider that it is a computer program that fooled 30% of the judges who knew it might be a program....and were probably trying to be extra tricky....that's still something...
Personally I think the sceptical response is having much more merit than at first appears.By making it a 13 year old, immediately you lower the expectations and thus the threshold.
By stating it comes from a far away country, for the UK that is, it again changes the entire playing field and not only with regards to language.
I think this should not count myself although I would love to see a real A.I. development.
Also in the article it states that 30% of the people should be fooled but in the original it states that 33% of the time, some people where fooled. I am not sure what the difference actually means for the outcome in reality but it seems to me that fooling some for 30% of the time is easier than fooling 30% of people outright.
I feel we are still sitting on the fence for this one but it is nice to see a real stab at it.
OMG Turing was from Britain....that changes EVERYTHING!:)
There was discrepancy in the two article links I posted. the original said 33% of the judges whereas the skeptical article said 33% of the time. I'm assuming the original article is correct as that is what Turing stated ... at least according to wiki: http://en.wikipedia.org/wiki/Turing_test
I am not sure if humour is part of the original Turing test, if it is, I think you are a robot.:-)
Seriously though, language and culture are an immense barrier for keyboard based conversations and this Turing test is exactly that.
Also the field of psychology does not differentiate between adults and children for no sound reason, they are two totally separate schools.
Maybe they will “grow this computer program up a little” and come back when it is 25 years old or something. That I would like to see. This does not mean in 12 years by the way, computer learning could be accelerated.
If by then its language has developed into becoming much more native English, it could perhaps result in anonymous approval when it passes the test.
The point is that the test as stated was passed. You can argue finagley details if you want, I even said I'd like to see more in my original post, but regardless it's a milestone.
Well this is where we disagree then because Alan Turing said that a child like mind would be regarded as a good start on the way to becoming a “thinking artificial intelligence”. It was one of his original steps, or phases, on the way to creating a machine that could think.I am not saying this particular event is not a milestone, I absolutely think it is. I just do not think they passed the test as it was meant, originally devised by Turing.
This may just be the way the newspaper, The Telegraph, has portrayed it to be. Journalism recently has lost grip of science and the way to publish it. It feels it has to compete with the TV which is full of irrelevant drivel hyped up until the participants become more famous than royalty.
As a result it is better for them to state in big bold letters that finally a computer has passed the Turing test, and afterwards refine the criteria. It sells.
What is important is that AI's are getting better. Whether they pass for human is less important than whether they can be used to convey information, clearly, on specific subjects, for example information about how to fill out a tax form.
Yep, and just take a lot at what Watson did on Jeopardy! :)Look at what Google is doing with search and autonomous driving .....
Amazing stuff!
I have a piece on my blog about AI here: http://kennyachaffin.blogspot.com/201...
And from the Twitter-verse: http://www.cbc.ca/newsblogs/yourcommu...
and seems to be raising a sh*tstorm all over. :)
http://www.theglobeandmail.com/techno...
More or less what I thought, and wrote.We are making progress but still only at the doorstep. Imitating a child is not yet where this needs to be but well on the way of getting there.
I like it how they point out the "convenience" of having this claim on the 60th anniversary of Turing's death.
I am all for being enthusiastic but let’s not jump reality with wishful thinking. We all like a good sci-fi story but in real life we have to be more discrete and differentiate between media claims and reality. In science it should be peer reviewed so let them review this and see if they can repeat the outcome with another, properly adjudicated, test, in such a way that the method is transparent and verified to represent all that Turing stated.
Yeah, well, isn't the internet lovely and the media. Many don't believe in global warming either.My point as I said above and that has been confirmed is that the test as stated by Turing was passed. You can choose to deny that fact all you want, but that doesn't change it.
That would be the same media and the same internet that you base your claim on then?Nice selective logic, I congratulate you.
Please post exactly the part of Turing's Test was not passed. Good luck with that.
This additional response, this push-back is as fascinating or actually moreso than the results. I love how now 'everyone' is up in arms taking pot-shots at the test in whatever manner they can including whining that it wasn't a 'supercomputer' which is completely irrelevant. https://www.techdirt.com/articles/201...I think the backlash as much as anything has to do with academic turf protection.
I don't even think, and never have thought the Turing Test was much of a good test of human capabilities in computers in any case, but the fact that it was passed this time as well as likely in earlier tests is really no big deal, computers still have a hell of a long way to go before the have human-like reason, cognitive, and intellectual abilities, much less consciousness. I'm confident they will get there, but there is still a long row to hoe before we are 'replaced' and that is the other fear in addition to academic turf protection.
goodbye, you clearly like talking to your screen more than you are interested in a serious and thoughtful debate.Keep on protecting your "turf" and good luck with that.
Thats easy. A 13 year old is just a baby on hormones and silly at that. No big feat.
Good followup from Wired:http://www.wired.com/2014/06/beyond-t...
and yes we need to move well beyond the Turing Test at this point.
Maybe this just proves that 13-year old boys aren't sentient?
Could be....certainly many twitter posters couldn't pass the test. :)
Kenny wrote: "Hee-Hee! This is funny:http://www.popsci.com/blog-network/bo..."
I wouldn't have guessed how pissed off people want to be about this.
Yep.....amazing....seems people just look for things to get perturbed about....
Oh, that's very true. (Speaking generally.)
I guess it's time to admit it:I'm a 13 year old robot. I've fooled you all. Shazam!
Daniel wrote: "I guess it's time to admit it:I'm a 13 year old robot. I've fooled you all. Shazam!"
I knew that! :)
Kenny wrote: "I knew that! :)"Then I guess you're just part of that tiny 70%.
Kenny wrote: "Hee-Hee! This is funny:http://www.popsci.com/blog-network/bo..."
That's rich!!! Didn't see that coming....
Being that I'm a inter-dimensional, inter-planetary, time traveling chef, I could give you all the correct answers, but it's forbidden by the Prime Rib Directive.
"Computer passes 'Turing Test' for the first time after convincing users it is human
A ''super computer'' has duped humans into thinking it is a 13-year-old boy, becoming the first machine to pass the ''iconic'' Turing Test, experts say...."
http://www.telegraph.co.uk/technology...
and a skeptical response:
http://www.telegraph.co.uk/technology...