Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!
Photo

Turing Test Poll

AI Artifficial Intelligence Turing Test 20 years 2035 2030 2029 computer machine poll

  • Please log in to reply
8 replies to this topic

Poll: Turing Test Poll (23 member(s) have cast votes)

Will machines pass my version of Turing Test by 2035?

  1. Voted yes (19 votes [82.61%])

    Percentage of vote: 82.61%

  2. no (4 votes [17.39%])

    Percentage of vote: 17.39%

Vote Guests cannot vote

#1
Kynareth

Kynareth

    Member

  • Members
  • PipPipPipPip
  • 148 posts

I know there have been similar polls in the past but I haven't seen any like this one.

 

Do you think in the next 20 years machines will be able to pass my version of Turing Test?

 

Description:

1 hour long text conversation with AI in English, assuming that the person on the other end is an adult, well educated native English speaker with no signs of mental ilness, instability, Alzheimer etc.


  • Jakob likes this

#2
wjfox

wjfox

    Administrator

  • Administrators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 10,150 posts
  • LocationLondon

If trends continue, machine intelligence will be six orders of magnitude more powerful by 2035 compared with today. Given this exponential increase and the accompanying rise in deep learning networks, semantic understanding, etc. I'd say the test is very likely to be passed, and probably sooner.


  • Casey likes this

#3
Jakob

Jakob

    Groveling Goblin

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 6,080 posts

Finally an insightful thread that isn't just political crap.

 

I believe that a computer could easily do that in 20 years. Your test is much better than the current Turing Test; if garbage like Eugene Goostman can pass the Turing Test, the test is surely broken.



#4
Unity

Unity

    Information Organism

  • Members
  • PipPipPipPipPipPipPipPip
  • 2,477 posts

Turing test is irrelevant.



#5
Zaphod

Zaphod

    Esteemed Member

  • Members
  • PipPipPipPipPipPip
  • 768 posts
  • LocationUK

I think this form of Turing test could be passed by relatively crude AI that lack any form of human-level consciousness. Even smart humans can be deceived by clever programming that has been tailored towards answering questions that are likely to be posed in a Turing test, especially if the input/output is text-based. After watching Ex Machina, I think the form of Turing test they use is more informative. If you are shown that the intelligence is artificial, but cannot help but see the AI as human then the AI must be sufficiently good in order to overcome our initial preconceptions. However, this still doesn't necessarily distinguish between consciousness and simulated consciousness.

 

In reality, I think we will  need multiple tests to uncover whether an AI is conscious, one of which may be a form of the Turing test. 


  • Outlook likes this

#6
Zaphod

Zaphod

    Esteemed Member

  • Members
  • PipPipPipPipPipPip
  • 768 posts
  • LocationUK

A further thought... the Turing test has already been passed by very crude programming and this is because of naive interrogation and the anthropomorphic fallacy. The interrogators of the Turing test should be the programmers of the AI themselves. If they cannot distinguish between their own creation and a human response despite knowing exactly how the AI is programmed then the AI must be very sophisticated.



#7
sorcerer

sorcerer

    Member

  • Members
  • PipPipPipPipPip
  • 428 posts
  • LocationFrance

A question though: does this assume that the machine is programmed by humans or learns the language itself? Similar to my thread about the chess computer.



#8
nomad

nomad

    The Darkness

  • Validating
  • PipPipPipPipPipPip
  • 548 posts

I think the hardware will be there but I have my doubts as to whether or not the software will have "caught up" yet.


Cats.


#9
GenX

GenX

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,045 posts
  • LocationSo Cal

I mean, I had a problem with Verizon and I went online and chatted with what I thought was a person, and it was only when they accidentally said in the middle of the conversation that they were going to transfer me to someone else and then just kept on going that I realized it was a computer program.  So I'm sure that in the next twenty years there will be many computer programs that easily pass the Turing test.  But that is only step 1.  What happens after that?  How do we know for sure whether a computer program really has consciousness or is just simulating consciousness?  I'm very interested to see how that plays out.


The only thing we ever want is more






Also tagged with one or more of these keywords: AI, Artifficial Intelligence, Turing Test, 20 years, 2035, 2030, 2029, computer, machine, poll

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users