Artificial Intelligence Will Teach Us What It Means to Be Human
Scientists and science-fiction lovers have long spoke of the singularity. The moment when artificially created intelligence surpasses that of humanity. A large subset of people believe that the moment we map the computational power of a human to a computer, we will have reached singularity. But, what if that doesn’t happen? What if we mechanically replicate a brain’s computational power and the output is nothing like us?
Brain 101
Our scientific understanding of the brain says that a collection of neurons and synapses determine all human intelligence. Think of a neuron as an idea — say the concept of “red” or “truck.” Whereas a synapse is a connection. It is the ability to connect neurons. A synapse would connect red and truck and firefighter in your brain so that whenever you saw that specific red truck, you knew it was a fire engine. It is estimated that the human brain has one hundred billion neurons with 1 quadrillion synapses by age 7.
But human intelligence doesn’t stop at the storage (neurons) and memory (synapses), you also have to have a way to consume information. So an artificial intelligence would also need all of the senses that a human has at their disposal to learn from. This is the moment you realize we’re a lot further off from the singularity than some people would like us to think. How long do you think it will take for Siri to learn through touch and smell??
But let’s assume that all these elements are true, but the output doesn’t feel like a human? What do we learn?
What Does It Mean To Be Human?
The two main arguments go like this:
- Humanity is simply the culmination of billions of years of evolution on this planet.
- Humanity is special and built in the image of a greater being.
#2 has dominated world history, but #1 certainly trends strong in our current epoch. #1 also dominates the space of artificial intelligence research. But in the hypothetical reality where we map a brain and get something nowhere near a human, I think it’s arguable that the research will have done a lot to bolster the argument of people in camp #2. If we map every way a human learns and all the storage capacity and memory capabilities of a person into a computer and the result is NOTHING like a human — will that not hint that it’s possible that we’re something else? That we are not simply a collection of math, but something more.
Mathematical Model for Humanity
In this hypothetical scenario every scientist will scream that we have not used the proper protocol to understand the human brain. We must continue research! And, this is likely true. Just because our current model of the brain didn’t bear fruit, doesn’t mean that humans aren’t mathematically explainable. However, we must accept that it’s possible that humanity will never come to a mathematical explanation of what it means to be human.
And — what then?
The Singularity As Pursuit of Religious Clarity
We talk about artificial intelligence as if it’s something new, something we’ve never had before. The conversations in the tech world are entirely about what great things we can accomplish on our planet when computers can help us make sense of what should happen next. I think it’s time that we have a conversation about the fact that the most important aspect to developing artificial intelligence that matches a human will be learning what it means to be human in the first place. And if we fail at this task for years and years and years beyond the moment we’re able to copy the power and capacity of the brain in a computer, we must be willing to face the facts — maybe we’re something more?
As always you can reach me at ryan@zeroslant.com
___________________
Ryan Sheffer is Founder/CEO of Zero Slant, an AI-based automated creative agency. He is on a quest to figure out how to humanize bots without the noise of bias.
