[ExI] The Turing Test

[Christian Saucier]

Let's also remember that this is a specialized system: it only handles reservations and appointments.  For sure, that will grow to include more complex interactions in the future, but the fact that the system is specialized is important.

Humans are specialized too. I don't know much and couldn't hold a serious conversation about microbiology.  Humans are also automatons, preferring to repeat habits rather than rethinking about all the social rules at every single moment.

As a result, I don't believe any of this should be mistaken for a proper Turing test.  

That said, based on the samples I've seen, I most likely wouldn't be able to tell if I was organizing a reservation with Duplex.

C. 

On May 16, 2018 2:46:34 PM UTC, Dave Sill <sparge at gmail.com> wrote:
>On Tue, May 15, 2018 at 5:40 PM John Clark <johnkclark at gmail.com>
>wrote:
>
>>
>> On Tue, May 15, 2018 at 5:08 PM, William Flynn Wallace <
>> foozler83 at gmail.com> wrote:
>>
>> *​>​Isn't the Turing test just whether people can tell it's a person
>or a
>>> computer?*
>>>
>>
>> ​Yes.
>>
>
>Actually, it's supposed to be a person determining which of two
>entities
>it's interacting with is a computer.
>
>​> ​
>>> *It seems to this psychologist that that is a very low bar to get
>over.
>>> People believe what they want to believe and are easily fooled.*
>>>
>>
>> But in this case the people didn't even know they were in a test,
>they
>> just assumed they were talking to a human. And the machine did a
>better job
>> making a reservation to the Chinese restaurant than I could have
>done; with
>> the thick accent and elliptical phrasing I had great difficulty
>> understanding what the human was trying to say, but obviously to the
>> computer it was clear as a bell.
>>
>
>That's not a Turing Test at all.  The tester has to be actively trying
>to
>make that determination.
>
>​>* ​*
>>> *I'd like to see a more meaningful test.*
>>>
>>
>> ​
>> Like what? All Turing is saying is that we should  judge how
>intelligence
>> a computer is the same way we judge the intelligence of other people,
>by
>> behavior. And I have to say that in this case the computer
>demonstrated
>> more intelligence than I have because I would have been constantly
>saying
>> "I don't understand what you're saying please repeat it".
>> ​
>>
>
>Yeah, and superhuman abilities are one thing that I, as a Turing
>Tester,
>would use to identify the computer. :-)
>
>-Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20180516/82849a57/attachment.html>