The emotional A.I. - A better parent than you?

Sounds like an absurd notion to even consider, doesn’t it? One thing is to let an intelligent machine square off in a battle of chess or the recently won trivia game by the amazing neurolingual “Watson” from IBM, but a whole another matter is leaving your kid with a computer algorithm. Even if the kid would emotionally survive the inital period, would you? Or to put it differently, would you willingly leave your child with an algorithm which at best can only imitate empathy and other subtle human interactions.
Or maybe one should look at this matter from another angle. I think it’s worth a try, because if you take away the human element, then just maybe it is exactly what is needed in a child’s upbringing? Let’s consider it for a moment. There are quite a few different parenting styles but if you want to simplify the subject matter for argument’s sake, I will split it into two rough camps, simply called mother style and father style.

 

0e1bf9fb22920ab97d92b6476fe32a6b

 

Mothers play often the nurturing side of parenting side. Mothers have a deeper sense of recognizing the baby’s specific needs, often much more so than the father. Call it a deeper emotional link. Some claim women in the mother role verbalize more adequatly the whole gamut of human emotions. This can be visible in Mother’s words of affirmation and being able to communicate issues where discipline is involved. But probably the most visible is the whole aspect of unconditional love. This kind of love which many claim manifests itself only between mother and child, brings with it  seldom experienced kind of altruism which may very well be hardwired into as as a species. Self sacrifice and putting children’s needs before one self is something men can not grasp or experience.

Fathers on the other hand are more inclined on having high expecation for their children and stimulating reaching them to reach those. So the weight is less on making the kid feel secure and more on helping them to prepare and toughening them up for the real world situations. Fathers are less verbal, more straight forward, using more pragmatic communication vocabulary. For the “mother style” orientated, the father style may be to disciplinary, too tough, but certainly this style helps to prepare the child for the hard realities of the world when they grow up. The element of unconditional love is certainly less pronounced in fathers, but may manifest itself more in the almost warrior like feeling of duty and need for protection of the whole family.

So now that we got some basic facts out of the way, where would the machines fall in these two categories? Would they be able to actually fulfil any of those roles adequately and if not would they maybe make up another altogether new category. Or maybe they would supplement the above mentioned parenting styles? It will be much easier to get to the core of this matter if we take a look at how different the machines think and might “feel” than what we would expect them to.

“I wonder who had the first computer dream, where, and when? I wonder if computers ever dream of humans.”
― David MitchellGhostwritten

Although it might become possible to “program” an A.I. to feel things like feelings of sickness, nausea, discomfort, jealousy, indigestion and so on, would one really want to? There is a whole gamut of feelings which maybe wouldn’t be besides the point for the machine to feel. Or would  machine need to experience all this discomfort to have an understanding of the other side of the spectrum? Would the A.I. have to experience soul crushing sadness to be able to feel euphoric joy? And even if this would be somehow possible to replicate, would this be the real thing? Or just some superficial imitation of the real thing? And woud an imitation be “good enough”to be able to be a care taker for our kids? Or to put it differently. Would you trust the machine if it told you it felt compassion for your kid and then leave it for the weekend together based on the assumption that the machine would never even consider hurting your kid since it has the compassion v.3.4 code running inside it.

“How will machines know what we value if we don’t know ourselves?”  – John C. Havens

twin-babies-02

Everything can be emulated and there is a very strong possibility it will be replicated to perfection. To such an extent that no human will ever be able to tell the difference. But based on that, will the machine actually ever have the surge of an emotion which provokes it to learn more about itself? If one considers the nature of emotions, they are there as an instrument to learn about ourselves. To gauge the truth about ourselves. You can’t fool the body can you? So here is the conundrum. Emotions come first and come from a place which we may never understand. So how do you replicate something as ephemeral as the truth which causes you to grow, to have insight, to mature as a human being?

“Emotions are a human feature. Will AI someday really feel? That’s like asking if submarines swim. If you call it swimming then robots will feel, yes.”  Noam Chomsky

Screen Shot 2015-09-20 at 20.46.56

Personally I think it’s a hack job. Imitating something we can’t explain at best will lead to a hack job. One which certainly won’t be easy to discern ut which at the core level is a hack job and reverse engineering human nature which we as humans have little clue about.

“You can’t hack your destiny, brute force…you need a back door, a side channel into Life.”
― Clyde DeSouza

But wait. Although we might never be able to understand ourselves – the unpredictaabliltiy of human nature, maybe this is not such a bad thing. Maybe, just maybe we don’t want true human nature inside those machines? The same human nature which led to the second world war and the eradication of millions of people. The same evil which leads us to the earth shattering oint of no return which is climate change. Maybe waht we need is human nature 2.0. A safer and more friendly version of urselves which will actually keep some sanity in the world? And with which we can feel safe to leave our kids and not come back and find them chipped away by some unppredictable psychopath babysitter?

“It’s amazing how once the mind is free of emotional pollution, logic and clarity emerge.”
― Clyde DeSouzaMemories With Maya

Ok. So this is maybe stretching it. But consider the more day to day aspects of it. Most parents don’t have the emotional or communicative capacity to effectively handle everything their kids throw at them. We certainly can’t be everything to our kids 24/7. We simply don’t have the capacity physically or mentally.  So an all coping, tireless A.I. would not even flinch or hesitate after the longest attention seeking bouts from the kids,

“I think they are a better race than humans ever were.”
― Angelo TsanatelisDirective 3.1

Despite this gross oversimplification, machines would probably have a much superior emotionally error proof existence. Just consider the countless inner conflicts and tomrent we are able to bring on ourselves daily. What kind of freedom must it be to be rid of that? nd what kind of performance would that give a machine attuned only to the needs of the baby?  And what about the unconditional love aspect? Yes, the machine would only replicate this emotion, never being able to touch upon the deeper unexplained emotion which is hardwire into us. But how far does this unconditional love stretch? Would we really do anything in the face of death,and unfathomable fear? Maybe we would but the machine would never even haev a microsecond of a doubt or be conflicted before it protects the child.

 

Brain2

 

 

 

And what about health, medical care and daily parenting chores? An A.I. as the perfect nanny you might say. Well here is a real world example of where teh technology is heading currently. A Claremont-based startup has developed an app to help families record, analyse and report information about their child’s development, which they can then use to gain better insights, as well as share data with health and education providers.

Called Kindom, the app has been developed by KinChip Systems directors Gry Stene and Catherine Resnick, who have turned to crowdfunding site Indiegogo with an initial target of raising $US50,000 to bring their product to market. The two businesswomen behind this endeavour, have backgrounds in software engineering and child psychotherapy respectively, and are hoping to garner support from parents who can provide feedback on the app and also raise funds to speed up the development of the app’s artificial intelligence capabilities used in Apple’s Siri and Google Now programs. So here is the

The women say:

“Say, for example, my three year old was developing normally but then has suddenly stopped talking. The app looks at all the data and it says … that could be because a new baby was born, it could be because dad’s in a new job and he’s really stressed. It could be something more serious, it could be autism spectrum disorder. So it asks questions related to that and looks at the data that you’ve collected such as is he doing repetitive behaviours as well as not speaking. It would say these are the things we’ve seen, do you think it’s this? And then you can narrow it down yourself. You get all that knowledge, you curate it and you actually train it to answer questions based on your own data that you’ve collected. ” she said.

So this interesting example only goes to show that the technology is opening up a whole new area of on-the-fly diagnostic possibilities and safety which the parents wouldn’t earlier have time/money for.

So, considering all of this, who would you trust with your child when going away? Personally I think we are onto something. I believe machines would not really make any of the common human mistakes. They will of course never replace the biological parents per se. But  as long as we reach a point where they would be able to effectively empathize and display affectionate and disciplinary behaviors they could make ideal parents.

o-ARTIFICIAL-INTELLIGENCE-facebook-3-940x564
Or maybe everything is just a little bit less complex. At least according to Terry Pratchet:
“Maybe the only significant difference between a really smart simulation and a human being was the noise they made when you punched them.”
― Terry PratchettThe Long Earth