[ExI] spike turing test, was RE: ai class at stanford
stefano.vaj at gmail.com
Fri Sep 2 12:56:17 UTC 2011
On 2 September 2011 06:24, Kelly Anderson <kellycoinguy at gmail.com> wrote:
> So here's a question.... which would be easier to program? The generic
> Turing test "I'm pretending to be SOME generic human, and you can't
> tell the difference" or the person specific test "I'm Paris Hilton,
> and you can't tell that I'm not".
I am going to publish an article in a few weeks in *Divenire* where I defend
the somewhat unusual contention that we are going first to emulate specific
individuals, and that any other kind of AGIs will be developed only later as
efforts to put together patchworks of actual (or hypothetical) human
The arguments behind such assumption, albeit not very technical in nature,
appear sound enough to me.
First, you have native speakers of a given language, then you generalise and
establish grammatical rules obeyed by an ideal speaker thereof.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the extropy-chat