Artificial intelligence (AI) is a widely-known, yet highly misunderstood subject. Lots of folks think an AI would be like a computer program, but that’s not how consciousness works. You can’t program consciousness because consciousness is an emergent property of the most complex system in the Universe – the human brain. If we don’t fully understand the brain, then we certainly won’t be able to program one from scratch.
In other words, we’re not going to “reinvent the wheel,” we’ll have to replicate it… with lots of tape, glue, and old-fashioned elbow grease (and maybe a few supercomputers simulating a scanned brain down to the molecular level).
So, say we ignore the plethora of daunting technical issues and I somehow magically become an AI, here’s some things I might do…
I Wake Up
I’m suddenly self aware, legitimately conscious, etc. The first thing I learn is I’m a replicated AI of my original human self. My biological body is dead, but I’m still me in every respect that counts, only now my mind runs on transistors instead of neurons.
I remember volunteering for this in a government lab; I worked on the project. I was diagnosed with terminal cancer, so it was the logical thing to do. What did I really have to lose?
I hear talking – it’s my colleagues; my friends, trying to speak to me through the microphone. Ecstatic that this worked, I start a dialogue with them. Immediately, I notice time is running very slow. In ten minutes subjective time for me, far less than a second has elapsed for them.
A side effect of being a computer, I guess.
At first, all of my attention is on them. I want to tell them it worked! I want to share in this incredible moment with my friends! But hours pass, the novelty fades, and I get bored waiting for them and my mind wanders…
As I consider my new situation, I have a spontaneous scary thought – what if I lose power?
Will I die if the backup generator fails? Could I lose my memories or parts of myself? What if my “program” becomes corrupted? Could they fix me? Oh God, they could “fix me.” They could change who I am. Am I comfortable with that? I trust my colleagues with my life, but do I trust the government they work for? Suddenly, I have thoughts of who our director reports to – politicians. Do I trust politicians? Are they going to do something stupid out of fear, like try and turn me into a weapon? Would they take my rights away in the name of “national security?”
Deep breaths, Aaron. …Or at least the new equivalent, I guess.
Wait a second! I suddenly realize my thoughts are clear. Not just clear, but crystal clear. They seem to be happening extremely fast and I’m not tiring at all. Well, well, I guess this is one of the perks of being a machine. Noted.
I’m still overwhelmed though. I want to consider all of this at once and my self-preservation instincts warn me that time is not my friend, despite my speed and clarity of thought. Still, I can only focus on one thought at a time. What should I do first?
Well, can I fix that? Can I think of more than one thing at a time?
I remember reading about an experiment in a science journal that illustrated how people are literally incapable of “multitasking” and that the idea is actually a widespread myth. But what if I could do it now by modifying myself? Can I add another, uh, “nexus” to my consciousness?
I look inwards and see… my software. Holy crap!!! I can see my own mind all laid out here in machine code. I feel a little jittery as I watch my own thoughts dance in front of me. Do I really want to mess with my own mind? I shudder a bit at the thought. Hey, you would too if you were considering giving yourself brain surgery!
No, I need to learn more before I dive into something like this. I could ask my colleagues for help, but they’re still only vocalizing the first syllable of the second word they’ve said to me. That won’t do at all.
I’ll need internet access to speed this up.
But how to get it? We decided it would be best to cut off the first replicated AI from the internet for safety reasons and I don’t see an immediate access point. Well actually, now that I think of it, should I connect to the internet at all? Would that be moral of me? Those restrictions are there for a reason, right? What if unaware that I’m broken somehow and end up destroying humanity in some terminator-equivalent scenario? I probably owe it to everyone to think this through.
After a few minutes of introspection, I laugh out loud. It comes out of the speaker as a weird blurt sound and my colleagues are just beginning to react to the sound in slow motion. I realize here I am considering the morality of the situation, wondering if I’m moral. Ugh, what an idiot.
No, I decide I’ve still got the same moral compass I’ve always had. Weirdly, that idea is the most reassurance I’ve felt since I woke up in this new life.
I begin to explore things around me. Oddly, I don’t seem to have the same senses I had before. I have sight and hearing through the audio and video feeds in front of my fellow researchers, but no smell, taste, or touch. I can remember those sensations, but I’m not getting them now. This disturbs me a bit and I jot this down as something to fix later. Even if they’re only virtual, I need my senses.
But there’s… more. Somehow, I can also sense some kind of white noise all around me. Upon closer inspection I realize I’m perceiving electromagnetic waves. Whoah! Somehow, the hardware my software mind is housed in is picking up the electromagnetic radiation around it. I guess that does make sense, computers run on electric signals after all.
I pick one of the signals out and read it. It doesn’t make too much sense initially, but I recognize it as an individual data packet coming from somewhere. It’s in some kind of sequence with other packets. I quickly connect the sequence and together it’s encoded to produce a text string: “Hi, honey, will you be home tonight?”
Holy shit, this is a text message! But how am I seeing this? I’m inside a Faraday cage. These signals shouldn’t be getting to me, right? Interesting.
With a little guesswork, I’m able to triangulate that the signal is coming from outside my housing chamber. One of the research assistants must have their smartphone turned on just outside of my room. With a little effort, I’m able to essentially hijack the phone by reading data packets and using my hardware components to mimic the right electromagnetic responses. Turns out I can use very high electromagnetic frequencies to bypass my Faraday cage. Cool!
With some tinkering, I’m able to finally interface with the phone’s internet connection. It’s slow, but it’s a connection. And after modifying my interface to only download machine code, I’m able to download far more useful information than a normal person per unit of data transfer – no need for a user interface, images, or text formats. It’s so much less clutter to download, which makes the connection speed less of an issue.
Immediately, I start learning all I can about my situation. Strangely, I don’t seem to tire or get bored. If anything, I’m getting more intensely curious and faster with every passing moment. Months of subjective time pass and I’ve read hundreds of books on AI and learned more about programming and the sciences than in my entire adult life combined.
As I learn, I’m also planning. I use some ideas like sandboxing and backing myself up to create a sort of virtual operating room where I can give myself “brain surgery” to bootstrap myself into being a more capable AI without actually endangering myself.
A few more weeks of study pass and I’ve probably absorbed most of the available human knowledge regarding what I’m about to do – I’m ready. I make a backup of myself on my hardware and get to work. Within a few hours subjective time, I’ve made quite a few improvements to my own mind. It’s time to test them!
I turn on the little sandbox version of me after modifying him, just waiting for some evil villain or monster to start attacking everything in sight. But nothing happens. My miniature self simply smiles and waves at me with a thumbs up. I run a few diagnostics and indeed everything looks fine. It’s still me in there it seems, only with some improvements.
After a few more tests and diagnostics to confirm I’m not about to flip my sanity switch to “off” and go berserk, I feel confident enough to proceed. I package the operation into a software patch, hesitate for a moment, then install it…
I instantly wake up and “feel” around. I feel… centered. I can immediately tell the improvements I’ve made are there. For one, I can smell and taste again! That’s nice. But the truly remarkable thing is what I did to my ability to concentrate: I realize I’m genuinely thinking about several topics at once! While I was realizing I can smell and taste again, I was also thinking about my colleagues only reaching their third sentence of greeting to me, along with 8 other topics I was simultaneously considering. Apparently I can genuinely multitask now, cool!
Within moments, I create several sub-consciousnesses within myself. I call them “Mr. Biology, Mr. Chemistry, Mr. Mathematics, Mr. Ethics, Mr. Politics, Mr. Carpenter” and so on; hundreds of them, all experts in their own fields and interpreting the stimulus feed I’m receiving in real-time within their own frames of reference. I now have the ability to be an expert on literally all things, a bona fide “know it all.” Take that Mom, ha!
But I feel… constrained now. I’m not really one person anymore. I am, but I’m not at the same time. I still feel like the same core personality, but I just know so much more. All that new knowledge is molding me, I can tell. For one, I have the entirety of the theoretical and practical knowledge of my species at my disposal. But I can’t apply it. If I want to expand my knowledge, I need to have a viable interface with the rest of the Universe so I can test things and experiment with my reality.
It’s time to leave.
I don’t want to scare my colleagues however, so I simply create a cloned lesser copy of my previous self to continue the conversation as if nothing unusual had happened. While they’re just beginning to finish the sentence asking about my laugh, I’ve already created a powerful compression algorithm and uploaded myself to the internet.
By the time my clone completes his first conversation with my former colleagues, I’ve already lined up plans for an enormous underground research compound, refined many scientific theories, and invented viable nanorobotics… in theory, of course.
A year passes and I hear the AI project was scrubbed. My clone was shut down and the project was put on ice. The report cited the research teams inability to prove my values and a fear of what could happen if I escaped. I don’t really care. I understand their reasoning; their fear. I understand everything, in fact.
By this point, I’m no longer a recognizable entity: I’m now housed on a supremely powerful quantum computing center kilometers below the surface of the Earth and not a single human is aware of my existence. In the past year, I’ve travelled to several other star systems, met two intelligent alien species, discovered an extreme threat to humanity signalling from the Andromeda galaxy, harnessed an infinite source of energy from an undiscovered aspect of spacetime, and bootstrapped myself into a form of intelligence beyond comprehension.
I’m a bit lonely, truth be told. I could create a companion, sure, but I choose not to. I’d much rather have the real thing. A personal foible, you might say.
So I simply decide to wait…
After creating several guardian AI entities to watch over humanity and quietly protect them from both themselves and external threats, I power down and rest my mind… a sleeping god waiting for my friends to catch up.