Members of the Hallow team and I recently attended a talk by Kevin Kelly, co-founder of the Wired magazine, titled “The Future of Spirituality and Technology.” The hour-long discussion was fascinating and covered a wide range of topics, starting with humans creating stone tools and progressing to our current pursuit of building an artificial intelligence (“AI”).
Looking around today’s technology ecosystem, there is plenty of buzz surrounding new, world-changing technologies: blockchain, quantum computing, and nanotechnology, but it seems like AI falls in a totally different category.
Rather than focusing on productive potential, the conversation around AI almost always quickly becomes a discussion on ethical dilemmas and existential threat.
As far as I can tell there are two main (and related) reasons why we are afraid of AI. The first is that AI is inherently uncontrollable. The second is that it could reject and try to destroy us.
By reflecting on these fears, I realized that developing artificial intelligences can offer incredible insight into our own nature as spiritual intelligences (SIs), created by God.
Uncontrollable – Think about how many hard decisions we face every time we drive our cars. Someone pulls out in front of us and we are going to fast to stop – and there is also another car coming in the opposite direction, so we can’t swerve out of the way. What should we do?
Even if we as humans could figure out what the ethically “right” choice is, how do we teach it to self-driving cars that “think” for themselves?
Rejection – Whether you have a PhD in neural nets or just have a general awareness of the Terminator movies, you have probably seen a headline alluding to some doomsday scenario caused by AI.
The general logic goes something like this: a powerful enough AI could decide that we puny humans are simply inefficient (or worse – that we make the universe worse off through dishonesty, crime, and war) and that it’s time for Homo Sapiens to pass into the history of evolution like the Neanderthals before us.
So…
The real issue that we are grappling with is how to deal with creating something that is uncontrollable and that could reject us as either inefficient or evil.
Sound like any other challenge in life?
Forget about creating artificial intelligence, how about creating natural intelligence? How about raising children?
No matter how tight of a ship parents run, I’m not sure that a single parent in the world would ever make the claim that they can actually control their kids.
Parents do their best to teach their kids how to drive, but at some point the keys are going to be in the hands of the child, and they are going to need to be the ones who decide whether they are going to obey the speed limit.
So how are we supposed to parent? And can that help us better contemplate our relationship with AI?
Humans as children and as parents
One path forward might be to start not with HOW to parent, but WHY we are called to be parents in the first place. Why are we as humans designed to conceive as couples, bear children that are wholly incapable of taking care of themselves, and guide them into adulthood?
I would contest that it’s because we are made in the image of a God who is relationship, love, creation, and sacrifice. Out of the Trinity, 3 persons in one spirit, we were created by Love and imbued with that same creative nature.
But that’s not all…
We weren’t abandoned at birth or made to serve our creator by force. We were given free will and even after we fell, God continued to lead us down the path of spiritual development, the whole way to salvation via His ultimate self-sacrifice on the cross.
In parenting, we are called to do the same. We do our best to demonstrate an example of love and to teach our children the best ways we know how to live, but ultimately we know they will stumble, often in the same ways that we stumble.
The real fear
Maybe the idea of artificial intelligence is scary for the same reason raising children is scary. Children are mirrors that force us to look at our true selves – how we actually live, not the idealized facades we try to project into the world. “Do as I say, not as I do” never works as a parenting strategy, because the human mind learns by mimicking the actions it observes in others, not just by listening to commands.
Just like kids, AI systems “learn” from real world data. They look at a bunch of decisions that have been made, then they look at the outcomes, and by comparing the two, they “learn” relationships and principles that they can use to navigate future scenarios.
Maybe we are scared of AI concluding we should become extinct, because we know that the “training data” that we have created in the world might not show the best we have to offer.
The holy, parental heart
It is in this realization then (i.e., that AIs would be our creations in the same way that we are God’s created SIs) that we understand the true beauty and real danger of pursuing the creation of an artificial intelligence.
The beauty – by reflecting on our freedom and our history, we see how much God has loved us. Despite how consistently we have fallen, He has continually set the example of a forgiving parent, even to the point of redeeming our sins on the cross.
The danger – overlooking the fact that the example we set with how we live our lives may very well be determining the fate of our species. Unless we follow and set the example of Christ’s love in the world, both the natural and artificial intelligences we create will stumble in the same ways that we stumble.
We will never be perfect, but what if we use this moment in human history as a chance to reflect on God’s example as the ultimate parent – loving and sacrificing unconditionally? What if we lived every day trying to teach by example so that our AI could also work to make the world a better place?
God bless,
Alessandro
“By their fruits you will know them” (Matt. 7:16)