It's the attitude as well as the latitude I was getting at.
Immaturity is the inability to use one's understanding without
guidance from another. This immaturity is self-imposed when its
cause lies not in lack of understanding, but in lack of resolve
and courage to use it without guidance from another.
"Have courage to use your own understanding!" -- that is the
motto of enlightenment.
(Immanuel Kant, 1784)
http://www.english.upenn.edu/~mgamer/Etexts/kant.html
This much about the attitude that end-users may all too easily
seduced into by systems that appear to possess intelligence. (Kant
may hardly have anticipated that the "another" in his statement might
once be a machine, but it couldn't have amused him.)
The latitude in the lavish use of metaphorics that developers
indulge in, and preferably anthropomorphic terms, like "intelligence",
becomes problematic when unprepared laypersons get confronted with it:
their expectations tend to leap far beyond the bounds of reason. Then,
they might all too readily abandon their courage to use their own
understanding. IOW, they succumb to the very attitude Kant nailed down
in his famous words in 1784. Even earlier, a well-known document
granted the "Pursuit of Happiness" as a basic human right - not the
right to obtain "Happiness" without effort.
Now, you rightfully bet I _do_ use Google's helpful hints. But I'm a
developer myself with quite a few years of work behind me. It still
horrifies me to see users of my software ready to accept guidance and
judgement and thinking from the machine. "Don't make me think!" is
a disastrous motto in Kant's regard, but people adopt it without a
second thought - without a first indeed when it comes to information
seeking.
What Janet Hill's mother reacted to was the fact that a machine
addressed her in a way she wouldn't have tolerated from a human in the
same situation. It was not the fact that it _was_ a machine that
addressed her! But _that's_ what matters and what the end-user should
always be acutely aware of.
My advice was _not_ to leave end-users on their own to figure out
what to do in a "zero hit" situation. My advice was just to avoid
patronizing and anthropomorphic language. More directly: avoid all
forms of personal address, questions, or even the word "please",
not to mention "welcome" and nonsense like that.
Try to present facts and suggestions in a neutral way, not in fact
declared as "suggestions" - there is no one who has the capacity to
issue suggestions in that setup. Not "Did you mean ..." but perhaps
"Maybe worth a look, maybe not: ...". And of course not a dry
"Sorry, your search found nothing" but something like
"Zero results for this search. There may be other terms to try,
like perhaps these: ... Or entirely different ones, like synomyms.
It is always a good idea to think of other approaches."
You get the idea. Try something that sets the user thinking, not
just clicking.
Thomas Dowling's statement:
> The real problem here is if the machine decides it knows better than
> you, replaces your search with its suggestion, and doesn't let you
> override it.
points to the extreme that of course no catalog or so-called "searc
engine" (an anthropomorphic term!) should be allowed to go to.
Received on Mon Feb 23 2009 - 02:35:32 EST