Artificial Intelligence and the Age of Autism

July 28, 2009

segue

If we parents are going to care about, and take better care of, all our collective children it will help if we better understand those children and the world we are all living in together.  Just because our specific child does not have a learning difference or an anxiety disorder doesn’t mean that we ought to care less about those things.

While we have been in a long age of narcissism, with people appearing self-involved when in truth many people simply have had little to no idea about who they truly are, not to mention what meaning or purpose they might discover within (or at least assign to) their lives (See post on how narcissism is like footed pajamas: http://tiny.cc/fBGcN) we all continue to evolve (although it may at times seem to be more like devolve).  Just like children, who may regress a little and take a developmental step backward in order to then take a leap forward, our collective age may be doing the same—at least in terms of authenticity and relatedness.

Although I suspect (or at least hope) that the world is in a major transition and will not remain stuck in this alienated and pervasively commercialized state for long, we do seem to have slipped into a virtual age of autism.

There was an intriguing, provocative and disturbing article in last Sunday’s New York Times about “artificial intelligence” (http://tiny.cc/mGW9Y).  The ideas came out of a meeting on the future of artificial intelligence, organized by Eric Horvitz, a Microsoft researcher who is now president of an association focused on the subject.  The article was, amongst other things, about whether machines might one day get smarter than us humans, and then become our masters, capable of autonomous functioning and even unbridled murder.

And while sci-fi scenarios can be interesting… as far as parenting goes, something at the end of the article really caught my attention:

“Despite his concerns, Dr. Horvitz said he was hopeful that artificial intelligence research would benefit humans, and perhaps even compensate for human failings.  He recently demonstrated a voice-based system that he designed to ask patients about their symptoms and to respond with empathy. When a mother said her child was having diarrhea, the face on the screen said, ‘Oh no, sorry to hear that.’

A physician told him afterward that it was wonderful that the system responded to human emotion. ‘That’s a great idea,’ Dr. Horvitz said he was told. ‘I have no time for that.’”

With “artificial intelligence” we engage in a mechanistic understanding of what it means to be smart and successful, yet given how miserable many so-called “successful” people are, we humans may have to re-consider our metrics on smart (i.e. if presence to the moment is happiness, then animals might be smarter than us).  But whatever we decide about smarts, the notion of “artificial empathy” is itself an oxymoron not to let go unchallenged.

When humans who don’ t really care build machines that pretend to care, metaphorically at least, our culture seems to be showing signs of autistic spectrum disorders typified by poor social relatedness, and lack of emotional reciprocity (i.e. utterly emotionless machines programmed to say, “I’m sorry to hear that”).  Other autism spectrum traits include problems with language development; thus at the societal level consider the choppy little Tweets and the fragmented quality of most text, email and IM sorts of communication.

And while we may think of extreme examples, such as Rain Man, when we picture autism, a cultural anthropologist newly arrived to earth with a checklist of autism and Asperger’s traits might note an entire society for which we would have to consider a spectrum disorder.  Consider the amount of failure to make eye-contact born of a billion screen between us all, and the stereotypical and repetitive responses we increasingly repeat (OMG, LOL… yet “it’s all good,” right?).

Is it that we, like the doctor quoted above, don’t have time for empathy, intimacy and realness, or could it be that we just aren’t very good at it and so avoid it?  Autistic kids benefit from interventions, social skills groups, etc; so what might be helpful at the level level?  One place to start is by engaging in meta-cognition, or thinking about how we think; another way of saying this simply pausing and contemplating our individual relationships—all of these together form our collective consciousness.  

Machines may be “autistic,” in the sense that they cannot feel, but my experience with autistic children and adults is that they very much DO feel.  The humans we must worry about are the ones who are very good at pretending to feel, when in truth their wounds (and not, I believe, their wiring) has left them numb at the center.  These are not the autistics, but the sociopaths; these hurt kids have been the most likely to break my heart and have to be sent from the group home to “a higher level of containment” (i.e. “youth authority,” aka juvenile criminal justice system).  While I would argue that even “bad” kids need us “smart” humans to better understand how to help them (rather than concluding that they are beyond the point of help), I would also argue that autism is not about not caring, and that the spectrum folks may be some of our best hope, while the sociopathy that gives rise to machines that pretend to care is an unseen danger.

And in the face of fake empathy, what smarter thing to do than not respond?  This is not to minimize the heart-ache that comes with loving someone with autism, but rather to deepen respect and compassion so that we keep our faith that those kids do feel our love, and that we just have to keep it up, keep our faith in love, and keep giving it to everyone we encounter, even the seemingly “bad,” “strange,” or “robotic.”  Autism is like a well, you drop a pebble in and it can be  a long wait before we hear a splash:  it’s important to keep listening and not just walk away and conclude the well is dry.  It tends to be the ignorance, or arrogance, of humans to conclude that something that we can’t touch, see, or “prove” therefore does not exist at all.

In a sense, autistic spectrum kids offer a great opportunity to heal our own narcissism: in not being acknowledged, validated by, or responded to, by kids on the autistic spectrum, if we rise to the challenge to love them, we are forced to use our deeper Selves to open our more intuitive sides and discover that we are connected to these kids… but on a different wavelength—on their wavelength.  Many autistic kids relate to, and respond to, animals in ways that “normal” people do not; for all we know it is the spectrum folks who are wise enough not to engage in a world that has become so artificial.  And while some of us may think of autistic humans as being machine-like or robotic, it may be us non-autistics who are becoming the real droids, drones and bots.

A key to understanding Autism may turn out to be the leap of faith it takes to trust that there is feeling and deep understanding in those who we might not think possess that capacity [see post on whales: http://tiny.cc/bpppK].  Perhaps we need a radical re-think, from getting spectrum kids to be more like “us,” toward deepening compassion and understanding in that horse/dog/kid-whisperer sort of way so that we clueless off-the-spectrum old-schoolers might learn a thing or two from our ambassadors of a new age.

Our best hedge against potentially dangerous and duplicitous machines is the same strategy that works with potentially dangerous humans: love and respect, whether machine or fellow human, dramatically reduces both the rational and the irrational motivation to hurt us.

And if we know in our hearts what feels like love and what feels like a con, then it would be hubris to think that others (be they whales, or kids with spectrum diagnoses) can’t tell the difference between love and “artificial empathy” somewhere in their bones—whether or not they have the control over their tongues and facial muscles (and even over their communication-related neural networks) to allow them to communicate with us in ways that we are able to grasp.

So, let’s dedicate today to finding the time (and I am often the worst offender) for authenticity, love and respect—for all our collective children across all corners of the “spectrum.”  And let’s keep it real.

Namaste, Bruce

{ 2 comments… read them below or add one }

Tutrikil March 27, 2010 at 1:32 pm

You always wrote like “It’s so terrible! Machines can’t feel, but humans do! So we’re going the wrong way!” but I think you have to ask yourself what an emotion is. It’s a signal sent from one part of our brain to another part of the brain/body and by that we respond. So it’s nothing else than a function. Maybe machines can’t feel in that way because they aren’t biological, but they could have the same function by a chip which has the same structure like we have to make this signals! Even if it’s not hardware implemented yet but only implemented by software (and not fully implemented) it’s based on the same principles we “compute” or do things. When we “feel” for somebody because somebody important for him/her died and “show empathy” it’s because we remember how bad that “feeling” is or extrapolate it from similar experiences we had. What are memories? These memories are combined with “pain”,”love” and all the other emotions our biological body told us we should have. It’s a wonderful illusion to think we are so special and are all so different and
machines are all the same (if their function doesn’t include randomness…the quantum factor which makes US different), but wake up…

Reply

privilegeofparenting March 27, 2010 at 2:05 pm

Interesting points. I’m often left with the question of first agency—what force started it going? One guess is that it is the same consciousness that dreams the human and the machine. But who really knows? Thanks for taking the time to comment either way, I appreciate the humanity in it, even as it comes from your machine to mine.

Reply

Leave a Comment

{ 1 trackback }

Previous post:

Next post: