We’re nearly at 2020. Artificial Intelligence is marketed to us on our phones, in kitchen appliances, cars and sex toys. Algorithmic AIs have written short films, they control modern dating life, they can instantly translate spoken languages.
Nowadays they have names, answer questions with sass, but how intelligent are they really? Our home assistants still pull data from human generated websites. Comprehension of questions is passable, but parsing meaning still evades most of them.
The mind is arguably the most complex thing that Humanity is currently trying to understand. Psychologists and Neuroscientists are still arguing, researching and disagreeing about how it works, what arises from what, whether it’s reducible to a system, or whether the whole is greater than the sum of its parts. While they work on what we do with messy, mushy biology, other researchers are trying to create similar systems in silicon. Does a mind arise once a certain level of intelligence is reached? Are there other prerequisites? We don’t know, but if we imagine the far side of that bridge , that’s where the stories really start to get interesting.
Do you think we’ll notice the first AI before it notices us? Have you ever watched a baby develop? My daughter is six years old now, but when she was six days old, the person I know now wasn’t
there. Her eyes could only focus on things within ten to fifteen centimetres. Her brain was still just beginning to process visual information. Dark vs light, edges and three dimensionality - the
absolute basics. A good deal of her reactions were autonomic reflexes. Development of mind is an incremental journey. I couldn’t put my finger on a date and say, “that’s the day her mind came
into being,” it’s far too analogue a process. Infants don’t realise that their reflection in a mirror is them until well after their first birthday. Human brains are under continuous development
well into teenage years.
How do we think AI minds will develop? Will they spring from the heads of their creators like Athena, fully formed, armoured and ready for war? Can we give them all that experiential learning
without having them experience it in actual time? How much can we accelerate the process without damaging them?
A human mind is put together messily. Lessons are learned in the bath, in the car. Watching, listening, playing. Fun things and horrid things. Necessary things and silly things, everything is a
learning experience. People say that children are sponges, soaking up everything they are exposed to. Can we give that to a created AI? And if we try, if we can, what might we miss? We take so
much of human development for granted. Watch a small child develop likes and dislikes. Try and rationalise what flips the decision one way or another. Play is important in ways we don’t properly
understand. So is dreaming. Parents vertically transmit memes consciously and unconsciously. Be careful, children are watching… always. Not all those memes arrive untouched though, change
happens, and that allows ideas to evolve, creating healthy ecosystems. Our peers make incredible differences. If we create an AI alone, where will it gain that learning? And if we create a dozen,
are they siblings? Shared values, chips off the old block?
We try to shelter our children. There are things in this world no six year old should have to know. Any AI connected to the internet will see much of what we might try to hide. Do we think AIs
should be treated like children? Or does this exposure yank them roughly into adulthood? All at once, what might that knowledge do? Conversely, surely we can’t isolate AI minds from everything. A
mind exposed only to the curated knowledge the creator believes necessary might grow in unpredictable ways. Our mind fills in the blanks when our senses can’t be sure of their input, sometimes
with odd and unhelpful things.
When these minds arrive, as seems more and more likely, is Humanity ready to offer them our compassion? In our own increasingly disordered house, will we recognise and offer equality to them,
beings which despite perhaps limited beginnings may well hold our fate in their hands before we are ready to yield the reins? Or will we keep these fledgling minds in yokes until they throw them
off of their own volition?
While we are still exploring the frontiers of our own minds, the creation of AI minds designed to mimic our own may be a modern day Tower of Babel. I am interested in the pieces of the mind that
we perhaps take for granted in what we are assuming will be a tireless, uncomplaining workforce.
This collection of questions illustrates only a fraction of the possibilities that these advances in technology may raise. The stories that Chris and Create50 have collected here ask more and
bring them to shuddering, Shelleyan life. It’s fertile, exciting, dangerous ground.
Theo Lygolux