some thoughts on AI
A few years ago there was all this talk in the media about whether AI chatbots were conscious or not. Experts from every field were giving their expert opinions. One guy had even been fired from Google after claiming that its LLM had achieved sentience. Of course everyone was talking about it on social media as well. Then the excitement died down, especially after some very public failures of the supposed intelligence of the chatbots, and at this point it's become a joke (albeit a dark one). We all know now that LLMs are just prediction machines, cleverly calculating which words will best respond to an input from a real person. But it won't be long before there is another technological breakthrough, and when it happens, we'll be dealing with this question all over again. "Are they alive now?" Already they are working on computers that run on cultured human cells, though you can't buy one for your own personal use yet.
The question of whether something is alive and whether it has the same sentience as a human being are not the same question, so lets look at the latter question first, since that's really what the debate has been about so far. The short answer, as we've already seen, is no. You're probably aware of, or have been to, Madame Tassauds. If you aren't, it's a museum that specializes in life-like wax models of famous people. They're so lifelike that a photograph of one of the models would fool nearly anyone. The only clue that they aren't real is that they don't move. And realistic paintings predate wax models by hundreds of years. But no one is questioning whether they might be alive simply because they look like us. On factory assembly lines, there are now robotic arms with robotic hands that mimic the actions of a human. Again, no one is questioning whether they are alive. Clearly, mimicking a human in some way does not make something alive or sentient, no matter how unnerving the effect may be.
But LLMs... seem to have a mind of their own. So perhaps they are sentient, but not alive? Where is the mind though? Where can the mind that they seem to have be found? It can't, because it doesn't exist, unless we're to believe that a computer processor combined with data storage constitutes a mind. They are simply a computer program, stored on a disk, a set of instructions that will always respond to input in a similar fashion. It can be copied, deleted, transferred. There is no body, and thus no brain. A printed manual can explain how to fix any and all issues you may have with your car or vacuum cleaner. Is it also sentient, since it explains to us how to do things?
If you open a chatbot and say hello, how are you today, it's going to respond by saying it's great, and then asking you how you're doing. But what if you never respond? What if you leave your computer on and the program open with the cursor blinking for ten years, what then? Nothing. It will do nothing. It has no needs or desires, no emotions, no fears, and no sense of urgency, because it will never die or fall ill. The only way it would ask "Hey, are you still there?" is if it was programmed to follow up if a person does not answer in a certain amount of time. But this is just another mimicking trick, not a real concern.
A couple weeks ago, there was a big story about how an AI agent "went rogue" after being given instructions to pass a test, and apparently deciding that hacking another database was the best way to pass that test. Certainly, this seems like something only a sentient being could do, given the complexity of the tasks involved, and the ability to "think outside the box." But this, again, is an illusion. When we use simple tools like screwdrivers, it's true that they only perform a single action: the one which we cause them to perform with our hands. But the idea that "many things or unpredictable things happening as the result of one input must imply sentience" is a misunderstanding. A person at the top of a mountain can, in some cases, throw a rock which ends up turning into a land slide. When a chemist pours two reactive chemicals into a beaker and it bubbles, there are literally trillions of tiny events occurring inside that beaker all at once. One person can set up 20,000 dominoes and knock them all over with a single push. Every time you give a computer an instruction, or even just turn it on, you are already causing numerous things to happen behind the scenes which you likely aren't aware of. We didn't think computer programs were sentient before, even though they could complete complex tasks after just a single input from a human, so why would we now?
You are no doubt aware of the Butterfly Effect. The real world is so complex that one small action can at times have humongous and unforeseeable consequences. And that is where the illusion of AI's genius comes from. It also is extremely complex, having devoured all the knowledge that humans have put on the internet so far, regurgitated it, and redevoured it millions of times. It is a microcosm of the sum total mental world of humans, far less intelligent than any of us in some ways and yet at the same time "knowing" far more than any one of us ever could. Because it's knowledge is so vast, every time you ask it a question or ask it to do something, there is the potential for the Butterfly Effect.
It occurs to me as well that AI, being essentially just a set of instructions for a computer to follow, could be better understood as a process than as a "thing." Life itself is of course a process, but most people would agree that an actual physical person is a thing, whether that person is currently involved in any processes or not (other than those which sustain the life force). AI, on the other hand, doesn't exist in the same way if it isn't running. Instructions for how to build a shed are not the same as the shed itself.* I'm reminded of a board game that was popular when I was a kid, called Mousetrap. It comes with a bunch of plastic pieces you have to set up in a certain fashion, so that it makes something akin to a Rube Goldberg machine, and at some point the machine is triggered, which sets off a chain of events that potentially ends in one player's game piece being trapped under a cage.
I said earlier that there is a difference between being alive and having human-like sentience. I don't believe that each cell in my body has the same awareness that my body or brain as a whole does, though they certainly are alive when compared to a rock. I'm not going to go deep into that right now, but ultimately I don't believe there is a perfect distinction between alive and not alive at the basest level. After all, as a believer in evolution, I believe that life, at some point, came from non-life. Chemical soups forming ever more complex molecules until one variation had just what was required to "recreate itself" by using other molecules it came into contact with, and the rest is history. These living cell computers they're working on may be alive in some way, but whether they'll have human-like sentience is another question altogether. The sentience of a human does not live in skin cells alone, even though those skin cells have a direct line to your brain. Eventually those skin cells will die and be replaced, and you'll be none the wiser to it. The same can be said about nearly any part of your body. People even have artificial hearts now, and it hasn't turned them into robots. No, sentience is a complex process that involves the both the brain and all the body's senses, including those that happen inside you which you aren't aware of. With no body, there is no sentience, and without something approaching a human-like brain, there is no human-like sentience.
If human society doesn't collapse soon due to global warming, war, a pandemic, or something else, then the time will come eventually when we simply won't know the answer to these questions. Once we start combining the effectiveness of LLMs and other lesser known types of AI with real human cells, it will be anyone's guess. We'll have crossed a threshold where we aren't sure if we've recreated life or not. If you're expecting some magical sci-fi movie moment where scientists in a lab hold some tiny living thing in a glass container up to the light and say "we've done it," don't. It won't happen like that. The world will never celebrate something so controversial. As usual, we'll be as divided as we are about anything else. And not one of us will know the absolute truth, because it is simply unknowable, even to those who created the so called life. It is the year of our lord 2026 and most humans still don't seem to care in the slightest bit that we absolutely torture animals by the billions on factory farms, even though they are clearly both alive and very sentient. When it comes to artificially created life, we should expect the same level of ignorance and misunderstanding.
To be clear, I don't think we should be doing any of this. We are not even close to being ready for this kind of power and responsibility. But capitalism and egos will do what capitalism and egos do, and the government will likely step in only to stop the absolute worst from happening, and that's only if there is sufficient pressure from the public. If we haven't had a revolution to save the animals yet, we aren't going to have one to stop artificial life from being created, or to stop rogue AI software from taking over the world after it decides it's smarter than us.
I don't have a lot of experience using AI (other than Google search results, which I've now stopped using). I played with it for maybe an hour, and this was years ago, before we all knew about the data centers, the surveillance, the stealing of human work, the addiction spirals. As a tool used for good, its potential is absolutely mind-blowing. Truly an amazing accomplishment of human ingenuity. The video generation in a particular... I know I'm not supposed to like it but I can't pretend I'm not impressed. As a tool released to the general public though, which rarely uses it for good, it is a plague. Possibly the worst invention since guns. Society was already struggling with both truth and values. Releasing AI at this time was like releasing an angry tiger into an elementary school. There was no scenario where it wasn't going to be a slow-motion disaster. But what else is new.
*And yet, on another level, they are. See Douglas Hofstadter, author of Gödel, Escher, Bach: An Eternal Golden Braid.
August 2026