Ron Barnette's Zeno's Coffeehouse Challenge #58 Result
This challenge prompted over 200 replies, which is appreciated! I have listed below the original challenge, followed by several respondents' thoughts on the matter. I want to thank ALL respondents for their thoughtful time taken with Zeno's Coffeehouse, and to encourage your continued support, as critical thinking exercises are explored for mental growth and recreation. Minds need exercising with shared, reflective thinking.
Thanks!...Ron Barnette
A perception and reality experiment was conducted sometime back, which has become a significant discussion item at the Coffeehouse. Many patrons have debated its relevance and its methodology; some are even convinced that its alleged claims are true. The experiment involved a so-called 'Mind Reading Program' posed as a thought experiment, for analysis in a university class, available here: http://www.ronbarnette.com/mental.htm.
Charles and Maggie now want to challenge you Zeno's patrons to check out this experiment, to analyze it, and to share your thoughts as to why so many folks have found the experiment convincing...or at least worth debating about. Come on....mind reading over the internet is impossible! Right? Something must be wrong. What's the catch? What's going on here? And why does it seem to work for so many? Your comments are invited, as your thoughtful responses can help illuminate some valuable reactions to this phenomenon, which has been discussed vigorously at Zeno's.
Some sample responses:
Mark Young in Canada writes:
comments: It seems to work because you are thinking only about one
card, and when that card fails to appear in the second list, the most accessible
explanation is the one offered by the program -- that that card was removed from the set
of cards you'd seen before. In order to notice what's going on, you need to notice
and think about two (or more) cards, but the instructions discourage that by instructing
you to "select *a* card and focus on *it*" (emphasis added). The social
role we are asked to play is familiar to us, and no trouble for us at all. It's too
easy to do, and so we (in our normal, socially cooperative way) do it. And in doing
it, we blind ourselves to the solution. Like any good magician, this program uses
our preconceptions to hide the trick from us.
Our tendency to cooperate, especially in communication, is what makes Eliza/Doctor and her
descendants so realistic. The program uses a bag of tricks (smaller or larger) to
translate our cooperative comments into natural seeming responses -- or to convincingly
return to some other subject when the going gets rough. The early versions of the
program were quite easy to find out: throw nonsense words into otherwise grammatical
constructions and watch them come at you. Just as with the mind-reading program, the
trick is uncovered by *not* cooperating.
The glibness of these chatting programs cause many people to object (rightly) to (the
popular misconception of) the Turing Test. This "test" says that any
machine that fools a human into thinking that it is human is intelligent. Eliza has
(supposedly) fooled some people into thinking that it is human, and it is clearly not
intelligent, and so the "test" is bogus.
Well, sure -- but that's not Turing's Test. Turing was quite clear that the test
involves a game -- a competition between two humans and a machine in which the person to
be fooled is well aware that one of the other participants is a computer. Indeed,
the judge is to tell the computer from the human. And winning one game is not enuf
to pass the test. The program must be able to perform at the level of one thinking
being imitating another: the man in the man-woman version of the imitation game.
The imitation game is set up so that the judge is an adversary of one of the other
players. That should encourage the judge to be less cooperative with the other
players than would otherwise be the case (as would be the case in the "secret
chatbot" version of the "test" mentioned above). But even if it
doesn't, using the man-woman version as a control case means that the tendency to
cooperate with our conversational (and magic-trick) partners does not give the computer an
undue!
advantage.
Interestingly (at least to me :-) Turing thought that one way a judge might be able to
find out the machine would be to use ESP! I guess the mind-reading program shows
that Turing was wrong about that....
Oli Bramford in the UK writes:
comments: Like any well executed 'magic' trick, this Mind Reading
Program expertly used misdirection to produce a seemingly miracuous result. It is
designed to make us think that our card will be guesses by analyzing how we move the mouse
- this is the only actual input we put in while supposedly having our minds read, so we
would logically deduce that the result of the trick depends entirely upon this.
The program is designed so that we are stunned with amazement when it inevitably works,
and want to immediately repeat it to prove it wrong, by altering our mouse movement.
By eliciting this state within us, and getting us to look specifically for OUR
chosen card, most of us neglect to notice a completely differnt way in which the trick
could work, that is, the way it actually does work: All the cards are replaced so
the trick has to work every time.
Sergei writes:
comments: Congrats, you've developed a way to read minds over the
internet. The only way I could get it to not work was to think one way and say another.
Keep up the good work.
Topi Linkala in Finland writes:
comments: Hello,
this one works for those who only concentrate one one card.
What ever card you consentrate on is missiing from those that are later shown.
Some might ask why is that, but it is quite simple.
The cards shown are king of diamonds and clubs, queen and jack of hearts and spades.
After you click the eyes you see king of hearts and spades, queens of diamonds and clubs
and a jack of diamonds.
So what ever you were focusing on is missing as none of the original cards are in the new
ones.
Further more the new cards are shown in flashes so noticing the whole change is made
harder.
Only on not following the rules ie. memorizing all the cards can you notice what is going
on.
And from Emilie Clark in the UK I received this jewel!:
comments: Sometimes I wear a low cut top so people don't pay
attention to
my backside. Its very clever really and so simple ! People are constantly
having their attention channelled or 'manipulated' to get the desired
result. I'll try this on my mates with a pack of cards...can't really go wrong can it ?
And thanks to all Zeno's contributors!!!