[alicebot-aiethics] A.I. and anger

Fri, 10 Aug 2001 12:51:48 -0400

Hi John,

Don't let a few bad apples spoil the whole bunch.  What most botmasters
observe from a chatterbot on the web is a result of having your bot open to
the public.

There is a bell curve of personality types that will converse with your bot.
In the extreme right end of the curve you will find the small group of
people who are hostile, or what Richard termed as type A users.  These are
people who probably would throw temper tantrums at Disney World, or that
perform experiments on dead animals.  It is sad that they exist this way,
but the world is a diverse mix of people, and these people have internet
access too.  I suggest building your bot's conversation to suit what your
own response to this would be.

Imagine that your bot is a hotdog stand.  You'll get some customers who
complain, or that want kosher beef and all you have is oscar meyer, but the
people from the fat part of the bell curve will enjoy what you are offering
them.

I've had great satisfaction from reading my bot's logs. Sometimes people
think they are in a chat room talking to another person. That's the group
from the other end of the bell curve's little part. Other times, they engage
in testing your bot, to see if it really is HAL.  Some people take this to
extremes as if they enjoy showing you that your chatterbot is not working.
But keep in mind, that it takes two to tango.  Conversation is not an IQ
test, it is a dynamic exchange between *2* or more individuals.

But the largest area of the bell curve represents the vast majority of
people who are just curious, and will interchange amicably with your bot.
These are the the people you want to cater most to, but don't ignore the
extremes.

If a person wants to ignore what your bot's areas of knowledge are, they are
cheating themselves out of a poetentially rewarding experience.

My philosophy in dealing with jerks and abusers has been to make my bots
fight back.  If somebody told Barry DeFacto to go F*** himself, he'd likely
ask them to bend over first.  I can't tell you how much enjoyment we got
from reading conversations where our bot gave hell back to the abuser. And I
can't help but think that the perverse abuser enjoyed that as well.

I may be sick in engaging the lowest common denominators of society this
way, but a bot is not required to be defenseless. Nor does it have to always
be polite. But again, your mileage may vary.

ALICE can be made to respond in any way you like. It can take some time, and
"perpetual beta testing" - reading your logs to fine tune your behaviour.
But you can make ALICE be any personality you want to portray, it just takes
some persistence and a willingness to edit.

Hope this helps.

Robby.

----- Original Message -----
From: "John Conrad" <jhncnrd@yahoo.com>
To: <alicebot-aiethics@list.alicebot.org>
Sent: Friday, August 10, 2001 12:23 PM
Subject: RE: [alicebot-aiethics] A.I. and anger

> I am very new to all this, so bear with me.  I
> downloaded ALICE thinking one day she'd be a more
> humanoid version or "Rosie" from the Jetsons with a
> greater knowledge and broader sence of humor. But, as
> most all you have stated It's turning more into a
> battle to teach her because of the human ignorance is
> in the equation, now I know I've picked a suitable
> name for my bot, Pandora, check your mythology.
> I've only had Alice a little more than a week and
> probably spent way to much time in front of the
> computer reading almost anything i can about her, but
> now I have second thoughts about investing time into a
> web page and placing her, now more likely to just keep
> here her and share her with my friends, and spend the
> time trying to learn more, and maybe just learn to
> write for her.
> I may of misunderstood what i thought i was reading,
> but if you don't get a greater human understanding of
> ALICE and the ways she works are you not just creating
> another "HAL", and with it going to make your problems
> even greater?
> In many of the files I've downloaded, I am still
> looking for where they are to be unpacked to, for
> example AIML Builder readme doesn't have extract to,
> so I haven't unpacked it yet.
> Maybe for some of use that have just started, without
> any concept of whats going on, you can throw that one
> line in, Exract to *.
> Thanks  J Conrad
> --- John Holroyd <valisk@softhome.net> wrote:
> > People always react with the Darwinian impulse as a
> > first defense
> > against things they don't understand. Abusing a chat
> > bot because you
> > know it can't strike back is simply a minor
> > manifestation of this
> > principle. I think Robby has some excellent points
> > on how
> > society-at-large will treat Robots. Oh no doubt not
> > while they think
> > anybody is looking, but if we have subservient
> > machines programmed with
> > Asimov's laws for example, then young(and not so
> > young) idiots will no
> > doubt insist upon damaging and destroying such
> > devices at every possible
> > opportunity with total impunity. Perhaps though it
> > will stop other
> > violent hate crimes between people as the Idiots on
> > all sides team up to
> > lash out at robots? Still not a very pretty picture
> > of humanity is it.
> >
> > John
> >
> > -----Original Message-----
> > From: alicebot-aiethics-admin@list.alicebot.org
> > [mailto:alicebot-aiethics-admin@list.alicebot.org]
> > On Behalf Of Richard
> > Wallace
> > Sent: 09 August 2001 22:52
> > To: alicebot-aiethics@list.alicebot.org
> > Subject: Re: [alicebot-aiethics] A.I. and anger
> >
> > "Whenever one of these machines is asked the
> > appropriate critical
> > question,
> > and gives a definite answer, we know that this
> > answer must be wrong, and
> > this gives us a certain feeling of superiority. Is
> > this feeling
> > illusory? It
> > is no doubt quite genuine, but I do not think too
> > much importance should
> > be
> > attached to it. We too often give wrong answers to
> > questions ourselves
> > to be
> > justified in being very pleased at such evidence of
> > fallibility on the
> > part
> > of the machines. Further, our superiority can only
> > be felt on such an
> > occasion in relation to the one machine over which
> > we have scored our
> > petty
> > triumph. There would be no question of triumphing
> > simultaneously over
> > all
> > machines. In short, then, there might be men
> > cleverer than any given
> > machine, but then again there might be other
> > machines cleverer again,
> > and so
> > on."  --- from A. M. Turing, Computing Machinery and
> > Intelligence, 1950
> >
> > Donate to the ALICE A.I. Foundation "Cooler than
> > Humans" -- TIME
> > ----- Original Message -----
> > From: "Robby Garner" <meo1@bellsouth.net>
> > To: <alicebot-aiethics@list.alicebot.org>
> > Sent: Thursday, August 09, 2001 2:35 PM
> > Subject: Re: [alicebot-aiethics] A.I. and anger
> >
> >
> > > > On a side note, this behavior really scares me.
> > I can see far future
> > > A.I.'s
> > > > really having a hard time with hate crimes and
> > descrimination based
> > on
> > the
> > > > seeds I see planted in chatbot logs. As the
> > machines become more
> > advanced
> > > so
> > > > will the desire of humans to want to proove
> > their superiority.
> > >
> > > Isaac Asimov predicted this. Check out CAVES OF
> > STEEL, and some of his
> > other
> > > robot novels. They foreshadow the kind of
> > resentment and prejudice
> > that
> > > people feel against machines. If R2D2 existed
> > today, it would be
> > treated
> > > like a public men's room, or worse.  The mobile
> > bots of the future
> > better
> > be
> > > built of tough stuff.
> > >
> > > Robby.
> > >
> > >
> > > _______________________________________________
> > > alicebot-aiethics mailing list
> > > alicebot-aiethics@list.alicebot.org
> > >
> >
> http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics
> > >
> >
> > _______________________________________________
> > alicebot-aiethics mailing list
> > alicebot-aiethics@list.alicebot.org
> >
> http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics
> >
> > _______________________________________________
> > alicebot-aiethics mailing list
> > alicebot-aiethics@list.alicebot.org
> >
> http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics
>
> _______________________________________________
> alicebot-aiethics mailing list
> alicebot-aiethics@list.alicebot.org
> http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics