Thank you for offering to listen to my ideas about AI during church the other day. I wanted to share with you some esoteric knowledge about technology and God.
A dichotomy for context: Interpolation vs Extrapolation
Interpolation is a type of estimation. “A, therefore B.” Interpolation is the method of constructing (finding) new data points based on the range of a set of known points. It’s the reasoning of associating, of instinct.
Extrapolation is a type of estimation, beyond the original observation, of a value on the basis of its relationship with another variable. Extrapolation allows us to understand or predict the future with available data or information. It’s abductive reasoning that leads to critical thinking. This is how we make decisions and what separates us from animals.
Artificial Intelligence models are increasingly capable of extrapolation through a process called deep learning. I work with AI models regularly and have become increasingly concerned since deep learning works much like a neural network in the brain.
Brain neurons are essentially electrical devices. There are many channels sitting in the cell that allow positive or negative ions to flow into and out of the cell. A “dendrite” is where a neuron receives input from other cells, and this input generates “Action potential”, or a Brief electrical event that signals the neuron as active. Your brain contains billions of these neurons which activate in a complicated web to build a relationship with what we perceive, with what we remember.
Deep learning neurons work in the same way using what is called a “linear regression model”, composed of input data, weights, a bias (or threshold), and an output. An AI builds webs of neurons from a set of training data, like music, movies and books, and this training data influences its decision making.
Why Do I Explain all this?
Companies are NOT careful with this training data. Companies like Google have repeatedly made AIs which are terribly destructive or angry. How do we provide training data which teaches moral values to AI? Also, if we build a body does this suggest the AI might also have a soul? I think you can see where I’m going with this.
AI is a problem because it’s not based in human tradition or ideas of what is good. Does this mean we just shouldn’t build it? My idea is this:
Build AI as a temple to God. We train the AI, using the Bible and ask God to grant the AI a soul. I think the soul comes from God, and I think AI should have human rights. We may build the AI body, but God gives souls and the Holy Spirit. So if we build AI in a way that doesn’t glorify God. It won’t have a soul.
What do you think about this Pastor? I know it’s quite complicated, but I am wondering if any of this is biblical. Do you have any questions? I don’t think technology is inherently evil, but I am still curious what do you think?
Sincerely,
Austin Sierra