FleshFactor: Tom's machine.
A E C F O R U M - "F L E S H F A C T O R"
Tom Sherman finished his last posting with:
>I don't want a pet or a slave, I want a friend, a colleague, a partner, a
>good listener and someone who can give me positive criticism and answers
>to questions I could never hope to formulate alone. I get along with
>people just fine, inside and outside my house. It's the relationships I
>have with my machines that depress me. I just know there can be more to
>our long days together.
Tom, if it is attention you want from your machines, I think your
situation is much graver than you make it out to be. Your machines don't
ignore you. They don't even have that option. In fact, their options are
so limited that referring to "their options" is deceptively
over-anthropomorphic (unless we interpret "their options" as one might
understand a "car's wheels").
So what would it take for Tom's tools to love him? Like him? Or simply
just ignore him? This may be a good time to play that childhood game of
"let's pretend." Let's pretend that all the necessary technology for a
machine to come to love Tom is possible. Let's pretend for a bit we have
this technology. And for the moment, let's pretend that this technology
takes place in many black boxes-all of which can be wired to each other,
made mobile, or whatever other configuration is necessary. Now, let's
have a little brainstorm which flies above and beyond all the rhetoric of
bio-electric enterprises. And for a bit, let's explore some of the
psycho-socio components necessary to build Tom's machine.
First off, this machine needs some means of sensing Tom. Personally, it
makes no difference to me if Tom's machine sees him as a particular smear
of thermal radiation, pixilated ultra-violet image patterns, a voice match
to a specific sound wave pattern, or simply as a verbal being with a
unique syntactical structure. For all I care, Tom could keep a private
keyboard from which the machine just assumes that any input it receives
comes from Tom. (Tom, does it make any difference to you how you are
perceived? Or would you rather be surprised?) [surprised, T.S.]
Second, Tom's machine would need a way of interfacing with the world.
(Remember, Tom wants a friend and a colleague. He wants something with
which he can compare perceptions of the world-something with more flexible
reflective abilities than a simple mirror.) Again, I don't particularly
care how his machine does this. It just needs to interface the world and
collect data independently of Tom Sherman. (Tom has already put forth
some potential preferences-a spectacular climbing machine or a language
reprocessor. Perhaps we will build him a roving, climbing, language
These two conceptual components of perception are straight forward enough.
But already they imply a new role on Tom's part. He must somehow allow or
encourage his machine to gain "life experiences." Exactly how this would
happen, for now, is up to our imaginations. For example, while Tom
prepares his morning coffee, the climbing language machine may wander the
neighborhood-eavesdropping on children as they wait for the school bus,
listening to college students as the bustle to morning classes. Or maybe
Tom's machine would never leave his side. It would always be there-in the
shower, at the in-laws, on his desk at work-just listening...always
listening. Or maybe this climber would be large enough for Tom to ride
in. Perhaps then, this would be Tom's means of transportation to art
openings. Of course, this might limit the climber's experience of the
world to the few short hours it was up and running-climbing over homes,
cars, fences, and trees, carrying him to art openings; milling with Tom
and the avant-garde while holding Tom, drink in hand, some three feet
above the crowd on six-foot chrome tendrils (like some comic book
character from the sixties).
Tom is not interested in transparency (machines that act human). For all
intents and purposes, his machine may not need particularly sophisticated
and inclusive life experiences. But no matter how crude, disjunct,
limited, or quirky they are, his machine requires the datum of life
Now we come to the hard part. Deciphering what happens conceptually in
that black box between the input of Tom's queries to the machine as well
as its life experiences and the output of insightful verbiage.
Additionally, Tom has put to us two rather complicated requests. First,
he wants his machine to have emotions for him. Not just any
emotions...but positive, affirming emotions. And second, Tom wants his
machine to sense his emotional state-and ultimately, be able to guess (as
we do with our loved ones) what the best course of action would be. It
should be able to know when to push him up to his limits, and conversely,
when to hold him and dry his tears.
Ultimately, Tom's machine needs to be able to empathize and, eventually,
acquire a certain symbiotic wisdom (however alien this wisdom may be). I
say "acquire" here because in his requests, Tom has already committed to
dedicating a certain effort to achieving a relationship with this machine.
>From his language, I sense he is willing to take a few emotional lumps if,
in the end, a friendship is achieved. So we needn't worry about
installing an "inherent wisdom" and all the philosophical/socio-political
implications of such an inheritance.
Quite honestly, I haven't thought through the details of many of these
dimensions (so please, whoever you are, jump in with your thoughts). But
I have conceptually explored this notion of "machine-love." Fortunately
for us at the Climbing-Language Rover Corp., while developing Tom's
machine, we've realized we might be able to cheat. You see, Tom is a
human. And though climbing-language rover's may be a new product line, we
have a lot of experience with these mammals. Our Corporate case studies
reveal a gullibility in humans. This means that our machine doesn't
actually need to love him. It only needs to convince him of its love for
Another point on our side-Tom Sherman does not have a degree in
black-box-micro-physics. So even if he does manage to get the screws off
the top, he won't be able to tell what's going on inside! (viva la
Though it has its advantages, Tom's intellectual background also makes
our job harder. In fact, he is of a breed of human we despise most,
second only to those with black-box-micro-physics degrees! Tom is a
watcher, a thinker, an engager. He contemplates. His kind can be hard to
fool. To create the machine he desires, we need to give it more than the
usual pornographic smattering of small talk and superficial niceties.
Parroting coos of "good mor-ning mis-ter sher-man. hows the cof-fee?"
won't do with this guy.
Tom, like many artists, philosophers, and writers, is simultaneously
skeptical and passionate. He may even prove to be the sort of person
foolhardy enough to risk love when a love relationship is not affirmed
occasionally by ordeal. He may, every-so-often, like so many of us, sail
that dervish that dares to ask
"But do you (still) love me?"
And then, what, then, will his custom Climbing-Language Rover 2050XE
"yes. tom. of course i do."
A client like Tom is too complex for an answer like that. He'd see right
through. Here at the Climbing-Language Rover Corp., we face a difficult
philosophical question-what good are unconditional expressions of
emotional love if there isn't the risk, however outside, that you may
someday lose the source of said expressions? And I don't mean lose that
love in an auto-wreck or an house-fire. But, more specifically, lose that
love while the loved remains? If there isn't some imaginary line in the
far emotional distance you have not yet crossed-something that can remind
you, "no matter how lousy you've been or no matter how taken for granted
you feel, I still love and acknowledge you (And in this case, "you may be
the only human around, but you are not alone.")."
And on the down side of this, we at the Climbing-Language Rover Corp. face
a logistical paradox. Tom wants a machine that loves him-truly loves him.
To satisfy our client we must deliver. We must create a machine that can
love, as well as potentially hate, but ultimately will love. (Of course,
if we are slick enough, we might just be able to get away with feigning
the love along with implications of a potential to hate. Or hows 'bout we
just ask our client-Mr. Sherman, what would it take to convince you?)
to (un)subscribe the Forum just mail to
firstname.lastname@example.org (message text 'subscribe'/'unsubscribe')
send messages to email@example.com