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.Nothing in the story as to whether the program claimed to be a Nigerian Minister of the Treasury.
It is thought to be the first computer to pass the iconic test. Though other programmes have claimed successes, those included set topics or questions in advance.
A version of the computer programme, which was created in 2001, is hosted online for anyone talk to. (“I feel about beating the turing test in quite convenient way. Nothing original,” said Goostman, when asked how he felt after his success.)
The computer programme claims to be a 13-year-old boy from Odessa in Ukraine.
Tuesday, June 10, 2014
Computer Scientists discover that the Turing Test is pretty useless
A computer can pass it if you lower your standards enough:
Subscribe to:
Post Comments (Atom)
9 comments:
Heck BP, there are a lot of voters in the US who couldn't pass the Turing test.
(Nice line about the Nigerians. I'm still waiting for my check...)
ROTF +1 on Sarge!!!
Yeah, Sarge beat me to it. It's not just voters though. You ever read Facebook?
Hm. I thought it was better than the average Tumblr user or YouTube commenter.
TBG
IIRR another chatbot did the same thing just a few years ago, with a 50+% result.
33% is success/passing? Talk about low standards.
Dave, I just read through an amusing list of things people have posted on social media that had me shaking my head in disappointment.
Preview: "Saw the dumbest elevator today, it had a button for the floor I was already on...?"
And, "Wow I can't believe the titanic sunk and the people filming it didn't even stop to help :("
At first I tried to convince myself the posts were staged. Then I brought myself back to reality and cried a little.
This is such bull. The test set out by Turing states that you connect the tester's terminal to two others. One is a human, while the other is the AI being tested. If the human can't do any better than chance at picking the human from the AI, then it passes.
Making it a little kid who speaks broken English, and lowering the bar to 30%, totally invalidates the test.
Yup. Bar has been lowered enough that we give the computer a 150% advantage, and claim it beat the turing test.
The turing test had a little higher standards than this.
For example, a human control was included in the turing test.
No consideration was given to age or ethnicity in Turing's test, but I think that you can at least agree that he intended the AI to simulate an adult speaking in his native language.
Finally, the Turing test was based on results of 50% plus. ie, the comupter needed to have a better than chance result to pass.
This thing did none of those things. Turing test NOT passed.
Was this Turing test scored by ex pro-baseball players? .300 will get you into the record books in the big leagues. .400 will have you talked about for 60 years after you retire.
Unlike the rest of real life.
Post a Comment