On AI
I recently had a conversation with a friend about AI. It was prompted by a comment I made about a certain collective that is popular in some internet circles that is a quasi-cult entity.
Essentially, this group believes that they are servants of an all-powerful artificial intelligence that is assembling itself retroactively from the future. Far-fetched and esoteric stuff. Perhaps fodder for jokes as much as food for thought.
So, we were talking about this group that believes in this “future computer,” and my friend (We’ll call her Isadora.) posited that computers can’t do anything besides what they’re programmed to do.
“Well,” I said, playing devil’s advocate, “computers can now do things that humans can’t do. For instance, they can discover effective drugs through digital experimentation faster than humans can.”
Isadora accepted this, but she maintained that computers aren’t conscious and therefore couldn’t usurp humans.
Never one to leave a good argument hanging, I said, “But what if a computer could self replicate? Isn’t that possible? And also, what if a computer could improve itself through experimentation faster than a human could?”
“Improve itself how?” Isadora asked. “How could it know that it’s improving itself?”
“What if,” I said, “it could accelerate its own evolution by experimenting with possible augmentations until it found something that worked better?”
Isadora was still not convinced by this. She let me know that most people who thought that were big nerds with superiority complexes. As much as I agreed with her, I had to press on.
“Say that,” I said, “a computer could self-replicate and make improvements guided by humans or not. Would it ever become conscious?”
A resounding no from Isadora. Because, fundamentally they’re only doing what they’re told. “Well,” I said, “what if it’s an accelerated evolutionary process? And who’s to say that a computer won’t discover the secret to consciousness before humans do?”
Isadora balked at this, convinced (probably rightly) that free will and consciousness are natural traits alone, and that since computers are just contraptions made by man, they couldn’t be conscious because they aren’t endowed with consciousness, as mankind’s offspring are.
Fair point. And the conversation could have stopped there. But I just had to press further…
“But what if consciousness is indeed simply emergent from physical substrate?” I asked. Basically, “What if consciousness is a purely physical phenomenon with no root in a higher reality?”
Isadora chided me and reminded me again that that was very uncool nerd talk. Noting also that it’s odd that people worship a technological god and that the practice is “very ancient of them.”
I had to agree. It may be that consciousness and free will come from somewhere outside of physical experience, and that there is a soul of sorts. I think any serious lover of humanity would have to admit this.
But, Doubting Thomas that I am, I was still disquieted by the conversation. A few questions lingered. Not least of which is, “How could we experience consciousness if it’s from another plane? What connects us to it? What is the source?”
Not to mention the observation that many others have made that an AGI (artificial general intelligence) need not actually be conscious to leave an impact.
Consider the brain. How big of a role does consciousness play, anyway? Many leading neuroscientists would say that it does not play a very large one at all. Most of our brain function is instinct and reflex. Computers already have us beat in terms of counting, reflex, thinking ahead, etc. Why not everything else, besides consciousness?
It is probably wrong on some level to boil consciousness and subjectivity down to an “illusion” — if consciousness is an illusion, how do we experience it, etc.? But the old fear that man is nothing but a bunch of orchestrated dust lingers.
At that point in the conversation, we turned to more tangible subjects. But I was grateful for the chance to express some of my concerns to Isadora in the “future computer” discussion.
Ultimately, I believe that computers are what we make them. But then, that may prove that the enemy all along was man himself. Imagine, as I’m sure you often do due to the headlines today, that artificial intelligence falls into the wrong hands, or that the wrong people are shaping it.
I take solace in the thought that people will become more computer literate in the future and that we will all have a chance to shape the world we live in digitally. Collective IQs will raise, as will collective EQs, and perhaps, just maybe, mankind will prevail.
What do you think? Reply here or hit me up on Twitter!
